Structural and Biochemical Investigation of PglF from Campylobacter jejuni Reveals a New Mechanism for a Member of the Short Chain Dehydrogenase/Reductase Superfamily.

Within recent years it has become apparent that protein glycosylation is not limited to eukaryotes. Indeed, in Campylobacter jejuni, a Gram-negative bacterium, more than 60 of its proteins are known to be glycosylated. One of the sugars found in such glycosylated proteins is 2,4-diacetamido-2,4,6-trideoxy-α-d-glucopyranose, hereafter referred to as QuiNAc4NAc. The pathway for its biosynthesis, initiating with UDP-GlcNAc, requires three enzymes referred to as PglF, PglE, and PlgD. The focus of this investigation is on PglF, an NAD+-dependent sugar 4,6-dehydratase known to belong to the short chain dehydrogenase/reductase (SDR) superfamily. Specifically, PglF catalyzes the first step in the pathway, namely, the dehydration of UDP-GlcNAc to UDP-2-acetamido-2,6-dideoxy-α-d-xylo-hexos-4-ulose. Most members of the SDR superfamily contain a characteristic signature sequence of YXXXK where the conserved tyrosine functions as a catalytic acid or a base. Strikingly, in PglF, this residue is a methionine. Here we describe a detailed structural and functional investigation of PglF from C. jejuni. For this investigation five X-ray structures were determined to resolutions of 2.0 Å or better. In addition, kinetic analyses of the wild-type and site-directed variants were performed. On the basis of the data reported herein, a new catalytic mechanism for a SDR superfamily member is proposed that does not require the typically conserved tyrosine residue.

Biosynthesis of Legionaminic Acid and Its Incorporation Into Glycoconjugates.

Legionaminic acids are analogs of sialic acid that occur in cell surface glycoconjugates of several bacteria. Because legionaminic acids share the same stereochemistry as sialic acid but differ at C7 and C9, they are interesting analogs to probe the impact of varying exocyclic moieties (C7-C9) on biological activities such as susceptibilities to sialidases, interactions with Siglecs and immunogenicity. There are currently no reports on the bacterial enzymes that transfer legionaminic acids to these cell surface glycoconjugates, but some mammalian and bacterial sialyltransferases display donor promiscuity and can use CMP-Leg5,7Ac2 efficiently enough to transfer Leg5,7Ac2 to their natural acceptor glycans. When the natural activity with CMP-Leg5,7Ac2 is significant but relatively low, an alternate strategy has been to engineer versions with improved activity to transfer Leg5,7Ac2. Importantly, we have found that some bacterial sialyltransferases are very efficient for transferring Leg5,7Ac2 to small synthetic glycans with various aglycones. The two mammalian sialyltransferases that have been tested so far (porcine ST3Gal1 and human ST6Gal1) were found to be more efficient than the bacterial sialyltransferases for the modification of glycoproteins. We provide a review of the sialyltransferases selected to modify different types of glycoconjugates with Leg5,7Ac2, including small synthetic acceptors, glycolipids, and glycoproteins. In the first part, we also propose an optimized biosynthetic pathway for in vitro preparation of the donor CMP-Leg5,7Ac2, based on enzymes selected from two bacteria that naturally produce legionaminic acid.

Molecular Basis for Recognition of the Cancer Glycobiomarker, LacdiNAc (GalNAc[ß1→4]GlcNAc), by Wisteria floribunda Agglutinin.

Aberrant glycosylation and the overexpression of specific carbohydrate epitopes is a hallmark of many cancers, and tumor-associated oligosaccharides are actively investigated as targets for immunotherapy and diagnostics. Wisteria floribunda agglutinin (WFA) is a legume lectin that recognizes terminal N-acetylgalactosaminides with high affinity. WFA preferentially binds the disaccharide LacdiNAc (ß-d-GalNAc-[1→4]-d-GlcNAc), which is associated with tumor malignancy in leukemia, prostate, pancreatic, ovarian, and liver cancers and has shown promise in cancer glycobiomarker detection. The mechanism of specificity for WFA recognition of LacdiNAc is not fully understood. To address this problem, we have determined affinities and structure of WFA in complex with GalNAc and LacdiNAc. Affinities toward Gal, GalNAc, and LacdiNAc were measured via surface plasmon resonance, yielding KD values of 4.67 × 10-4 m, 9.24 × 10-5 m, and 5.45 × 10-6 m, respectively. Structures of WFA in complex with LacdiNAc and GalNAc have been determined to 1.80-2.32 Å resolution. These high resolution structures revealed a hydrophobic groove complementary to the GalNAc and, to a minor extent, to the back-face of the GlcNAc sugar ring. Remarkably, the contribution of this small hydrophobic surface significantly increases the observed affinity for LacdiNAc over GalNAc. Tandem MS sequencing confirmed the presence of two isolectin forms in commercially available WFA differing only in the identities of two amino acids. Finally, the WFA carbohydrate binding site is similar to a homologous lectin isolated from Vatairea macrocarpa in complex with GalNAc, which, unlike WFA, binds not only αGalNAc but also terminal Ser/Thr O-linked αGalNAc (Tn antigen).

Preparation of legionaminic acid analogs of sialo-glycoconjugates by means of mammalian sialyltransferases.

Legionaminic acids are analogs of sialic acid that occur in several bacteria. The most commonly occurring form is Leg5Ac7Ac, which differs from Neu5Ac only at the C7 (acetamido) and C9 (deoxy) positions. While these differences greatly reduce the susceptibility of Leg compounds to sialidases, several sialyltransferases have been identified that can use CMP-Leg5Ac7Ac as a donor (Watson et al. 2011). We report the successful modification with Leg5Ac7Ac of a glycolipid, GM1a, and two glycoproteins, interferon-α2b and α1-antitrypsin, by means of two mammalian sialyltransferases, namely porcine ST3Gal1 and human ST6Gal1. The Leg5Ac7Ac form of GD1a was not recognized by the myelin-associated glycoprotein (MAG, Siglec-4), confirming the importance of the glycerol moiety in the interaction of sialo-glycans with Siglecs.

Structure of the external aldimine form of PglE, an aminotransferase required for N,N'-diacetylbacillosamine biosynthesis.

N,N'-diacetylbacillosamine is a novel sugar that plays a key role in bacterial glycosylation. Three enzymes are required for its biosynthesis in Campylobacter jejuni starting from UDP-GlcNAc. The focus of this investigation, PglE, catalyzes the second step in the pathway. It is a PLP-dependent aminotransferase that converts UDP-2-acetamido-4-keto-2,4,6-trideoxy-d-glucose to UDP-2-acetamido-4-amino-2,4,6-trideoxy-d-glucose. For this investigation, the structure of PglE in complex with an external aldimine was determined to a nominal resolution of 2.0 Å. A comparison of its structure with those of other sugar aminotransferases reveals a remarkable difference in the manner by which PglE accommodates its nucleotide-linked sugar substrate.

Draft Whole-Genome Sequence of Morganella morganii Serotype O:1ab.

Morganella morganii is a facultative pathogen of humans, causing urinary tract and postsurgical infections. Here, we report a high-quality draft assembly of the O:1ab serotype.

Sialyltransferases with enhanced legionaminic acid transferase activity for the preparation of analogs of sialoglycoconjugates.

Legionaminic acids (Leg) are bacterial analogs of neuraminic acid, with the same stereochemistry but different substituents at C5, C7 and C9. Hence they may be incorporated into useful analogs of sialoglycoconjugates, and we previously reported two sialyltransferases that could utilize cytidine monophosphate (CMP)-Leg5Ac7Ac for preparation of Leg glycoconjugates, which were resistant to sialidases [Watson DC, Leclerc S, Wakarchuk WW, Young NM. 2011. Enzymatic synthesis and properties of glycoconjugates with legionaminic acid as a replacement for neuraminic acid. Glycobiology. 21:99-108.]. These were the porcine ST3Gal1 and Pasteurella multocida sialyltransferases. We now report two additional sialyltransferases with superior Leg-transferase properties to the previous two. These are (i) a truncated form of a Photobacterium α2,6-sialyltransferase with an Ala-Met mutation in its active site, and (ii) an α2,3-sialyltransferase from Neisseria meningitidis MC58 with a higher transferase activity than the P. multocida enzyme, with either CMP-Neu5Ac or CMP-Leg5Ac7Ac as the donor. These enzymes will enable the production of useful Leg5Ac7Ac glycoconjugate derivatives with either α2,6 or α2,3 linkages and unique biological properties.

The structure of the Morganella morganii lipopolysaccharide core region and identification of its genomic loci.

The core region of the lipopolysaccharide of Morganella morganii serotype O:1ab was obtained by hydrolysis of the LPS and studied by 2D NMR, ESI MS, and chemical methods. Its structure was highly homologous to those from the two major members of the same Proteeae tribe, Proteus mirabilis and Providencia alcalifaciens, and analysis of the M. morganii genome disclosed that the loci for its outer core, lipid A and Ara4N moieties are similarly conserved.

Predicting the Origins of Anti-Blood Group Antibody Specificity: A Case Study of the ABO A- and B-Antigens.

The ABO blood group system is the most important blood type system in human transfusion medicine. Here, we explore the specificity of antibody recognition toward ABO blood group antigens using computational modeling and biolayer interferometry. Automated docking and molecular dynamics simulations were used to explore the origin of the specificity of an anti-blood group A antibody variable fragment (Fv AC1001). The analysis predicts a number of Fv-antigen interactions that contribute to affinity, including a hydrogen bond between a His(L49) and the carbonyl moiety of the GalNAc in antigen A. This interaction was consistent with the dependence of affinity on pH, as measured experimentally; at lower pH there is an increase in binding affinity. Binding energy calculations provide unique insight into the origin of interaction energies at a per-residue level in both the scFv and the trisaccharide antigen. The calculations indicate that while the antibody can accommodate both blood group A and B antigens in its combining site, the A antigen is preferred by 4 kcal/mol, consistent with the lack of binding observed for the B antigen.

The intrinsic cysteine and histidine residues of the anti-Salmonella antibody Se155-4: a model for the introduction of new functions into antibody-binding sites.

New functions can be incorporated into anti-hapten or anti-protein antibodies by mutating selected residues in the binding-site region either to Cys, to allow alkylation with reagents bearing the desired functional groups, or to His, to create metal-binding sites or to make antigen binding pH-sensitive. However, choosing suitable sites for these mutations has been hampered by the lack of antibodies with these features, to serve as models. Remarkably, the anti-carbohydrate antibody Se155-4, specific for the Salmonella group B lipopolysaccharide, already has a Cys and two pairs of His residues close to the antigen-binding pocket in its structure, and shows pH-dependent antigen binding. We therefore investigated modification of its Cys94L in an scFv version of the antibody with the aims of creating a 'reagentless' fluorescent sensor and attaching a metal-binding group that might confer lyase activity. These groups were successfully introduced, as judged by mass spectrometry, and had only slightly reduced antigen binding in enzyme-linked immunosorbent assay. The fluorescent product was sensitive to addition of antigen in a solution format, unlike a modification of a more distant Cys introduced into the VH CDR4 loop. Two other routes to modulate antigen binding were also explored, metal binding by the His pair alongside the antigen-binding pocket and insertions into CDR4 to extend the antigen-contact area. His residues adjacent to the antigen-binding pocket bound copper, causing a 5-fold decrease in antigen binding. In CDR4 of the VH domain, the preferred insert length was four residues, which gave stable antigen-binding products but did not improve overall antigen affinity.

Review of phosphocholine substituents on bacterial pathogen glycans: synthesis, structures and interactions with host proteins.

Among the non-carbohydrate components of glycans, the addition of phosphocholine (ChoP) to the glycans of pathogens occurs more rarely than acetylation or methylation, but it has far more potent biological consequences. These arise from ChoP's multiple interactions with host proteins, which are important at all stages of the infection process. These stages include initial adherence to cells, encountering the host's innate immune system and then the adaptive immune system. Thus, in the initial stages of an infection, ChoP groups are an asset to the pathogen, but they can turn into a disadvantage subsequently. In this review, we have focussed on structural aspects of these phenomena. We describe the biosynthesis of the ChoP modification, the structures of the pathogen glycans known to carry ChoP groups and the host proteins that recognize ChoP.

Specificity analysis of the C-type lectin from rattlesnake venom, and its selectivity towards Gal- or GalNAc-terminated glycoproteins.

The rattlesnake (Crotalus atrox) venom lectin is a readily-prepared decameric C-type lectin, specific for Gal and GalNAc. Glycan microarray analysis showed it reacted with a wide range of glycans, chiefly recognizing sets of compounds with Galß1-4GlcNAc (LacNAc), α-Gal or α-GalNAc non-reducing termini. Its array profile was therefore distinctly different from those of four previously studied mammalian C-type lectins with the same Gal/GalNAc monosaccharide specificity, and it was more broadly reactive than several Gal- or GalNAc-specific plant lectins commonly used for glycan blotting. Though a general reactivity towards glycoproteins might be expected from the avidity conferred by its high valence, it showed a marked preference for glycoproteins with multiple glycans, terminated by Gal or GalNAc. Thus its ten closely-spaced sites each with a K(D) for GalNAc of ~2 mM appeared to make RSVL more selective than the four more widely-spaced sites of soybean agglutinin, with a ten-fold better K(D) for GalNAc.

Structural characterization and MHCII-dependent immunological properties of the zwitterionic O-chain antigen of Morganella morganii.

Morganella morganii is a commensal Gram-negative bacterium that has long been known to produce an antigen bearing phosphocholine groups. We determined the structure of this O-chain antigen and found that its repeating unit also contains a free amino group and a second phosphate: This alternating charge character places the M. morganii O-chain polysaccharide into a small family of zwitterionic polysaccharides (ZPSs) known to induce T-cell-dependent immune responses via presentation by class II major histocompatibility complex (MHCII) molecules. In vitro binding assays demonstrate that this O-chain interacts with MHCII in a manner that competes with binding of the prototypical ZPS antigen PSA from Bacteroides fragilis, despite its lack of a helical structure. Cellular studies also showed that the M. morganii polysaccharide induces activation of CD4(+) T-cells. Antibody binding experiments using acid hydrolyzed fragments representing the monomer and higher oligomers of the repeating unit showed that the phosphocholine group was the dominant element of the epitope with an overall affinity (K(D)) of about 5 × 10(-5) M, a typical value for an IgM anti-carbohydrate antibody but much lower than the affinity for phosphocholine itself. These data show that the structure of the M. morganii polysaccharide contains a unique zwitterionic repeating unit which allows for immune recognition by T-cells, making it the first identified T-cell-dependent O-chain antigen.

Enzymatic synthesis and properties of glycoconjugates with legionaminic acid as a replacement for neuraminic acid.

In addition to sialic acid, bacteria produce several other nonulosonic acids, including legionaminic acid (Leg). This has exactly the same stereochemistry as sialic acid, with the added features of 9-deoxy and 7-amino groups. In order to explore the biological effects of replacing sialic acid residues (Neu5Ac) in glycoconjugates with Leg in its diacetylated form, diacetyllegionaminic acid (Leg5Ac7Ac), we tested CMP-Leg5Ac7Ac as a donor substrate with a selection of bacterial and mammalian sialyltransferases. The CMP-Leg5Ac7Ac was synthesized in vitro by means of cloned enzymes from the bacillosamine portion of the Campylobacter jejuni N-glycan pathway and from the Leg pathway of Legionella pneumophila. Using fluorescent derivatives of lactose, Galß1,4GlcNAcß and T-antigen (Galß1,3GalNAcα) as acceptors, we tested eight different sialyltransferases and found that the Pasteurella multocida PM0188h and porcine ST3Gal1 sialyltransferases were significantly active with CMP-Leg5Ac7Ac, showing ∼60% activity when compared with CMP-Neu5Ac. The Photobacterium α2,6 sialyltransferase was weakly active, with ∼6% relative activity. The Leg5Ac7Ac-α-2,3-lactose product was then tested as a substrate with six sialidases of viral, bacterial and mammalian origin. All showed much lower activities than with the corresponding sialic acid substrate, with the influenza virus N1 being the most active and human NEU2 being the least active. These results show the feasibility of producing glycoconjugates with Leg5Ac7Ac residues as the terminal sugars, which should display novel biological properties.

Characterization of the secondary binding sites of Maclura pomifera agglutinin by glycan array and crystallographic analyses.

The Maclura pomifera agglutinin (MPA) recognizes the T-antigen disaccharide Galß1,3GalNAc mainly through interaction of the α-GalNAc moiety with its primary site, but the interactions of the two flanking subsites A and B with aglycones and substituents other than Gal, respectively, are not well understood. We therefore characterized the specificity of MPA in more detail by glycan microarray analysis and determined the crystal structures of MPA without ligand and in complexes with Galß1,3GalNAc and p-nitrophenyl α-GalNAc. In both sugar complexes, pairs of ligands created inter-tetramer hydrogen-bond bridging networks. While subsite A showed increased affinity for hydrophobic aglycones, it also accommodated several sugar substituents. Notably, a GalNAc-O-tripeptide, a Tn-antigen mimic, showed lower affinity than these compounds in surface plasmon resonance (SPR) experiments. The glycan array data that showed subsite B accepted compounds in which the O3 position of the GalNAc was substituted with various sugars other than Gal, but substitutions at O6 led to inactivity. Additions to the Gal moiety of the disaccharide also had only small effects on reactivity. These results are all compatible with the features seen in the crystal structures.

Specificity of Campylobacter jejuni adhesin PEB3 for phosphates and structural differences among its ligand complexes.

PEB3 is a glycoprotein adhesin from Campylobacter jejuni whose structure suggested a role in transport. We have investigated potential ligands for PEB3 and characterized their binding properties using biophysical methods in solution and by X-ray crystallography. A thermal aggregation assay of PEB3 with a library of physiological compounds identified three possible ligands [3-phosphoglycerate (3-PG), phosphoenolpyruvate (PEP), and aconitate], which stabilized wild-type PEB3 but did not stabilize either a PEB3 form containing two mutations at the ligand-binding site, T138A/S139A, or a second PEB3 mutant, K135E, at a site approximately 14 A away. Fluorescence titration experiments and cocrystal structures with various ligands were used to characterize the binding of 3-PG, PEP, and phosphate to PEB3. Further, a C. jejuni growth experiment in minimal medium supplemented with 3-PG showed that this molecule enhances the growth of wild-type C. jejuni, but not of the PEB3 mutants. Crystallographic analysis of PEB3 complexes revealed that the Ser171-Gln180 region in the presence of 3-PG or other phosphates is helical and similar to those of other transport proteins, but it is nonhelical when citrate is bound. The K135E mutation resulted in expression of a more highly glycosylated form of PEB3 in vivo, and its crystal structure showed the conformation of the first two residues of the glycan. On the basis of our findings, we suggest that PEB3 is a transport protein that may function in utilization of 3-PG or other phosphate-containing molecules from the host.

Biosynthesis of CMP-N,N'-diacetyllegionaminic acid from UDP-N,N'-diacetylbacillosamine in Legionella pneumophila.

Legionaminic acid is a nine-carbon alpha-keto acid that is similar in structure to other members of the sialic acid family that includes neuraminic acid and pseudaminic acid. It is found as a component of the lipopolysaccharide in several bacterial species and is perhaps best known for its presence in the O-antigen of the causative agent of Legionnaires' disease, Legionella pneumophila. In this work, the enzymes responsible for the biosynthesis and activation of N, N'-diacetyllegionaminic acid are identified for the first time. A cluster of three L. pneumophila genes bearing homology to known sialic acid biosynthetic genes ( neuA,B,C) were cloned and overexpressed in Escherichia coli. The NeuC homologue was found to be a hydrolyzing UDP- N, N'-diacetylbacillosamine 2-epimerase that converts UDP- N, N'-diacetylbacillosamine into 2,4-diacetamido-2,4,6-trideoxymannose and UDP. Stereochemical and isotopic labeling studies showed that the enzyme utilizes a mechanism involving an initial anti elimination of UDP to form a glycal intermediate and a subsequent syn addition of water to generate product. This is similar to the hydrolyzing UDP- N-acetylglucosamine 2-epimerase (NeuC) of sialic acid biosynthesis, but the L. pneumophila enzyme would not accept UDP-GlcNAc as an alternate substrate. The NeuB homologue was found to be a N, N'-diacetyllegionaminic acid synthase that condenses 2,4-diacetamido-2,4,6-trideoxymannose with phosphoenolpyruvate (PEP), although the in vitro activity of the recombinant enzyme (isolated as a MalE fusion protein) was very low. The synthase activity was dependent on the presence of a divalent metal ion, and the reaction proceeded via a C-O bond cleavage process, similar to the reactions catalyzed by the sialic acid and pseudaminic acid synthases. Finally, the NeuA homologue was shown to possess the CMP- N, N'-diacetyllegionaminic acid synthetase activity that generates the activated form of legionaminic acid used in lipopolysaccharide biosynthesis. Together, the three enzymes constitute a pathway that converts a UDP-linked bacillosamine derivative into a CMP-linked legionaminic acid derivative.

Structure and active site residues of PglD, an N-acetyltransferase from the bacillosamine synthetic pathway required for N-glycan synthesis in Campylobacter jejuni.

Campylobacter jejuni is highly unusual among bacteria in forming N-linked glycoproteins. The heptasaccharide produced by its pgl system is attached to protein Asn through its terminal 2,4-diacetamido-2,4,6-trideoxy-d-Glc (QuiNAc4NAc or N,N'-diacetylbacillosamine) moiety. The crucial, last part of this sugar's synthesis is the acetylation of UDP-2-acetamido-4-amino-2,4,6-trideoxy-d-Glc by the enzyme PglD, with acetyl-CoA as a cosubstrate. We have determined the crystal structures of PglD in CoA-bound and unbound forms, refined to 1.8 and 1.75 A resolution, respectively. PglD is a trimer of subunits each comprised of two domains, an N-terminal alpha/beta-domain and a C-terminal left-handed beta-helix. Few structural differences accompany CoA binding, except in the C-terminal region following the beta-helix (residues 189-195), which adopts an extended structure in the unbound form and folds to extend the beta-helix upon binding CoA. Computational molecular docking suggests a different mode of nucleotide-sugar binding with respect to the acetyl-CoA donor, with the molecules arranged in an "L-shape", compared with the "in-line" orientation in related enzymes. Modeling indicates that the oxyanion intermediate would be stabilized by the NH group of Gly143', with His125' the most likely residue to function as a general base, removing H+ from the amino group prior to nucleophilic attack at the carbonyl carbon of acetyl-CoA. Site-specific mutations of active site residues confirmed the importance of His125', Glu124', and Asn118. We conclude that Asn118 exerts its function by stabilizing the intricate hydrogen bonding network within the active site and that Glu124' may function to increase the pKa of the putative general base, His125'.

The complete genome sequence of Actinobacillus pleuropneumoniae L20 (serotype 5b).

There are 16 capsule-based serotypes of Actinobacillus pleuropneumoniae, all of which are capable of causing disease in pigs. Here we report the finished and annotated genome sequence of the reference serotype 5b strain L20. This strain has a rough appearance and readily forms biofilms, as is typical for most field isolates.

The cytoplasmic phosphoproteome of the Gram-negative bacterium Campylobacter jejuni: evidence for modification by unidentified protein kinases.

We have undertaken a comprehensive analysis of cytoplasmic protein phosphorylation in Campylobacter jejuni by mass spectrometric identification of phosphoproteins and localization of the sites of modification by phosphopeptide analyses. Cell extracts, enriched for phosphoproteins using Fe(III) IMAC or commercial phosphoprotein purification kits, were analyzed by 1-D and 2-D SDS-PAGE and subjected to mass fingerprinting by in-gel tryptic digestion and MALDI-TOF MS. Fifty-eight phosphopeptides were identified from 1-D gel bands by nano-LC-MS/MS and automated searching in a C. jejuni ORF database resulting in the unequivocal identification of 36 phosphoproteins of diverse function. In addition to elongation factors and chaperonins, which have been reported to be phosphorylated in other bacteria, the major phosphoproteins included bacterioferritin and superoxide dismutase. The sequences around the phosphorylated Ser and Thr residues are indicative of specific kinases being responsible for some of the modifications. However, many of the other identified proteins are enzymes that have phosphorylated substrates, including ATP, hence other modifications may arise from autophosphorylation. Comparative analyses of IMAC extracts from the Escherichia coli strain AD202 and Helicobacter pylori resulted in the identification of homologs of six of the C. jejuni phosphoproteins, though their overall phosphoproteome maps were distinctly different.