Glycobiology Advance Access originally published online on November 13, 2007
Glycobiology 2008 18(2):177-186; doi:10.1093/glycob/cwm126
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Characterization of the 
-2,8-polysialyltransferase from Neisseria meningitidis with synthetic acceptors, and the development of a self-priming polysialyltransferase fusion enzyme
National Research Council Canada, Institute for Biological Sciences, Glycobiology Program, 100 Sussex Drive, Ottawa, ON, K1A 0R6, Canada
1 To whom correspondence should be addressed: Tel: +1-613-952-4299; Fax: +1-613-941-1327; e-mail: warren.wakarchuk{at}nrc-cnrc.gc.ca
Received on July 27, 2007; revised on November 5, 2007; accepted on November 6, 2007
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Glycoconjugates containing polysialic acid have many biological activities and represent target molecules for therapeutic interventions. Enzymatic synthesis of these glycoconjugates should give access to these important molecules to evaluate their potential. The polysialyltransferases from both Neisseria meningitidis and Escherichia coli were cloned and expressed as recombinant proteins in E. coli. We have used synthetic acceptors to probe the acceptor requirement of these enzymes and to examine the basic enzymology. The minimum number of sialic acid residues (Neu5Ac) on the acceptor for activity in vitro was shown to be 2 for both enzymes, but a large increase in activity was seen if the acceptor had three Neu5Ac residues. The polysialyltransferase from N. meningitidis generated longer reaction products than the enzyme from E. coli on FCHASE acceptors. Examination of the products showed them to be a heterogeneous mixture, but products with >50 Neu5Ac residues could be seen using capillary zone electrophoresis analyses. In addition we made fusion proteins of these polysialyltransferase enzymes with the bifunctional
-2,3/-2,8-sialyltransferase from Campylobacter jejuni to create self priming polysialyltransferases. These bifunctional sialyltransferases utilized various synthetic disaccharide acceptors with a terminal galactose, and we demonstrate here that the PST enzyme from N. meningitidis and its fusion protein with the C. jejuni sialyltransferase can be used to create polysialic acid on O-linked glycopeptides.
Key words:
Fusion enzyme
/
glycosyltransferase
/
polysialic acid
/
-2,8-polysialyltransferase
/
synthetic acceptor
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Some bacterial pathogens that invade the mammalian host have taken advantage of the presence of sialic acid glycoconjugates on the host. These bacteria display some of these same carbohydrate chains on bacterial cell surfaces, and indeed a role for these carbohydrates in pathogenesis has been demonstrated (Moran et al. 1996
; Kahler and Stephens 1998). Capsular polysaccharides from group B Neisseria meningitidis, Escherichia coli K1, Moraxella nonliquifaciens, and Mannheimia haemolytica A2 (formerly known as Pasteurella haemolytica A2) include sialic acid homopolymers of -2,8-linked Neu5Ac. This is a molecular mimic of the mammalian polysialic acid (PSA) structure seen mainly on the neural cell adhesion molecule (NCAM), a brain specific protein integral to neuronal function. Other bacterial capsules have variations of PSA such as homopolymers of -2,9-linked residues, copolymers in which the linkage is mixed -2,8/-2,9, and finally polymers in which other sugars are included, as in the group B Streptococcus agalactiae, N. meningitidis group Y and W (Table I). It is thought that the presence of these carbohydrate mimics allows the pathogens to escape detection by the immune system since these molecules are not considered foreign. Further, the presence of these carbohydrates presents a physical barrier for the killing action of serum complement (Vogel et al. 1996). The -2,8-polysialic acid capsules are also required for crossing the blood–brain barrier to cause neuroinvasive disease in the case of E. coli, N. meningitidis, and the animal pathogen M. haemolytica (Silver and Vimr 1990).
|
Because of the prominent role of these sialylated oligosaccharides in the biology of both the pathogen and the host, the enzymes involved in their biosynthesis in bacterial pathogens and in humans represent potential targets for therapeutic intervention and catalysts to produce biologically active glycoconjugates. For example, the homopolymer of
-2,9-linked Neu5Ac residues can be used as a vaccine for protection against group C meningococci (Girard et al. 2006), and efforts are underway to develop such a vaccine based on the group B meningococcal homopolymer of -2,8-linked Neu5Ac (Pon et al. 1997). Recently it has been shown that the chemical coupling of -2,8-polysialic acid to therapeutic proteins can extend their half-life in vivo, making them in effect more potent (Gregoriadis et al. 2005). Additionally, the role of PSA in repair of the central nervous system has been examined in the context of using gene transfer to stimulate the production of PSA-NCAM and thus the regeneration of CNS axons (Franz et al. 2005; El Maarouf et al. 2006). Access to all of these PSA conjugates would be improved if their in vitro synthesis could be optimized with suitable enzymes.
To date there has been little detailed work on the fundamental aspects of sialyltransferase enzymology from bacterial pathogens. We and others (Gilbert et al. 1996; 2000; Chiu et al. 2004; Yu et al. 2005) have shown that it is possible to express, purify, and crystallize some of those enzymes responsible for lipooligosaccharise (LOS) sialylation. However no such work has been done with the enzymes involved in the generation of the sialic acid homopolymeric capsules. The genetic loci for the PSA capsule production have been identified in both E. coli and N. meningitidis, and some work has been done on the recombinant enzymes (NeuS) from E. coli K1, and K92 (Cho and Troy 1994; Shen et al. 1999), but again no detailed enzymology on the isolated sialyltransferase has been reported. The study of the enzymology has been hampered by the inability to isolate soluble enzyme and the lack of a simple synthetic acceptor from which enzymology data could be obtained. We have been able to clone active fusion proteins of both the E. coli K1 and N. meningitidis group B PST enzymes. We report here the characterization of the isolated recombinant enzymes using synthetic acceptors generated chemi-enzymatically. We also present data on a multifunctional fusion protein which functions as a self-priming polysialyltransferase that has potential for the in vitro synthesis of various PSA glycoconjugates which may be of therapeutic use.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Comparison of N. meningitidis and E. coli polysialyltransferases
Both the genes for these polysialyltransferases are from the glycosyltransferase family GT-38 (Coutinho et al. 2003
) which contains only bacterial members. The two proteins show only 33% identity (Figure 1) even though they make identical structures in vivo. The proteins were expressed by themselves and as fusion proteins with the maltose binding protein (MalE) of E. coli which lead to the production of a more soluble protein. The enzymes are not truly soluble, as they can be isolated in both the supernatant and the pellet fraction if the lysates are centrifuged at 26,000 x g. With a lower speed centrifugation (1000 x g) more of the enzymes stay in solution and can be purified on the amylose resin with excellent recovery (Figure 2).
|
|
The activity of the enzymes was determined using the synthetic acceptors derived from lactose-aminophenyl-(6-(5-(fluorescein-carboxamido)-hexanoic acid amide) (Lac-FCHASE), GM3-FCHASE, GD3-FCHASE, GT3-FCHASE, and GQ3-FCHASE (Table II). These acceptors are named for the carbohydrate portion of the glycoside which resembles the head of a ganglioside. The enzymatic synthesis of these types of acceptors has been described previously (Blixt et al. 2005
). With these substrates the activity of these enzymes was shown to increase upon changing the number of Neu5Ac residues from 2 to 3. With the PSTNm enzyme the specific activity increased threefold comparing GT3-FCHASE with GD3-FCHASE. There was no detectable activity when only one Neu5Ac residue was present. In our hands the Neisseria enzyme is more active toward these small synthetic substrates than the E. coli enzyme; in fact, we saw only oligosialic acid production and never saw long polymers with the PSTEc construct so we focused our efforts on the Neisseria enzyme (PSTNm).
|
PST does not appear to follow Michaelis–Menten kinetics when using a CE-based assay
We made several attempts to measure kinetic parameters of PSTEc and PSTNm using the synthetic acceptors described above. The CE-based assay however does not permit the determination of substrate interconversion during the assays. Hence we could only follow the production of all products as a function of their combined peak area on electropherograms (Figure 3), but we could not look at specific products formed as the result of generating a better acceptor as the reaction proceeded. Trying to fit these data with the Michaelis–Menten equation did not show a very good fit (Figure 4). For this reason we were only able to estimate the apparent Km for GD3-FCHASE as 120 µM with PSTNm and 145 µM with PSTEc, along with an apparent Km for CMP-Neu5Ac of 1.4 mM for PSTNm and 3.2 mM for PSTEc.
|
Ex 488 nm, with emission filter at 520 nm). Each peak corresponds to the addition of one Neu5Ac residue.
|
When performing assays to examine kinetic parameters, we found that PSTNm showed acceptor inhibition with higher concentrations of GD3-FCHASE. At concentrations over 2 mM the productivity of the enzyme decreased significantly (data not shown). In some experiments we observed degradation of the starting material, which was more significant at high acceptor concentrations (5 mM), but this was not always reproducible.
Addition of Mg2+ results in a fourfold increase in the activity of PSTNm
In order to ensure the best possible activity, assays were optimized for metal and buffer composition and pH optimum. The enzyme showed a broad pH optimum, but had the highest activity around pH 8 (data not shown). The enzyme showed a strong preference for Mg2+ compared to other divalent cations (data not shown). After removing endogenous metals by dialysis with EDTA, the enzyme was titrated with MgCl2 to determine the optimum concentration required for activity. N. meningitidis PST exhibits a fourfold increase in activity with the addition of 20 mM MgCl2.
Functional characterization of the CST–PST fusion enzymes
In order to use PST as a synthetic tool, it is necessary to provide it with a primer of two sialic acid units. We chose to create a unique fusion protein with the CST-II enzyme from Campylobacter jejuni which can generate this primer. Fusions were made between the CST-II (Ile53Gly) from C. jejuni OH4384 (Gilbert et al. 2002) and both E. coli (CST–PSTEc) and N. meningitidis (CST–PSTNm) polysialyltransferases. We used both of the fusions and compared both Lac-FCHASE and T-Ag-FCHASE in a time course assay (Figure 5). Traces from these assays clearly show a qualitative difference in the product distribution between these CST–PST fusions. Even though the acceptor is consumed completely with CST–PSTEc, there are no long polymers formed that can be seen by CE analysis. With CST–PSTNm the formation of long polymers occurs from the earliest time points, and in longer reactions, products greater than DP-50 can be seen. This reflects the observation we made with the MalE fusion version of PSTEc that it does not make long polymers on FCHASE-aminophenylglycosides in vitro (Figure 6).
|
), +1 NeuAc (
), +2 NeuAc (
), +3 NeuAc (
), +4 NeuAc (
), <4 NeuAc (
) Panel A is CST-PSTEc with Lac-FCHASE, panel B is CST-PSTNm with Lac-FCHASE, Panel C is CST-PSTEc with T-Ag-FCHASE, Panel D is CST-PSTNm with T-Ag-FCHASE.
|
The CST–PSTNm enzyme was shown to be a very good catalyst for making long polymers starting from simple galactosides. A comparison of acceptors was performed with lactose, N-acetyllactosamine, Gal-β-1,3-GalNAc-
(T-Ag) disaccharide acceptors, and two glycopeptide acceptors containing an O-linked Gal-β-1,3-GalNAc--Ser moiety. We observed that Lac- and LacNAc-FCHASE were equally good acceptors (data not shown), and that T-Ag-FCHASE was a slightly weaker acceptor. In a time course experiment with CST–PSTEc and CST–PSTNm we showed that with a single addition of enzyme and fixed donor concentration essentially all of the Lac-FCHASE is consumed but only 72% of the T-Ag material is consumed (data not shown). The glycopeptides were also very good acceptors for the CST–PSTNm fusion protein, and long PSA polymers could be seen in CE analysis of these reactions.
Production of CST–PSTEc and CST–PSTNm
Partial purification of the fusion proteins between CST-II and either PST was achieved by removing other proteins by binding them to a HiPrepQ FF column, followed by ammonium sulfate precipitation of the flow through fraction. It is not possible to know the exact level of enrichment, but using gel densitometry we showed that CST–PST represents around 12–14% of the total protein (Figure 2). The protein can also be recovered by the ultracentrifugation of this flow through fraction. This would suggest that the enzyme is present as a micellar or aggregated preparation.
Comparison of the fusion protein with mixtures of the individual parent enzymes
In order to evaluate if the fusion protein functions more efficiently than the two parent enzymes mixed together we performed experiments where we controlled the stoichiometry of the individual sialyltransferases. Because the fusion protein could not be purified to homogeneity, we estimated how much enzyme was in the preparations with densitometry of SDS–PAGE analysis. We then made reaction mixtures with an equivalent amount of the purified parent enzymes (CST-II and PSTNm). In this comparison (Figure 7), it appears that the fusion protein is a more efficient catalyst making more than four times as much polysialic acid product on the T-Ag-containing IFN2b glycopeptide (20–40% PSA versus 3–4% PSA for the combination), and this is also shown with the monosialylated T-Ag-glycopeptide acceptor, although here only a twofold increase is seen (60–75% PSA versus 35–45% PSA). What is significant is that the level of disialylated intermediate is much higher in the reaction with the separate enzymes. We applied ANOVA (t-test) to the production of the PSA in these experiments and found that level of PSA formed in the different reactions with separate enzymes and those with the fusion was statistically significant. When the acceptor was the monosialylated T-Ag-glycopeptide acceptor P = 0.002 for PSA formation in this series of experiments. When the acceptor was T-Ag-containing IFN2b glycopeptide P = 0.03 for PSA formation in this series of experiments. Another significant observation is that at very short reaction times, the fusion protein generates PSA material from both types of acceptors, whereas the separate enzymes do not. With longer incubation times CST–PSTNm will convert >50% of the T-Ag-glycopeptide to polysialylated product (Figure 7), and if the peptide is already monosialylated then the reaction goes to 
75% polysialylated material (data not shown). We have also used a second glycopeptide acceptor, this time derived from the O-linked glycosylation site of the granulocyte colony stimulating factor (G-CSF Thr-134) (DeFrees et al. 2006). The data on the extent of the polysialylation of these two peptides are very similar so we have not shown all the data for both peptide acceptors.
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Polysialyltransferases are found in both humans and some human bacterial pathogens. The human PST enzymes are classified in the CAZY family GT-29 along with all the other human sialyltransferases because of the presence of various sialyl-motif sequences common to all mammalian sialyltransferases. The bacterial PST enzymes are found in a unique GT-38 family and are not sequence-related to any other sialyltransferases, and have no apparent sialyl-motif sequences. This difference in structures is accompanied by some fundamental changes in activity as well.
The human enzymes ST8SiaII (STX) and ST8SiaIV (PST) can effectively use -2,3/2,6 monosialylated acceptors (Angata and Fukuda 2003) whereas the bacterial enzymes appear to have a minimal requirement for a glycan structure containing -2,8-disialic acid as a primer for the reaction, although the in vivo primer structure is still not known. We felt it was important to revisit the minimal acceptor requirement to ensure our recombinant protein was not enzymatically compromised because of the nature of the protein fusion. There are conflicting reports about the mammalian enzymes having different divalent metal requirements and pH optima (Kitazume-Kawaguchi et al. 2001 and references within), but the PSTNm enzyme clearly has a requirement for Mg 2+ where it shows a fourfold increase in activity at 20 mM Mg2+ and a slightly basic pH optimum. The cloning and expression of soluble forms of the human PST enzymes has been described in the literature for many years (see references in Angata and Fukuda 2003), but so far large amount of the enzymes have not been available for in vitro synthesis of PSA containing glycoconjugates. Our expression and purification of the bacterial PST enzymes yields more active protein than has been seen for the mammalian enzymes, which has been the case for a number of bacterial enzymes which make human glycan structures (Gilbert et al. 1998; Blixt et al. 2001, 2005). The ease of expression of many bacterial glycosyltransferases including the bacterial PST is likely due to the absence of critical disulfide bonds and the lack of posttranslational modification which are required for many eukaryotic glycosyltransferases (Datta et al. 2001; Uemura et al. 2005).
The bacterial PSTs have been reported to be peripheral membrane proteins which have previously been found only in the membrane fraction as proteins, which we found were not active after solubilization with detergents (data not shown). We chose to use the MalE tag as it has been a useful solubilizing fusion partner for many of the glycosyltransferases we have studied, especially another sialyltransferase, LST, from N. meningitidis (Wakarchuk et al. 2001). While this fusion protein is not completely soluble at 100,000 x g, it can be isolated in multimilligram quantities and has been shown to be able to work very well in vitro for the acceptors we have tested so far. A recent manuscript (Freiberger et al. 2007) also describes a MalE fusion of the PSTNm and they have also shown the activity with our FCHASE-based acceptor. The MalE fusion produced in our laboratory has a slightly higher purity than described by Freiberger et al. (based on the SDS–PAGE profiles) and appears to be less soluble for reasons that are not clear at this time. Clearly some differences exist between the recombinant constructs and the method of enzyme preparation, which makes a direct comparison of these enzymes difficult. We have shown that the purified MalE fusion of the PST enzyme is stable enough to be stored for several weeks in the presence of glycerol at –20°C which is another property that makes this enzyme much more accessible than the mammalian ST8SiaII/IV enzymes which have not been reported to be expressed in this quantity or as stable. The mammalian enzymes have been examined mainly as chimeras with the IgG-Fc fragment which are secreted and captured on beads and assayed in that way (Angata et al. 2004).
When comparing the two bacterial PST enzymes, we found that the in vitro activity of the PSTEc was qualitatively different from the PSTNm in that we only ever saw short polymers with the PSTEc enzyme with our aminophenylglycoside-based synthetic acceptors. Both enzymes will however make polymers on the glycopeptide substrates we used, although the activity of PSTEc is again very weak (data not shown). Clearly in vivo long polymers are formed in both organisms, so the reason for this altered in vitro activity on the FCHASE-glycoside acceptors is not clear. There is some evidence that the PSTEc requires additional proteins in order to function as a PST in vivo (Andreishcheva and Vann 2006) which might explain its poor performance compared to the PSTNm.. The E. coli polysialic biosynthesis operon contains the NeuE protein which has been shown to be required for efficient polymer formation in vivo and to form a complex with NeuS (PSTEc) as well as the capsule export protein KpsC/S. There are NeuE homologues in Neisseria as well, but their function has not yet been investigated. A rather more trivial explanation for the lower activity of the MalE fusion might be that it simply prevents the PSTEc from folding correctly, so that its activity is no longer optimal. This may be the case as recombinant NeuS has been shown to be fully capable of forming polymers on various acceptors when expressed as a membrane bound protein (Cho and Troy 1994). This observation means that with both of these enzymes we may have the ability to form both shorter oligosialic acid chains and longer polymers. This opens the possibility of evaluating both types of chains for their contribution to the function of glycans.
We were unable to perform detailed kinetic assays on the PST enzyme due to the interconversion of acceptors during the assay and how we were detecting these products. What this means is that the products of the reaction are better substrates than the starting material, and therefore we cannot quantitate the level of conversion of all the products, just the starting material relative to all products. We have also investigated using the continuous assay described by Freiberger et al. (2007), but in our hands, the background from concentrations of CMP-Neu5Ac over 0.5 mM prevented us from collecting data at higher donor concentrations. Using the CE-based assay we obtained a Km which is three- to fivefold higher than what they reported, which is reflected in the much higher activity we see with concentrations of CMP-Neu5Ac around 10 mM. We are planning to investigate modified CMP-Neu5Ac donors as chain terminators as a means of examining the acceptor binding when there is only one turn over event which eliminates the substrate interconversion. We have already examined the 9-O-acetyl sialic acid, which is a chain terminator, but this donor has the drawback that the O-acetyl group hydrolyses slowly during the reaction leading to a background level of polymeric material which makes the quantitation of product difficult (data not shown). A similar approach has been used to examine the K92 PSTEc where a 9-azido-NeuAc was used as a chain terminator to examine the alternating -2,8/2,9 NeuAc linkage formed by this enzyme (Vionnet and Vann 2007). Use of these modified NeuAc donors should provide data toward the understanding of the chemical mechanism of this class of enzymes.
Because of the requirement for the disialic acid acceptor we decided to investigate enzyme fusions with the bifunctional -2,3/2,8 sialyltransferase from C. jejuni (CST-II). Our work with CST-II to make di-, and trisialylated ganglioside mimics made us think we could have a single polypeptide which could then produce polysialic acid from galactoside acceptors. There is no indication that Campylobacter has polysialic acid on any of its ganglioside mimics. Our first attempts to make such fusion proteins were successful, and we did compare the ability of both PSTEc and PSTNm to function in this construct. The CST–PSTNm fusion performed better than the CST–PSTEc fusion because of the PSTNm making longer polymers on our substrates. We investigated the ability of this fusion to make polysialylated material using a variety of simple disaccharide acceptors and showed that the nature of the linkage of the galactose containing acceptor had only a minimal effect on the activity of the enzyme. The Gal-β-1,3-GalNAc--R disaccharide is a typical O-linked glycan on many glycoproteins which would provide a convenient handle for the addition of sialic acid to the protein. We were able to show that with model glycopeptide acceptors with an O-linked disaccharide we could produce polysialylated material and the fusion does function better than the two proteins provided separately. However, we still need to improve the production of this fusion protein, as it is not as soluble as the MalE fusion of PST, and cannot yet be purified to homogeneity.
The work of Gregoriadis et al. (2005) has shown that the chemical coupling of preformed PSA chains can increase the circulating half-life of therapeutic proteins, but the addition depends on having available lysine residues, and not all residues are modified. The great advantage of using an enzyme is that we can control where the PSA will go, and this will permit the synthesis of PSA at defined sites in therapeutic proteins which are less likely to interfere with the function of that protein and are easier to characterize for regulatory agencies. Having shown that model glycopeptides are good acceptors we are now looking at in vitro protein polysialylation with these enzymes.
|
Material and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Bacterial strains, plasmids, and growth conditions
N. meningitidis 992B NRCC4726, obtained from Dr. Harold Jennings, was grown on Columbia Blood Agar at 37°C and 5% CO2. E. coli O18:K1:H7 LCDC 84-1743, obtained from Health Canada, was grown on 2YT plates at 37°C. E. coli AD202 (CGSC 7297) was used for expression of cloned enzymes. Recombinant E. coli strains were grown in 2YT broth with 150 µg/mL ampicillin at 37°C for 2 h and then induced with 0.5 mM isopropyl-1-thio-β-D-galactopyranoside and grown at 20°C for 24 h. The plasmid used for cloning was a modified pCWori+ (Wakarchuk et al. 1994
). N-Terminal fusions to the maltose binding protein of E. coli were made using the plasmid pCWmalE-thrombin (this plasmid has a thrombin cleavage site between the MalE and the cloned polysialyltransferase). In this plasmid, the malE gene from E. coli (strain BMH-71-18) with an added thrombin recognition sequence at the 3' end was inserted at the NdeI site of pCW and the 5' NdeI site was changed to CAGATG, leaving the 3' site intact for subsequent cloning.
Basic recombinant DNA method
All DNA isolations, restriction enzyme digestions, ligations, and transformations were performed as recommended by the supplier. Enzymes were obtained from New England Biolabs (Mississauga, ON, Canada). Genomic DNA was isolated using a DNeasy Tissue kit (Qiagen Inc., Mississauga, ON, Canada). PCR was performed using Phusion polymerase and the program: 94°C for 5 min, 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 60 s, and finally 72°C for 10 min. Primers for the E. coli PST were as follows: 5'-AAGGTATAAGACATATGA- TATTTGATGCTAGTTTAAAGAAG and 3'-CCTAGGTCGA- CTTACTCCCCCAAGAAAATCCTTTTATCGTGC. The N. meningitidis PST was amplified in two stages to remove an internal NdeI site (T474C). Two separate PCR reactions were performed to generate two overlapping gene fragments that both contained the silent mutation due to either the 5' or the 3' primers. The two PCR products were then used with primers 5'-GCTGGAGCTGGACATATGCTAAAGAAAATAAAAAA- AGCTCTTTTTCA and 3'-GCTGGAGCTGGAGTCGACCTA- TTATCTATCTCTACCAATTCTATTGTC to amplify the full-length gene containing the silent mutation. DNA was purified using either Qiaquick or Minelute kits from Qiagen Inc. (Mississauga, ON). Genes digested with NdeI and SalI were ligated into pCW or pCWmalE-thrombin and then used to transform E. coli AD202 by electroporation. Plasmids were isolated using High Pure Plasmid Isolation kit (Roche Diagnostics, Laval, PQ, Canada). DNA sequencing was performed using an Applied Biosystems (Montreal, PQ, Canada) model 3100 automated DNA sequencer and the manufacturer's cycle sequencing kit.
Construction of CST–PST fusions
The CST used was from C. jejuni OH4384 and consists of the first 260 amino acids with the mutation I53S, which is known to stabilize the CST-II proteins (Chiu et al. 2004). The gene was amplified with primers 5'-CTTAGGAGGTCATATGAAAAAA- GTTATTATTGCTGGAAATG which contains an NdeI site and 3'-GCTGGAGCTGGACATATGTCCGCCTCCAAAATTAAT- ATTTTTTGAAAATTTTCC which corresponds to the 3' end of CST without a stop codon, followed by three glycine codons to act as a linker and then an NdeI site. The pCWPST plasmids were digested with NdeI and then ligated with NdeI digested CST before transforming E. coli AD202 by electroporation. The orientation of the CST gene was determined with HindIII digestion and confirmed by sequencing.
Enzyme preparation
Cells grown in the presence of IPTG were lysed with an Avestin C5 Emulsiflex cell disrupter (Avestin Ottawa, ON, Canada) and centrifuged at 1000 x g to pellet unbroken cells. MalE-PST fusions were purified using amylose resin as described by the manufacturer (New England Biolabs, Mississauga, ON, Canada). Partial purification of CST–PST fusion protein was achieved by removing other proteins by binding them to a HiPrepQ FF column (GE Healthcare, Montreal, PQ, Canada), followed by the precipitation of the flow through fraction with ammonium sulfate at a concentration of 2 M, or by ultracentrifugation at 100,000 x g for 30 min. This material was then resuspended in a small volume of buffer and analyzed by SDS–PAGE. Densitometry of the stained gels was used to estimate the quantity of enzyme in the preparation. Protein concentrations were determined using a BCA assay (Pierce Biotechnology Inc., Rockford, IL).
Measurement of sialyltransferase activity
Basic assays were performed at 37°C in 10–40 µL volumes containing 50 mM NaHEPES pH 7.5, 10 mM MgCl2, 0.1–50 mM CMP-NeuAc, 0.01–5 mM labeled acceptor and various amounts of enzyme. Buffer optimization assays were done with the following buffers: 20 mM citrate-phosphate pH 6.0–8.0, 50 mM Na2PO4 pH 6.5–7.5, 50 mM NaHEPES pH 6.5–8.0, 50 mM Tris–HCl pH 7.5–9.0. Metal optimization assays were performed with 10 mM CaCl2, CoCl2, MgCl2, MnCl2, NiCl2, or ZnSO4. For the magnesium titration, endogenous Mg2+ was removed by addition of EDTA and then dialyzed against 20 mM Tris–HCl pH 7.5. Reactions were stopped in an equal volume of 50% acetonitrile, 1% sodium dodecylsulfate, 10 mM EDTA. Analysis of reactions was done by capillary electrophoresis as previously described using a TBE buffer at pH 8.8, where TBE was (Tris–HCl 90 mM, Boric acid, 90 mM, EDTA 2 mM) (Wakarchuk and Cunningham 2003).
Comparsion of CST-II/PSTNm mixtures with CST–PSTNm
The stoichiometry of the mixtures was calculated using gel densitometry. Care was taken not to overload the gels so that the protein bands were not saturated. Gels were stained with SYPRO Orange and protein was quantitated with software provided with the imager system (Alpha Imager, Alpha Innotech Corporation, CA). Reactions contained equimolar concentrations of enzyme (2.5–4 µM) based on the percentage present in the fusion protein preparations.
Acceptor synthesis
The GM3/GD3/GT3-FCHASE acceptors (Wakarchuk and Cunningham 2003) were synthesized from Lac-FCHASE on a milligram scale using purified CST-II essentially as described by Blixt et al. (2005). Purification was performed by reversed phase HPLC on a Hamiton PRP-1 10 x 300 mm reversed phase column using a gradient made from 10 mM ammonium acetate pH 5.5 and 80% acetonitrile. Fractions were recycled to improve purity.
Glycopeptide acceptors, NH2-VGV-[GalNAc--]-TETP-COOH (Tn-Interferon 2b peptide), and NH2- APALQP-[GalNAc--]-TQGAMPA-COOH (Tn-Granulocyte colony stimulating factor) were either purchased from Sussex Research Laboratories (Ottawa, Canada) or synthesized at NRC-IBS and then N-terminally labeled with FCHASE as described for the aminophenyl glycosides. GalNAc- was added with ppGalNAcT-2 (a gift of NEOSE technologies) and β-1,3-linked galactose was added to produce the Gal-β-1,3-GalNAc--R structure with the CgtB enzyme from C. jejuni (Bernatchez et al. 2007). Sialic acid was added to the O-linked structure Gal-β-1,3-GalNAc--Peptide using CST-I and CST-II to give mono-, di-, and trisialylated acceptors. All peptide substrates were purified by RPC on the column described above.
Mass spectrometry
After synthesis and purification the FCHASE labeled compounds were analyzed by mass spectrometry. A Prince CE system (Prince Technologies, The Netherlands) was coupled to a 4000 QTRAP mass spectrometer (Applied Biosystems/MDS Sciex, Canada). A sheath solution (isopropanol–methanol, 2:1) was delivered at a flow rate of 1.0 uL/min. Separations were obtained on about 90 cm length bare fused-silica capillary using 15 mM ammonium acetate in deionized water, pH 9.0. The 5 kV of electrospray ionization voltage was used for positive ion mode detection.
|
Funding |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Canadian Institutes for Health Research (MOP-64180); NEOSE Technologies Horsham PA, USA.
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
None declared.
|
Acknowledgements |
|---|
We thank Denis Brochu, Dr. Stéphane Bernatchez, and Cynthia Bainbridge for technical help, Dr. Jianjun Li and Jacek Stupak for mass spectrometry analysis and Tom Devecseri for assistance with the graphics.
|
Abbreviations |
|---|
CE, capillary electrophoresis; FCHASE, 6-(fluorescein-5-carbaxamido)hexanoic acid, succinimidyl ester; G-CSF Thr-134, granulocyte colony stimulating factor; HEPES, N-(2-hydroxyethyl)piperazine-N'-2-ethanesulfonic acid; IPTG, isopropyl-1-thio-β-D-galactopyranoside; LOS, lipooligosaccharide; NCAM, neural cell adhesion molecule; PSA, polysialic acid
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterial and methodsFundingConflict of interest statementReferences |
|---|
Aalto J, Pelkonen S, Kalimo H, Finne J. Mutant bacteriophage with non-catalytic endosialidase binds to both bacterial and eukaryotic polysialic acid and can be used as probe for its detection. Glycoconj J (2001) 18:751–758.[CrossRef][Web of Science][Medline]
Adlam C, Knight JM, Mugridge A, Williams JM, Lindon JC. Production of colominic acid by Pasteurella haemolytica serotype A2 organisms. FEBS Microbiol Lett (1987) 42:23–25.[CrossRef][Web of Science]
Andreishcheva EN, Vann WF. Gene products required for De Novo synthesis of polysialic acid in Escherichia coli K1. J Bacteriol (2006) 188:1786–1797.
Angata K, Chan D, Thibault J, Fukuda M. Molecular dissection of the ST8Sia IV polysialyltransferase: Distinct domains are required for neural cell adhesion molecule recognition and polysialylation. J Biol Chem (2004) 279:25883–25890.
Angata K, Fukuda M. Polysialyltransferases: Major players in polysialic acid synthesis on the neural cell adhesion molecule. Biochimie (2003) 85:195–206.[Medline]
Bernatchez S, Gilbert M, Blanchard MC, Karwaski MF, Li J, Defrees S, Wakarchuk WW. Variants of the β-1,3-galactosyltransferase CgtB from the bacterium Campylobacter jejuni have distinct acceptor specificities. Glycobiology (2007) 17:1333–1343.
Bhattacharjee AK, Jennings HJ, Kenny CP, Martin A, Smith IC. Structural determination of the sialic acid polysaccharide antigens of Neisseria meningitidis serogroups B and C with carbon 13 nuclear magnetic resonance. J Biol Chem (1975) 250:1926–1932.
Blixt O, Brown J, Schur MJ, Wakarchuk W, Paulson JC. Efficient preparation of natural and synthetic galactosides with a recombinant beta-1,4-galactosyltransferase-/UDP-4'-gal epimerase fusion protein. J Org Chem (2001) 66:2442–2448.[CrossRef][Web of Science][Medline]
Blixt O, Vasiliu D, Allin K, Jacobsen N, Warnock D, Razi N, Paulson JC, Bernatchez S, Gilbert M, Wakarchuk W. Chemoenzymatic synthesis of 2-azidoethyl-ganglio-oligosaccharides GD3, GT3, GM2, GD2, GT2, GM1, and GD1a. Carbohydr Res (2005) 340:1963–1972.[CrossRef][Web of Science][Medline]
Chiu CP, Watts AG, Lairson LL, Gilbert M, Lim D, Wakarchuk WW, Withers SG, Strynadka NC. Structural analysis of the sialyltransferase CstII from Campylobacter jejuni in complex with a substrate analog. Nat Struct Mol Biol (2004) 11:163–170.[CrossRef][Web of Science][Medline]
Cho J, Troy FA II. Polysialic acid engineering: Synthesis of polysialylated neoglycosphingolipids by using the polysialyltransferase from neuroinvasive Escherichia coli K1. PNAS (1994) 91:11427–11431.
Coutinho PM, Deleury E, Davies GJ, Henrissat B. An evolving hierarchical family classification for glycosyltransferases. J Mol Biol (2003) 328:307–317.[CrossRef][Web of Science][Medline]
Datta AK, Chammas R, Paulson JC. Conserved cysteines in the sialyltransferase sialylmotifs form an essential disulfide bond. J Biol Chem (2001) 276:15200–15207.
DeFrees S, Wang ZG, Xing R, Scott AE, Wang J, Zopf D, Gouty DL, Sjoberg ER, Panneerselvam K, Brinkman-Van der Linden E, et al. GlycoPEGylation of recombinant therapeutic proteins produced in Escherichia coli. Glycobiology (2006) 16:833–843.
El Maarouf A, Petridis AK, Rutishauser U. Use of polysialic acid in repair of the central nervous system. PNAS (2006) 103:16989–16994.
Franz CK, Rutishauser U, Rafuse VF. Polysialylated neural cell adhesion molecule is necessary for selective targeting of regenerating motor neurons. J Neurosci (2005) 25:2081–2091.
Freiberger F, Claus H, Gunzel A, Oltmann-Norden I, Vionnet J, Muhlenhoff M, Vogel U, Vann WF, Gerardy-Schahn R, Stummeyer K. Biochemical characterization of a Neisseria meningitidis polysialyltransferase reveals novel functional motifs in bacterial sialyltransferases. Mol Microbiol (2007) 65:1258–1275.[CrossRef][Web of Science][Medline]
Gilbert M, Bayer R, Cunningham AM, DeFrees S, Gao Y, Watson DC, Young NM, Wakarchuk WW. The synthesis of sialylated oligosaccharides using a CMP-Neu5Ac synthetase/sialyltransferase fusion. Nat Biotechnol (1998) 16:769–772.[CrossRef][Web of Science][Medline]
Gilbert M, Brisson JR, Karwaski MF, Michniewicz J, Cunningham AM, Wu Y, Young NM, Wakarchuk WW. Biosynthesis of ganglioside mimics in Campylobacter jejuni OH4384. Identification of the glycosyltransferase genes, enzymatic synthesis of model compounds, and characterization of nanomole amounts by 600-mhz (1)h and (13)c NMR analysis. J Biol Chem (2000) 275:3896–3906.
Gilbert M, Karwaski MF, Bernatchez S, Young NM, Taboada E, Michniewicz J, Cunningham AM, Wakarchuk WW. The genetic bases for the variation in the lipo-oligosaccharide of the mucosal pathogen, Campylobacter jejuni. Biosynthesis of sialylated ganglioside mimics in the core oligosaccharide. J Biol Chem (2002) 277:327–337.
Gilbert M, Watson DC, Cunningham AM, Jennings MP, Young NM, Wakarchuk WW. Cloning of the lipooligosaccharide -2,3-sialyltransferase from the bacterial pathogens Neisseria meningitidis and Neisseria gonorrhoeae. J Biol Chem (1996) 271:28271–28276.
Girard MP, Preziosi MP, Aguado MT, Kieny MP. A review of vaccine research and development: Meningococcal disease. Vaccine (2006) 24:4692–4700.[CrossRef][Web of Science][Medline]
Gregoriadis G, Jain S, Papaioannou I, Laing P. Improving the therapeutic efficacy of peptides and proteins: A role for polysialic acids. Int J Pharm (2005) 300:125–130.[CrossRef][Web of Science][Medline]
Jennings HJ, Bhattacharjee AK, Bundle DR, Kenny CP, Martin A, Smith IC. Structures of the capsular polysaccharides of Neisseria meningitidis as determined by 13C-nuclear magnetic resonance spectroscopy. J Infect Dis (1977) 136(Suppl):S78–S83.[Web of Science][Medline]
Kahler CM, Stephens DS. Genetic basis for biosynthesis, structure, and function of meningococcal lipooligosaccharide (endotoxin). Crit Rev Microbiol (1998) 24:281–334.[Web of Science][Medline]
Kitazume-Kawaguchi S, Kabata S, Arita M. Differential biosynthesis of polysialic or disialic acid structure by ST8Sia II and ST8Sia IV. J Biol Chem (2001) 276:15696–15703.
Kogan G, Uhrøn D, Brisson JR, Paoletti LC, Blodgett A.E, Kasper DL, Jennings HJ. Structural and immunochemical characterization of the type VIII group B Streptococcus capsular polysaccharide. J Biol Chem (1996) 271:8786–8790.
Lifely MR, Lindon JC, Williams JM, Moreno C. Structural and conformational features of the Escherichia coli K92 capsular polysaccharide. Carbohydr Res (1985) 143:191–205.[CrossRef][Web of Science][Medline]
McGuire EJ, Binkley SB. The structure and chemistry of colominic acid. Biochemistry (1964) 3:247–251.[CrossRef][Web of Science][Medline]
Moran AP, Prendergast MM, Appelmelk BJ. Molecular mimicry of host structures by bacterial lipopolysaccharides and its contribution to disease. FEMS Immunol Med Microbiol. (1996) 16:105–115.[CrossRef][Web of Science][Medline]
Pon RA, Lussier M, Yang QL, Jennings HJ. N-Propionylated group B meningococcal polysaccharide mimics a unique bactericidal capsular epitope in group B Neisseria meningitidis. J Exp Med (1997) 185:1929–1938.
Shen GJ, Datta AK, Izumi M, Koeller KM, Wong CH. Expression of 2,8/2,9-Polysialyltransferase from Escherichia coli K92. Characterization of the enzyme and its reaction products. J Biol Chem (1999) 274:35139–35146.
Silver RP, Vimr ER. Polysialic acid capsule of Escherichia coli K1. Iglewski BHaCVL, ed. (1990) Vol. XI. San Diego (CA): Academic Press. p. 39–60.
Uemura S, Kurose T, Suzuki T, Yoshida S, Ito M, Saito M, Horiuchi M, Inagaki F, Igarashi Y, Inokuchi Ji. Substitution of the N-glycan function in glycosyltransferases by specific amino acids (SUNGA): ST3Gal-V as a model enzyme. Glycobiology (2005) 6:258–270.
Vionnet J, Vann WF. Successive glycosyltransfer of sialic acid by Escherichia coli K92 polysialyltransferase in elongation of oligosialic acceptors. Glycobiology (2007) 7:735–743.
Vogel U, Hammerschmidt S, Frosch M. Sialic acids of both the capsule and the sialylated lipooligosaccharide of Neisseria meningitidis serogroup B are prerequisites for virulence of meningococci in the infant rat. Med Microbiol Immunol (Berl) (1996) 185:81–87.[CrossRef][Medline]
Wakarchuk WW, Campbell RL, Sung WL, Davoodi J, Yaguchi M. Mutational and crystallographic analyses of the active site residues of the Bacillus circulans xylanase. Protein Sci (1994) 3:467–475.[Web of Science][Medline]
Wakarchuk WW, Cunningham AM. Capillary electrophoresis as an assay method for monitoring glycosyltransferase activity. Methods Mol Biol (2003) 213:263–274.[Medline]
Wakarchuk WW, Watson D, St Michael F, Li J, Wu Y, Brisson JR, Young NM, Gilbert M. Dependence of the bi-functional nature of a sialyltransferase from Neisseria meningitidis on a single amino acid substitution. J Biol Chem (2001) 276:12785–12790.
Yu H, Chokhawala H, Karpel R, Yu H, Wu B, Zhang J, Zhang Y, Jia Q, Chen X. A multifunctional Pasteurella multocida sialyltransferase: A powerful tool for the synthesis of sialoside libraries. J Am Chem Soc (2005) 127:17618–17619.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. J. Morley, L. M. Willis, C. Whitfield, W. W. Wakarchuk, and S. G. Withers A New Sialidase Mechanism: BACTERIOPHAGE K1F ENDO-SIALIDASE IS AN INVERTING GLYCOSIDASE J. Biol. Chem., June 26, 2009; 284(26): 17404 - 17410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R S. Houliston, S. Bernatchez, M.-F. Karwaski, R. E Mandrell, H. C Jarrell, W. W Wakarchuk, and M. Gilbert Complete chemoenzymatic synthesis of the Forssman antigen using novel glycosyltransferases identified in Campylobacter jejuni and Pasteurella multocida Glycobiology, February 1, 2009; 19(2): 153 - 159. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||