The Campylobacter jejuni glycome - Wiley Online Library

FEMS Microbiology Reviews 29 (2005) 377–390 www.fems-microbiology.org

The Campylobacter jejuni glycome

q

Andrey V. Karlyshev a, Julian M. Ketley b, Brendan W. Wren a

a,*

Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel St., London WC1E 7HT, UK b Department of Genetics, University of Leicester, Leicester LE1 7RH, UK Received 14 December 2004; received in revised form 21 January 2005; accepted 21 January 2005 First published online 1 February 2005

Abstract Microbial cell surface glycans in the form of glycolipids and glycoproteins frequently play important roles in cell–cell interaction and host immune responses. Given the likely importance of these surface structures in the survival and pathogenesis of Campylobacter jejuni, a concerted effort has been made to characterise these determinants genetically and structurally since the genome was sequenced in 2000. We review the considerable progress made in characterising the Campylobacter glycome including the lipooligosaccharide (LOS), the capsule and O- and N-linked protein glycosylation systems, and speculate on the roles played by glycan surface structures in the life-cycle of C. jejuni. Ó 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Campylobacter jejuni; Glycome; Lipooligosaccharide; Capsule; Glycosylated flagellin; N-linked glycosylation system

Contents 1. 2.

3.

4.

5.

q *

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lipooligosaccharide (LOS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Lipopolysaccharide biosynthesis by C. jejuni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. LOS structure in NCTC 11168 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Genetic basis of LOS biosynthesis in C. jejuni NCTC 11168 . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Variation of LOS structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Role of LOS in C. jejuni-mediated disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Capsular polysaccharide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Capsule polysaccharide structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Capsule genetic diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . O-linked protein glycosylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Glycan structure of modified flagellin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Genetic organisation and diversity of the O-linked flagella glycosylation system . . . . . . . . . . . . N-linked protein glycosylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Structure of N-linked glycan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Genetic basis of N-linked glycosylation system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

Edited by Charles W. Penn. Corresponding author. E-mail address: [email protected] (B.W. Wren).

0168-6445/$22.00 Ó 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsre.2005.01.003

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

378 378 379 379 379 381 383 384 384 384 385 385 386 386 386 386

378

6.

A.V. Karlyshev et al. / FEMS Microbiology Reviews 29 (2005) 377–390

5.3. Role of N-linked glycosylation system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction The Gram-negative bacterium Campylobacter jejuni is the principal bacterial cause of human gastroenteritis world-wide [1]. C. jejuni is also the most frequent antecedent infection responsible for the post-infectious peripheral neuropathy Guillain-Barre´ syndrome, which is thought to be triggered by mimicry with surface exposed carbohydrate structures. The main source of transmission is through the consumption of livestock, particularly poultry where invariably flocks are heavily contaminated with up to 109 Campylobacter, yet are asymptomatic. By contrast as few as 102 cells is sufficient to cause severe disease in humans. Given the medical and public health importance of Campylobacter infection, it is remarkable that C. jejuni is one of the least understood enteropathogens. The lack of understanding of how the organism causes disease was the driving force for the instigation of the C. jejuni NCTC 11168 genome sequencing project that was completed in 2000 [2]. The sequencing of NCTC 11168 revealed a previously unexpected capacity for the organism to produce a variety of carbohydrates [2]. Subsequent analyses have demonstrated the presence of several cell surface glycoconjugates such as a novel capsule and N-linked glycosylation pathway, as well as providing rich data on the lipooligosaccharide (LOS) and the O-linked glycosylation system that decorates the flagellum (Fig. 1). A striking feature from the genome project was the presence of several hypervariable homopolymeric repeats responsible for slipped-strand mispairing and phase variation in glycan moieties present in LOS, capsule and flagella [2]. Further studies have shown that the loci

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

387 387 388 388

Forming an integral part of the cell, lipid A substituted polysaccharide (PS) forms an essential cell surface structure as part of the outer leaflet of the outer membrane lipid bilayer. In addition to maintaining the structural integrity of the bacterium via a barrier function to substances entering or leaving the cell, modulation of the attached PS provides a strategy for avoidance of the host immune system. In most gram negative bacteria, lipid A is substituted with a core PS to which is attached a set of repeating subunits termed the O-chain; this structure is termed lipopolysaccharide (LPS). How-

Capsular polysaccharide VARIABLE

N-linked glycoproteins CONSERVED

Polysaccharide repeat unit

Inner core Phospholipid

outer

. . . .

2. Lipooligosaccharide (LOS)

Lipo-oligosaccharide VARIABLE Outer core

. . . .

responsible for the biosynthesis of these molecules are highly variable among different strains [3–7]. This is in contrast to the N-linked general glycosylation pathway that appears highly conserved, even among some other Campylobacter species [12]. Recent studies have allowed the detailed structural analysis of the glycan components of the glycolipids and glycoproteins [9], which has facilitated studies to determine gene function in the respective biosynthetic pathways [10,11]. This review will summarise the post genome genetic and structural studies on the C. jejuni glycolipids (LOS and capsule) and the O- and N-linked glycosylation pathways. Because there have been two recent reviews on the Campylobacter glycosylation systems [10,12], the emphasis on this current review will be on Campylobacter LOS and capsule. We will also speculate on the possible biological roles of these structural surface determinants and why some appear to be hypervariable.

n

O-linked glycosylated flagellum VARIABLE

. . . .

Lipid A

membrane periplasm Fig. 1. The known glycolipids and glycoproteins in C. jejuni.


ever, in other bacteria, for example, Neisseria and Haemophilus sp. a lipooligosaccharide is synthesised where only core oligosaccharide is attached and the repeating O-chain is absent.

2.1. Lipopolysaccharide biosynthesis by C. jejuni Historically, C. jejuni was believed to produce either LPS or LOS in a strain-dependent manner. Early reports describe the presence of a high molecular weight (HMW) PS in some strains that, in the absence of evidence to the contrary, was deemed to reflect the synthesis of LPS. Nevertheless, the HMW PS could only be visualised by immunoblotting with serotyping antibodies [13] whereas, in contrast to the core oligosaccharide, it could not be stained using silver. In addition, despite the structure of several core and ‘‘O-chain’’ polysaccharides being determined, the attachment point of the repeating subunits to the core was not defined. Initial genetic evidence suggested that C. jejuni possessed genes involved in LPS biosynthesis [14,15]. However, mutation of genes in this locus did not affect the expression of the putative O-chain as expected [16]. In addition, gene content did not vary in strains of different serotype (‘‘heat stable’’, or HS antigen serotype), although it was noted that an adjacent locus did vary between strains [16]. The observation that the genes initially thought to be involved in LPS/LOS biosynthesis had a role in protein glycosylation [17] showed why initial reservations as to a role in LPS biosynthesis were auspicious. The somewhat confusing genetic profile of LPS/LOS biosynthesis began to be unraveled when the full potential for polysaccharide biosynthesis by C. jejuni was realised [2]. Two important observations defined the relationship between the HMW PS and core PS linked to lipid A. Karlyshev et al. [18,19] demonstrated that the previously assumed ‘‘O-chain’’ containing serotype-specific epitopes is capsular and that mutation of capsule biosynthesis genes disrupted the HMW PS (Section 3.1). Conversely, Oldfield et al. [20] showed that a

379

mutation that truncated the core PS did not affect the HMW PS. 2.2. LOS structure in NCTC 11168 The lipid A moiety of C. jejuni LOS has been found to exhibit differences to that found in Escherichia coli [21]. In NCTC 11168 the major species of disaccharide backbone in lipid A is GlcN3N-GlcN (2,3-diamino2,3-dideoxy-D-glucose and D-glucosamine) phosphorylated with PPEtn (pyrophosphorylethanolamine) [9] and this is acylated with palmitic or lauric acid. However, the disaccharide backbone was found to vary, with GlcN3N–GlcN3N and GlcN–GlcN also present, and every backbone having differing phosphorylation patterns. The structure of the NCTC 11168 core PS was first postulated [16] based on a structure from a strain with a shared HS type [22], in hindsight an unsound strategy given HS type is now known to be capsular-derived. Direct chemical analysis of the core of NCTC 11168 was first reported by Oldfield et al. [20] and this was confirmed and modified by others [9,23]; the core structure is shown in Fig. 2. The core PS of NCTC 11168 contains a terminal structure analogous to the human gangliosides GM1a and GM2 (Fig. 2). In addition, phase variation is also evident in the core PS with the terminal b-1,3-linked galactose residue (Gal) being absent (GM2 core type) in a proportion of the population (see below; [23]).

2.3. Genetic basis of LOS biosynthesis in C. jejuni NCTC 11168 The presence of homologues of the lipid A biosynthesis genes lpxA, lpxC, lpxD, lpxB and lpxK in NCTC 11168 suggests that the lipid A biosynthesis pathway in C. jejuni is similar to that of E. coli [24]. These genes are not closely linked in C. jejuni and the absence of the production of LpxH (UDP-2,3-diacylglucosamine hydrolase) is not novel to C. jejuni, likely reflecting the

P/PEtn ↓ 6 β-Gal-(1→3)-β-D-GalNAc-(1→4)-β-Gal-(1→3)-β-Gal-(1→3)-L-α-D-Hep-(1→3)- L-α-D-Hep-(1→5)-Kdo

GM1a GM2

3 ↑ 2 α-Neu5Ac

2 ↑ 1 α-D-Gal

2 ↑ 1 β-D-Glc

4 ↑ 1 β-D-Glc

Fig. 2. The complete structure of the core oligosaccharide of NCTC 11168 LOS. The structure is based on those derived by Oldfield et al. [20], St Michael et al. [23] and Szymanski et al. [9]. The carbohydrate structures analogous to the gangliosides GM1a and GM2 are bracketed. Kdo, 3-deoxya-D-manno-oct-2-ulopyranosonic acid; Hep, L-glycero-D-mannoheptose; Glc, glucose; Gal, galactose; Neu5Ac, N-acetylneuraminic acid or sialic acid; GalNAc, N-acetyl-D-galactosamine.

380


presence of an unrecognised, but functionally equivalent enzyme. The gene encoding the Kdo (3-deoxy-a-D-manno-oct-2-ulopyranosonic acid) transferase WaaA (KdtA) is present, but only one late acetylase gene, htrB (lpxL, waaM), has been identified. Unusually, in C. jejuni htrB is located in the gene cluster encoding the core PS. Although mutation of many of the lpx genes in E. coli is lethal, mutation of lpxA is possible in N. meningitidis [25]. It is unknown if similar mutants can be made in C. jejuni, but mutation of htrB is possible ([26]; Millar and Ketley, unpublished data). The genes responsible for the biosynthesis of the core oligosaccharide of LOS are situated adjacent to the locus involved in N-linked glycosylation of proteins (Fig. 3). Heptosyltransferase I encoded by waaC attaches the first heptose (Hep-I) to Kdo and complements a Salmonella enterica serovar Typhimurium waaC mutant [27]. Attachment of the second heptose (Hep-II) by the Heptosyltransferase II, WaaF, was also verified by complementation and a mutant produced a truncated core oligosaccharide [20]. Genes likely to be involved in the biosynthesis of the nucleotide-activated L,D-heptose substrate (ADP-L-glycero-D-manno heptose) of the LOS heptosyltransferases are adjacent to waaF (Fig. 3) [28– 30]; only the function of the sedoheptulose-7-phosphate isomerase (gmhA) and D-a-D-Heptose-1,7-bisphosphate phosphatase (gmhB) have been investigated [7,30]. Hep-I and Hep-II have attached glucose residues (Fig. 2) that, based on sequence analysis, are likely to be attached by Cj1135 (Fig. 4), a putative glucosyltransferase containing two domains. One domain has similarity to b-1,4 glucosyltransferases from Neisseria and Haemophilus (LgtF) and therefore may transfer the glucose (Glc) to Hep-I [4]. The role of the second domain in the transfer of the Glc onto Hep-II is inferred from the observa-

protein glycosylation gene cluster

l ga

tion that in some C. jejuni strains cj1135 contains a frameshift mutation that disrupts the C-terminal domain; these strains lack a b-1,4 Glc attached to Hep-II [4]. The transfer of the b-1,3 galactose (Gal I) to Hep-II and subsequent attachment of a b-1,3 Gal (Gal II) and a-1,2 Gal (Gal III) to Gal I are possibly mediated by Cj1136, Cj1138 and Cj1137, respectively (Fig. 4). Cj1136 and Cj1138 have similarity to galactosyltransferases and Cj1137 to various glycosyltransferases [4]. Unfortunately, the predicted function of these genes is based entirely on sequence similarity and a choice of best candidate; experimental confirmation is awaited. A UDP-glucose 4-epimerase activity to interconvert between UDP-Gal and UDP-Glc is encoded by the galE gene (cj1131) [2] which maps to the boundary of the LOS and pgl gene clusters (Fig. 3). Originally reported to epimerise UDP-Glc based on complementation and enzymatic activity [31], recent studies have shown this gene (with a reannotation suggested as gne) to encode an enzyme with dual UDP-GlcNAc/Glc 4-epimerase activity. Mutation of galE in strain 81116 truncated core PS but did not affect capsule [31], whereas gne (galE) is required for three surface carbohydrate biosynthesis pathways (LOS, capsule and N-linked glycosylation) in NCTC 11168 [32]. The reason for this discrepancy is likely to reflect the lack of GalNAc in strain 81116 capsule [33]. The terminal sugars of the outer part of the NCTC 11168 core PS are part of a structure that mimics the GM1a ganglioside (Fig. 2). To the second core Gal is attached to an a-2,3 Neu5Ac (N-acetylneuraminic acid or sialic acid) (Fig. 2) and a set of three genes (cgtA/neuA1, neuB1 and neuC1) have been shown to be required for sialic acid biosynthesis (Fig. 4) [4,34–36]. In NCTC

LOS biosynthesis gene cluster

E

5 8 7 6 C N t 13 11 3 11 3 113 la aa trB j1 j j j w cs C w h C C C

45 tA 11 cg / 4 1 1 1 B uC uA 14 aV aF hA aD aE hB u 1 a a a a j w C ne w gm w w gm ne ne

Fig. 3. Representation of the genetic organisation of the LOS core oligosaccharide biosynthesis locus of NCTC 11168. The close linkage of the LOS biosynthesis gene cluster with the locus involved in N-linked glycosylation of proteins is illustrated. Names of genes involved in LOS biosynthesis are shown above each ORF with the direction of transcription indicated by an arrow. The genes containing homopolymeric tracts are denoted with a triangle.


381

Heptose biosynthesis GmhA D-sedoheptulose 7-phosphate

Cj1139 (WlaN)

WaaE

D,D-heptose 7phosphate

CgtA/NeuA1

GmhB D,D-heptose 1,7bisphosphate

Cj1138?

(Gal II)

WaaE

D,D-heptose 1phosphate

Cj1136?

WaaD ADP-D,Dheptose

WaaF

ADP-L,Dheptose

WaaC P/PEtn ↓ 6

(Gal I)

β-Gal-(1→3)-β-D-GalNAc-(1→4)-β-Gal-(1→3)-β-Gal-(1→ 3)-L-α-D-Hep-(1→3)- L-α-D-Hep-(1→5)-Kdo 3 ↑ 2 α-Neu5Ac Cst

2 ↑ 1 (Gal III) α-D-Gal

Cj1137?

2 ↑ 1 β-D-Glc

Cj1135?

4 ↑ 1 β-D-Glc Cj1135? Sialic acid biosynthesis

CgtA/NeuA1 CMP- N-acetylneuraminic acid

NeuB1

N-acetylneuraminic acid

NeuC1 N-acetyl-Dmannosamine

Fig. 4. Hypothetical model for the biosynthesis of C. jejuni NCTC 11168 LOS core oligosaccharide. Based on the discussion in the text, the biosynthetic role in the formation of core oligosaccharide by the enzymes encoded by the genes of the LOS gene cluster is illustrated. The point at which each enzyme acts is indicated by an arrowed box. Where functional role is not supported with direct experimental evidence, the enzyme name is followed with a question mark. The proposed pathways for heptose biosynthesis and sialic acid biosynthesis are shown in the boxes. For the bi-functional CgtA/NeuA1 the relevant domain is highlighted in bold.

11168 the gene annotated as neuA1 (cj1143 and here termed cgtA/neuA1) encodes a two domain bi-functional protein, whereas in other strains (see below) these domains are encoded by separate ORFs (Figs. 3, 5). The CgtA/NeuA1 N-terminal domain has b-1,4-N-acetylgalactosaminyltransferase activity (see below) with the C-terminal being a CMP-Neu5Ac synthase [4]. The neuB1 gene encodes one of three functional Neu5Ac synthetases in the NCTC 11168 genome but only mutation of neuB1 affects core PS mobility [36]. Mutation of the neuC1 gene that encodes a putative N-acetylglucosamine-6-phosphate-2-epimerase verified a role in sialic acid biosynthesis [34,35]. In addition to the a-2,3 Neu5Ac, a b-1,4 GalNAc (N-acetyl-D-galactosamine) is attached to the distal b1,3 Gal of the core (Fig. 2). In NCTC 11168 this attachment is mediated by the N-terminal domain of CgtA/ NeuA1 (Cj1143) (Fig. 4) that has a b-1,4-N-acetylgalactosaminyltransferase activity demonstrated both genetically and enzymatically [35,37,38]. Finally, the terminal b-1,3 Gal is attached to the b-1,4 GalNAc by WlaN (Cj1139 or CgtB) that has b-1,3 galactosyltransferase activity [37] (Fig. 4). Like other mutations that disrupt the formation of the GM1a mimic in NCTC 11168 LOS, a wlaN mutant is no longer able to bind cholera toxin (CT), which can be used as a molecular probe [39].

2.4. Variation of LOS structure Modulation of the expression of surface structures is likely to be an important C. jejuni strategy for avoiding host defences and possibly adaptation to microenvironments. Various Campylobacter LOS structures have been described (reviewed in [40]). Phenotypic variation of C. jejuni LOS arises from both differences in the content of specific genes as well as high frequency sequence variation that leads to changes in gene expression. Reversible gene switching, known as phase variation [41], is associated with homopolymeric tracts and short sequence repeats and has been observed in Haemophilus influenzae, Helicobacter pylori and Neisseria species [42– 44]. Such variation is believed to arise due to slipped strand mispairing during chromosomal replication [45]. 2.4.1. Gene content variation With respect to core structures, it is well established in E. coli/S. enterica that structure reflects the genes present in individual strains [46]. Therefore, within certain limitations, for example the presence of other polymorphisms, determination of LOS gene cluster content will be informative with respect to LOS structure. In the LOS biosynthesis locus, gene content variation occurs predominantly between htrB and the possible

382


Fig. 5. Schematic diagram of different patterns of genetic organization in the LOS gene cluster (waaC to waaF region) found in strains of C. jejuni. A new classification scheme that groups related organization classes together is used and new groupings are named on the left. Each ORF is shown with the direction of transcription indicated by the arrow. Gene names according to the scheme used by Gilbert et al. [4,47] are shown between htrB and waaV for comparison. Genes shown as black arrows are shared by all groups, those shown as grey arrows are shared by members of each group and those as white arrows vary within each group. Homology between genes in each group is illustrated by the stripped boxes. Genes that contain homopolymeric tracts are indicated by a diamond.

glucosyltransferase gene cj1146 (waaV) [4,26,47] (Fig. 5). In addition, some strains have additional genes inserted between waaF and gmhA [26] (Fig. 6). Focusing on the main region of variation (htrB to waaV), observed gene content diversity between strains ranges from single gene differences to all the genes present (Fig. 5), involves from 4 to 14 genes (Fig. 5) and some genes appear specific to a

content type [4,26,47]. As more information on the gene organisation in different strains has accumulated, different gene contents have been classified [4,47]. In order to reflect a more organisational and less historical system, a new gene content grouping scheme is used in Fig. 5. As more strains are investigated it is now possible to devise models for the evolutionary events that link cer-

Fig. 6. Schematic diagram of different patterns of genetic organization in the LOS gene cluster (waaV to gmhB region) found in strains of C. jejuni. Each ORF is shown with the direction of transcription indicated by the arrow. Genes shown as black arrows are shared by all classes and those as white arrows vary within each type. Variable genes are named provisionally to illustrate different content classes, using a scheme different to that used for those between waaC and waaF (Fig. 5). Homology between genes is illustrated by the striped boxes. Group type for the waaC–waaF region is shown on the right.


tain gene classes [47] with intra and intergenomic recombination events involved in the generation of novel gene content. The organisation patterns in group 1 (classes A, B and C; [47]) appear closely related, particularly 1a and 1b (Fig. 5). A model has been proposed [47] for the most parsimonious recombination events that explain the derivation of group 1a from 1b by gene duplication. Another example has been proposed to demonstrate horizontal transfer where the comparison of molecular typing data with LOS gene organisation suggests different phylogenetic relatedness therefore indicating intergenomic transfer of the LOS region [47]. The detailed investigation of gene function in NCTC 11168 has revealed evidence of genetic re-arrangements of domains. In many strains the cgtA and neuA1 genes are independent whereas in NCTC 11168 a genetic re-arrangement has formed a gene encoding a bi-functional protein CgtA/NeuA1 (Cj1143) [4]. 2.4.2. Sequence variation in homologues For many genes direct experimental functional data combined with sequence comparison provides a clue as to the role of genes found in new gene cluster variants [4,26,47]. However, small differences between apparently related ORFs can lead to different functionality and therefore as ever, care must be taken when predicting function on sequence comparisons alone. For example, the cgtA genes in group 1 strains appear to be allelic despite not having high levels of sequence identity [4]. Although all encode b-1,4-N-acetylgalactosaminyltransferases (CgtA, CgtA-I and CgtA-II) the receptor specificity varies markedly [4]. A similar situation has also been observed for the cst gene that encodes a sialyltransferase where different amino acid substitutions are associated with a-2,3 and a-2,8 activity or different acceptors [4,47]. Finally, in some genes non-phase variable frame shift mutations inactivate or truncate different transferases [47]. 2.4.3. Phase variation arising from homopolymeric tracts With the analysis of gene content in a large number of strains, homopolymeric tracts have been found in genes from all classes of LOS cluster organisation (Fig. 5). At least nine genes have been found to contain tracts [47], but few have been studied. In NCTC 11168 two genes contain homopolymeric G/C tracts, cj1139 (wlaN, cgtB) and cj1144/cj1145 both of which were seen to vary in clones analysed during genome sequencing [2] (Fig. 3). The consequence of tract phase variation was shown convincingly with wlaN [39], where variants that lacked the terminal b-1,3 Gal were linked to a particular tract length (9 G residues) that introduced a frameshift mutation and premature stop codon thus converting a GM1a structure to a GM2 structure (Fig. 2). In contrast, a tract length of G(8) resulted in a full length product and a GM1a structure. Different degrees of tract length varia-

383

tion have been reported for wlaN, ranging from G(7-9) [26] through G(8-9) [2,39], to an absence of variation [48]. Phenotypically, such diversity in observed tract length variation of cj1139 results in isolate-specific differences that determine which ganglioside mimic predominates [4,23]. In cj1144/1145 tract lengths of C(8) and C(9) truncate cj1145 and C(10) leads to read-through translation into the cj1144 ORF. Only C(8) and C(9) were seen in the sequenced shotgun library clones [2] whereas C(5-9) were seen in a larger analysis ([26]; Millar and Ketley, unpublished data) suggesting that the formation of the fused Cj1144-45 protein is possible. In strain 81-176 the homopolymeric tract in the cgtA gene varied between G(9) and G(12) with G(10) allowing a full length product associated with a GM2 mimic [35]. Other variants produced a truncated core PS forming a GM3 mimic. As expected from the proposed mechanism of slipped strand mispairing [45], tracts consist of G/C residues, however, one example of a variable homopolymeric A tract has been reported in the cst-II sialyltransferase gene of one strain [47]. 2.5. Role of LOS in C. jejuni-mediated disease The degree of genetic variation that generates the diversity apparent in LOS structure signifies the functional importance of LOS for C. jejuni. The natural environment for C. jejuni is intestinal and therefore it is reasonable to suggest that LOS diversity is important for the ability to colonise a wide variety of hosts and intestinal niches. In addition, different LOS structures may be advantageous for survival in different non-intestinal environments. C. jejuni varies LOS biosynthesis genes through strategies that act at different rates, from rare multi-gene recombination events to high frequency homopolymeric tract variation. Such variation will facilitate the ability of C. jejuni to adapt to dynamic and hostile environments. The early work of McSweegan and Walker [49] indicated that C. jejuni LPS had a role in adherence and subsequent studies using defined LOS mutants support a role in both host cell adherence and invasion for LOS [31,35]. The mutation of galE that truncated LOS did not remove the ability to colonise the avian intestine, but did affect colonisation at lower doses [31]. The ability to generate variation at high frequency, the molecular mimicry evident in LOS structure and the phenotypes of certain LOS mutants support a role in the avoidance of host defences. Phase variable changes in LOS structure will present the host with a shifting set of surface antigens, many of which have similarity to host structures. The importance of sialic acid for immune avoidance is supported by the observation that a mutant lacking LOS sialic acid residues (but not GalNAc) showed greater immunoreactivity and decreased serum resistance [34,35].

384


The similarity of C. jejuni LOS structures to host gangliosides and the subsequent ability to generate crossreacting antibodies forms the pathological basis for the association of preceding C. jejuni infection with GBS [47,50,51]. The pathogenesis of GBS remains unclear; however, the mimicry of gangliosides associated with the generation of anti-ganglioside antibodies is thought to precipitate the clinical features of GBS. Nevertheless, given that most enteric infections with C. jejuni strains expressing ganglioside mimics do not result in GBS [52], other host and/or bacterial factors are also essential; these factors remain unknown.

3. Capsular polysaccharide Another class of Campylobacter cell surface molecules with a potential for structural variation is capsular polysaccharide (CPS). A large gene cluster encoding various genes with similarities to type II/III capsular polysaccharide-related genes from other bacteria was found in the course of the C. jejuni NCTC 11168 genome sequencing project [2]. This and a previous preliminary observation of capsule-like molecules [53] prompted a search for C. jejuni CPSs. The studies resulted not only in the detection and characterisation of the CPS [18,54], but also in demonstration of the related capsule by electron microscopy [19]. 3.1. Capsule polysaccharide structure The CPS was found to be the major component of the Penner serotyping scheme [18], which had long been associated with ‘‘LPS’’ [55], or its O-chain (Section 2.1). Therefore, the previously published structures of ‘‘O-chain’’ were in fact those of the CPS. The published data on the chemical analysis of ‘‘O-chain’’ samples (CPS) extracted from different Campylobacter strains demonstrated high variation in their structure (e.g. [56–58]). The structure of the CPS from the NCTC 11168 strain has been determined recently and contains an unusual phosphoramide side chain modification [9] that is now also known to be present in the CPS of other strains [7]. Interestingly, despite having almost identical cps gene clusters, CPS from strain HS:36, but not from strain HS:23, has this modification. The difference may be attributed to the variation in one or more of the contingency genes present in these clusters [23]. Another interesting feature of some, but not all CPSs, is the presence of heptosyl residues (Hep). This is correlated with the presence of genes responsible for the biosynthesis of Hep residues hddC, gmhA2 and hddA in the respective cps clusters [7]. The presence of phosphoramide side chains and Hep residues may be essential for the biological function of CPSs. The availability of the CPS structure will

facilitate studies assigning function to all 38 genes that constitute the NCTC 11168 CPS locus. 3.2. Capsule genetic diversity The cps gene clusters of C. jejuni contain three major regions. The internal region consists of the genes required for the biosynthesis of the CPS repeating units. This region is highly variable between the strains belonging to different serotypes reflecting the variety of CPS structures. In contrast, the two flanking regions consisting of the kps genes, which are required for polymerisation of the repeating units and for the translocation of the CPS to the cell surface, are more conserved [7,18]. Hybridisation analysis using DNA microarrays has demonstrated a correlation between strain serotypes and gene composition of the respective cps clusters [3]. Detailed mechanisms of genetic variation were revealed by comparative genomics, when complete sequences of the entire cps gene clusters from several C. jejuni strains became available [7]. These data demonstrate the presence of highly conserved genes in some strains, especially those involved in heptose biosynthesis, along with various gene insertions and deletions (Fig. 7). The presence of such highly conserved genes may be the result of inter-strain horizontal gene transfer. Another potential for genetic variation is provided by the presence of homopolymeric G tracts in several cps genes. Cps regions from six other strains in addition to NCTC 11168 (HS:2) have been sequenced: HS:1, HS:19, HS:23, HS:36, HS:23/36 and HS:41. The internal biosynthetic region (between the flanking kps genes involved in CPS transport) was found to vary between 15.2 kb (HS:1) and 34.1 kb (HS:2 and HS:41). In all cases a number of genes with a potential for phase variation have been revealed [7]. A reversible mechanism of switching these genes ON and OFF is likely to result in production of modified CPSs. Variation in the CPS structure may be essential for evasion of host immune response and this is supported by the finding that C. jejuni CPS is involved in virulence [59]. It was found that the 81-176 mutant lacking CPS revealed reduced attachment and invasion efficiency in the experiments with the INT407 cell line. In addition, the cps mutant of this strain had increased hydrophobicity and decreased resistance to human serum. Moreover, the CPS-deficient mutant was less potent in developing diarrhoea in a ferret model of infection [59]. The presence of heptose residues in the CPS of C. jejuni may also be important for virulence. Indeed, heptose residues found in some cell surface-located glycoconjugates are required for adhesion [60]. The CPS may also be involved in enhanced survival of C. jejuni outside its animal hosts due to, for example, protection from desiccation, which may partly explain the persistent nature of the pathogen in the environment.


385

Fig. 7. Example of mosaic gene structure in C. jejuni cps regions. Genetic comparison of sections of the capsule loci from strains 81-176 (bottom) and the sequenced strain NCTC 11168 (top) using Artemis Comparison Tool [84]. The genes encoding sugar biosynthesis-related and hypothetical products are shown in green and brown, respectively. The genes resulted from insertion/deletion events, the highly variable copies of fcl genes and the regions containing highly conserved genes hddC, gmhA2 and hddA involved in the biosynthesis of Hep residues are marked by blue, yellow and pink ellipses, respectively.

4. O-linked protein glycosylation Glycosylation of the Campylobacter flagellum was first reported in 1989 [61]. Initial studies demonstrated that flagellar modifications in Campylobacter coli VC167 are linked to immunoreactivity and are affected by mutations in genes ptmA (cj1332) and ptmB (cj1331) [62,63]. Subsequent studies of flagellar modification in both C. coli and C. jejuni demonstrated the chemical nature of the glycan moieties decorating flagellins and revealed some other genes involved. These results were thoroughly summarised in a review article published two years ago [10]. These and some more recent data are described briefly below. 4.1. Glycan structure of modified flagellin The glycan moieties were initially considered to contain sialic acid (Neu5Ac) residues due to interaction with sialic acid-specific lectins. However, further analyses showed the presence of the residues of pseudaminic acid

(Pse5Ac7Ac), a nine-carbon sugar that is similar to sialic acid [64]. The observed cross-reaction in lectin studies could be attributed to structural similarities between pseudaminic and sialic acids. Among the modifications of pseudaminic acid is an acetamidino substituted analogue of pseudaminic acid (PseAm). The PseAm forms may also vary among different strains/species due to additional modification(s). For example, the PseAm residues in strains 81-176 and VC167 are slightly different possibly due to the position of attachment of the acetoamidino group, and, more importantly, are synthesised via different pathways [6,64]. In contrast to the Campylobacter N-linked protein glycosylation system that modifies dozens of proteins (see below), the only known O-linked glycosylated molecules are flagellar subunits (reviewed in: [10]). The reason for this difference may be attributed to the specific mechanisms of flagellar biosynthesis as well as to the particular function performed by flagella. The modification appears to be important for flagellar assembly, as mutations in some genes involved in pseudaminic acid

386


biosynthesis (eg cj1293, pseB) [11] result in non-motile and aflagellate cells [36]. Interestingly, in these mutants unmodified flagellar subunits are accumulated intracellularly [6], suggesting that glycosylation may be required for recognition by the flagellar secretion/assembly apparatus. However, the exact role of glycosylation in protein secretion is unknown. Despite similarities between the genetic basis and overall mechanism of flagella biosynthesis in various bacteria, in most cases (e.g. in E. coli) flagellin modification is not required for assembly. As many as 19 Ser and Thr sites on the FlaA and FlaB flagellin subunits can be glycosylated. In contrast to the N-linked glycosylation system, O-linked glycosylation does not depend on any amino acid sequence motif, and is apparently determined by the tertiary structure of the protein, with mainly surface exposed regions being modified [64]. 4.2. Genetic organisation and diversity of the O-linked flagella glycosylation system The O-linked glycosylation island of C. jejuni NCTC 11168 contains about 50 genes [2]. The genetic organisation of the flagellar glycosylation locus varies markedly not only between different species, but also between strains. For example, the 81-176 strain of C. jejuni lacks a large contiguous region corresponding to genes cj1318–cj1332 present in C. jejuni strain NCTC 11168 [6]. This deletion may be a result of the presence of several highly similar genes of the maf family in the latter strain, which may promote recombinational rearrangements. To date structural studies on the flagellin modification system have focused on C. jejuni 81-176 and the C. coli strain VC167 with no structural data reported for NCTC 11168. Given that NCTC 11168 contains several additional genes in its glycosylation locus compared to 81-176, it is likely that, in addition to Pse5Ac7Ac and PseAm the NCTC 11168 flagellins are modified by other unknown glycans. Homologues of the genes missing in 81-176 are involved in flagellin modification in the C. coli strain VC167 and are designated ptmA–ptmE [6]. Some other genes, for example, neuB2 and/or neuB3, may also participate in the biosynthesis of pseudaminic acid [64] in the process similar to the biosynthesis of sialic acid, which involves the product of gene neuB1 [36]. Another gene affecting glycan biosynthesis is cj1293. This gene was first shown to be involved in the biosynthesis of pseudaminic acid [11]. However, it was later speculated that this gene may be involved in bacillosamine biosynthesis [65] although no strong experimental evidence supporting this possibility was presented. When comparing the flagellar biosynthesis locus with those involved in the biosynthesis of other glycans/glycoconjugates, the most striking feature is the absence of homologues of putative glycosyltransferases. As the

function of a larger number of genes from the region becomes established it raises the possibility that some products of the remaining hypothetical genes from the locus may perform a transferase role. Among the candidates for this role are the Maf proteins. The presence of several similar but not identical putative transferases may reflect variation in their specificity to various activated forms of pseudaminic acid as a substrate. A number of genes in the flagellar glycosylation locus contain homopolymeric nucleotide tracts making them prone to phase variation. There are two paralogous groups of genes, cj0617/0618 and maf cj1318-like) families, most of which are located in the flagellar glycosylation locus and contain polyG tracts. Although the exact function of these genes is unknown, the maf genes were found to be involved in reversible production of flagella [5]. The availability of a mechanism of phase variable flagellar production and consequently motility, as well as a potential to vary the structure of the highly immunogenic glycan moieties on flagellin may be essential for bacterial adaptation to changing environmental conditions. In addition, as with LOS biosynthesis, variation in cell-surface exposed flagellar glycan structure may assist in bacterial evasion of the host immune response.

5. N-linked protein glycosylation In addition to the O-linked glycosylation system, C. jejuni has a second glycosylation system which is N-linked [10,12]. The possession of both glycosylation systems in C. jejuni is almost unprecedented among bacteria. The genes in the 16 kb N-linked glycosylation island were named pgl (Cjpgl) for protein glycosylation and were subsequently demonstrated to play a role in the glycosylation of dozens of C. jejuni proteins [17]. The locus encoding this glycosylation system is adjacent to the genes responsible for LOS biosynthesis (Fig. 3) and initially the pgl genes were attributed a role in LOS biosynthesis (Section 2.1) [14,15]. 5.1. Structure of N-linked glycan Recently, mass spectrometry and NMR spectrometry analyses have been used to determine the structure of the glycan as a heptasaccharide GalNAc-a1,4-GalNAc-a1,4-(Glcb1,3-)GalNAc-a1,4-GalNAc-a1,4-GalNAc-a1,3-Bac-b1,N-Asn, where Bac is bacillosamine, 2,4-diacetamido-2,4,6-trideoxyglucose [66,67]. Structural determination facilitated the elucidation of the roles of the genes in the N-linked glycosylation locus. 5.2. Genetic basis of N-linked glycosylation system The genetic locus encoding the N-linked general glycosylation system (Fig. 8) is highly conserved, with


387

Fig. 8. N-linked glycosylation locus and proposed gene function in C. jejuni NCTC11168.

variation in the presence of only one gene (wlaJ) being reported [16]. The key gene product in N-linked general glycosylation is the putative oligosaccharyltransferase PglB that is likely to be responsible for coupling the glycan to the asparagine residue to form the N-linked glycoprotein [10,12]. PglB has significant similarity to STT3, which is found almost exclusively in eukaryotes and in archaea [68]. In eukaryotes, STT3 has been shown to play a central role in the oligosaccharyltransferase complex and is evolutionarily conserved [69]. The potential roles of the rest of the genes in the pathway can be hypothesised based on sequence similarity to glycosyltransferases and enzymes required for sugar modification as well as recent information from mutagenesis and mass spectroscopy studies on the glycan structure [70]. These hypothesised roles are summarised in Fig. 8. Additionally, some components of the pathway have similarity to the O-linked glycosylation pathway of Neisseria meningitidis pilin and to other organisms that synthesise bacillosamine derivatives [71]. Based on these studies a model for the synthesis of N-linked glycoproteins has been proposed [10]. Initially UDP-HexNAc is converted to bacillosamine via sequential modification by PglF (dehydratase), PglE (aminotransferase) and PglD (acetyltransferase) in the cytoplasm [10,12]. PglC then attaches the Bac residue to a lipid carrier. Based on the role of the PglA orthologue in N. meningitidis that attaches the a-1,3-linked Gal to diacetamido-trideoxy hexose (DATDH) [69,72], it is proposed that PglA in C. jejuni adds the a-1,3-linked GalNAc to Bac. PglH, an a-1,4-GalNAc transferase [73], has similarity to PglJ suggesting that both may be involved in adding the next four a-1,4-linked GalNAc residues [10,12]. The final glycosyltransferase in the pathway, PglI, adds the branched glucose. Using the analogy of the characterised N-linked eukaryotic pathway in Saccharomyces cerevisiae in which the assembled sugars are flipped from the cytoplasm into the lumen of the endoplasmic reticulum [74], in C. jejuni it is proposed

that the entire heptasaccharide is flipped across the inner membrane into the periplasm by the putative ABC transporter, WlaB [10,12]. 5.3. Role of N-linked glycosylation system Determining why C. jejuni requires the N-linked general glycosylation system is difficult to ascertain, because disruption of the pathway effectively removes the heptasaccharide from several dozen glycoproteins. However, disruption of N-linked glycosylation in Campylobacter results in cells that have a reduced capacity to attach to and invade in vitro cultured eukaryotic cells and a reduced ability to colonise the intestines of chickens and mice [75–78]. Additionally, in some strains of C. jejuni, a loss of type IV complex protein assembly and competence has been found [79] when N-linked glycosylation is absent. Other possible functions of the N-linked glycosylation system may include protection against proteolytic cleavage, or enhancement of protein stability or signals for cellular sorting as has been suggested for analogous eukaryote N-linked glycans [80,81].

6. Conclusion Since the completion of the NCTC 11168 genome sequence in 2000, rapid progress has been made towards the genetic and structural characterisation of the known glycolipids (LOS and capsule) and the two glycosylation pathways in C. jejuni. A relatively large proportion of the genome (>8%) is dedicated to the biosynthesis of C. jejuni cell surface carbohydrates, and further structures may yet be revealed. In the near future spectroscopic analyses complemented with metabolomic studies should enhance progress towards the full Campylobacter glycome and this will include knowledge of the activated sugar intermediates involved in the respective biosynthetic and metabolic pathways.

388


An interesting feature emerging from Campylobacter glycome studies is that some enzymes can be ‘‘promiscuous’’ in their specificity or functionally redundant, thereby complicating the analysis of metabolic synthesis or conversion. A recent example of this phenomenon is the bifunctional UDP-GlcNAc 4-epimerase (Gne, formerly annotated as GalE) that provides Gal and GalNAc residues for the biosynthesis of LOS, capsule and the N-linked glycan [32]. Further examples of shared glycan pathways are likely to be found. Another common feature is the extraordinary structural variability of the LOS, capsule and O-linked glycan between strains, which is mediated by multiple genetic mechanisms of variation. This is likely to enhance the ability of the organism to increase its host range and/or avoid the host immune defence. In contrast, the N-linked glycan that modifies at least 30 proteins is highly conserved among strains and even among Campylobacter species [82]. The cellular location of the N-linked glycoproteins appears variable [8], but at least some are likely to be surface located making them good vaccine candidates [83]. The reasons why the glycan modification of these surface structures is so conserved is unknown, but perhaps they are masked to some extent by protruding glycolipids. Despite the deluge of sequence and structural data, we still require a deeper understanding of the biological significance of the Campylobacter cell surface carbohydrates. These determinants are likely to play important roles in both avoiding and interacting with the avian and human host immune systems. Such endeavours may be particularly rewarding in terms of understanding both the commensal and pathogenic nature of this intriguing bacterium.

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

Acknowledgements

[12]

The authors acknowledge funding from the BBSRC exploiting functional genomics initiative for research into deciphering the Campylobacter glycome. We acknowledge Christine Szymanski for reviewing the manuscript.

[13]

[14]

[15]

References [1] Allos, B.M. (2001) Campylobacter jejuni infections: update on emerging issues and trends. Clin. Infect. Dis. 32, 1201–1206. [2] Parkhill, J., Wren, B.W., Mungall, K., Ketley, J.M., Churcher, C., Basham, D., Chillingworth, T., Davies, R.M., Feltwell, T., Holroyd, S., Jagels, K., Karlyshev, A.V., Moule, S., Pallen, M.J., Penn, C.W., Quail, M.A., Rajandream, M.A., Rutherford, K.M., van Vliet, A.H., Whitehead, S. and Barrell, B.G. (2000) The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403, 665–668. [3] Dorrell, N., Mangan, J.A., Laing, K.G., Hinds, J., Linton, D., AlGhusein, H., Barrell, B.G., Parkhill, J., Stoker, N.G., Karlyshev,

[16]

[17]

[18]

A.V., Butcher, P.D. and Wren, B.W. (2001) Whole genome comparison of Campylobacter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11, 1706–1715. Gilbert, M., Karwaski, M.F., Bernatchez, S., Young, N.M., Taboada, E., Michniewicz, J., Cunningham, A.M. and Wakarchuk, W.W. (2002) The genetic bases for the variation in the lipooligosaccharide of the mucosal pathogen, Campylobacter jejuni. Biosynthesis of sialylated ganglioside mimics in the core oligosaccharide. J. Biol. Chem. 277, 327–337. Karlyshev, A.V., Linton, D., Gregson, N.A. and Wren, B.W. (2002) A novel paralogous gene family involved in phase-variable flagella-mediated motility in Campylobacter jejuni. Microbiology 148, 473–480. Logan, S.M., Kelly, J.F., Thibault, P., Ewing, C.P. and Guerry, P. (2002) Structural heterogeneity of carbohydrate modifications affects serospecificity of Campylobacter flagellins. Mol. Microbiol. 46, 587–597. Karlyshev, A.V., Champion, O.L., Churcher, C., Brisson, J.R., Jarrell, H.C., Gilbert, M., Brochu, D., St Michael, F., Li, J., Wakarchuk, W.W., Goodhead, I., Sanders, M., Stevens, K., White, B., Parkhill, J., Wren, B.W. and Szymanski, C.M. (2005) Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol. Microbiol. 55, 90–103. Young, N.M., Brisson, J.R., Kelly, J., Watson, D.C., Tessier, L., Lanthier, P.H., Jarrell, H.C., Cadotte, N., St Michael, F., Aberg, E. and Szymanski, C.M. (2002) Structure of the N-linked glycan present on multiple glycoproteins in the Gram-negative bacterium, Campylobacter jejuni. J. Biol. Chem. 277, 42530–42539. Szymanski, C.M., St Michael, F., Jarrell, H.C., Li, J., Gilbert, M., Larocque, S., Vinogradov, E. and Brisson, J.R. (2003) Detection of conserved N-linked glycans and phase variable lipo-oligosaccharides and capsules from Campylobacter cells by mass spectrometry and high resolution magic angle spinning NMR spectroscopy. J. Biol. Chem. 278, 24509–24520. Szymanski, C.M., Logan, S.M., Linton, D. and Wren, B.W. (2003) Campylobacter – a tale of two protein glycosylation systems. Trends Microbiol. 11, 233–238. Goon, S., Kelly, J.F., Logan, S.M., Ewing, C.P. and Guerry, P. (2003) Pseudaminic acid, the major modification on Campylobacter flagellin, is synthesized via the Cj1293 gene. Mol. Microbiol. 50, 659–671. C. Szymanski, B.W. Wren, Protein glycosylation in mucosal bacterial pathogens, Nat. Rev. Microbiol. (in press). Penner, J.L., Hennessy, J.N. and Congi, R.V. (1983) Serotyping of Campylobacter jejuni and Campylobacter coli on the basis of thermostable antigens. Eur. J. Clin. Microbiol. 2, 378–383. Fry, B.N., Korolik, V., ten Brinke, J.A., Pennings, M.T., Zalm, R., Teunis, B.J., Coloe, P.J. and van der Zeijst, B.A. (1998) The lipopolysaccharide biosynthesis locus of Campylobacter jejuni 81116. Microbiology 144 (Pt 8), 2049–2061. Korolik, V., Fry, B.N., Alderton, M.R., van der Zeijst, B.A. and Coloe, P.J. (1997) Expression of Campylobacter hyoilei lipooligosaccharide (LOS) antigens in Escherichia coli. Microbiology 143 (Pt 11), 3481–3489. Wood, A.C., Oldfield, N.J., OÕDwyer, C.A. and Ketley, J.M. (1999) Cloning, mutation and distribution of a putative lipopolysaccharide biosynthesis locus in Campylobacter jejuni. Microbiology 145, 379–388. Szymanski, C.M., Yao, R., Ewing, C.P., Trust, T.J. and Guerry, P. (1999) Evidence for a system of general protein glycosylation in Campylobacter jejuni. Mol. Microbiol. 32, 1022–1030. Karlyshev, A.V., Linton, D., Gregson, N.A., Lastovica, A.J. and Wren, B.W. (2000) Genetic and biochemical evidence of a Campylobacter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 35, 529–541.

A.V. Karlyshev et al. / FEMS Microbiology Reviews 29 (2005) 377–390 [19] Karlyshev, A.V., McCrossan, M.V. and Wren, B.W. (2001) Demonstration of polysaccharide capsule in Campylobacter jejuni using electron microscopy. Infect. Immun. 69, 5921–5924. [20] Oldfield, N.J., Moran, A.P., Millar, L.A., Prendergast, M.M. and Ketley, J.M. (2002) Characterization of the Campylobacter jejuni heptosyltransferase II gene, waaF, provides genetic evidence that extracellular polysaccharide is lipid A core independent. J. Bacteriol. 184, 2100–2107. [21] aMoran, A.P. (1995) Biological and serological characterization of Campylobacter jejuni lipopolysaccharides with deviating core and lipid A structures. FEMS Immunol. Med. Microbiol. 11, 121–130. [22] Aspinall, G.O., McDonald, A.G., Raju, T.S., Pang, H., Kurjanczyk, L.A., Penner, J.L. and Moran, A.P. (1993) Chemical structure of the core region of Campylobacter jejuni serotype O:2 lipopolysaccharide. Eur. J. Biochem. 213, 1029–1037. [23] St Michael, F., Szymanski, C.M., Li, J., Chan, K.H., Khieu, N.H., Larocque, S., Wakarchuk, W.W., Brisson, J.R. and Monteiro, M.A. (2002) The structures of the lipooligosaccharide and capsule polysaccharide of Campylobacter jejuni genome sequenced strain NCTC 11168. Eur. J. Biochem. 269, 5119–5136. [24] Raetz, C.R.H. (1996) Bacterial lipopolysaccharides: a remarkable family of bioactive macroamphiphiles In: Escherichia Coli and Salmonella: Cellular and Molecular Biology (Neidhardt, F.C., Ed.), pp. 135–1071. ASM Press, Washington, DC. [25] Steeghs, L., den Hartog, R., den Boer, A., Zomer, B., Roholl, P. and van der Ley, P. (1998) Meningitis bacterium is viable without endotoxin. Nature 392, 449–450. [26] Millar, L.A. (2003) Diversity and Function of the Lipopolysaccharide Biosynthesis Genes from Campylobacter Jejuni. University of Leicester, Leicester. [27] Klena, J.D., Gray, S.A. and Konkel, M.E. (1998) Cloning, sequencing, and characterization of the lipopolysaccharide biosynthetic enzyme heptosyltransferase I gene (waaC) from Campylobacter jejuni and Campylobacter coli. Gene 222, 177–185. [28] Kneidinger, B., Graninger, M., Puchberger, M., Kosma, P. and Messner, P. (2001) Biosynthesis of nucleotide-activated D-glyceroD-manno-heptose. J. Biol. Chem. 276, 20935–20944. [29] Linton, D., Karlyshev, A.V. and Wren, B.W. (2001) Deciphering Campylobacter jejuni cell surface interactions from the genome sequence. Curr. Opin. Microbiol. 4, 35–40. [30] Szymanski, C.M., St Michael, F., Monteiro, M.A., Shen, H., Conlan, W. and Wakarchuk, W.W. (2001) Heptose biosynthetic pathways in Campylobacter jejuni NCTC 11168. Int. J. Med. Microbiol. 291, 80. [31] Fry, B.N., Feng, S., Chen, Y.Y., Newell, D.G., Coloe, P.J. and Korolik, V. (2000) The galE gene of Campylobacter jejuni is involved in lipopolysaccharide synthesis and virulence. Infect. Immun. 68, 2594–2601. [32] Bernatchez, S., Szymanski, C.M., Ishiyama, N., Li, J., Jarrell, H.C., Lau, P.C., Berghuis, A.M., Young, N.M. and Wakarchuk, W.W. (2004) A single bifunctional UDP-GlcNAc/Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni. J. Biol. Chem... [33] Muldoon, J., Shashkov, A.S., Moran, A.P., Ferris, J.A., Senchenkova, S.N. and Savage, A.V. (2002) Structures of two polysaccharides of Campylobacter jejuni 81116. Carbohydr Res 337, 2223–2229. [34] Guerry, P., Ewing, C.P., Hickey, T.E., Prendergast, M.M. and Moran, A.P. (2000) Sialylation of lipooligosaccharide cores affects immunogenicity and serum resistance of Campylobacter jejuni. Infect. Immun. 68, 6656–6662. [35] Guerry, P., Szymanski, C.M., Prendergast, M.M., Hickey, T.E., Ewing, C.P., Pattarini, D.L. and Moran, A.P. (2002) Phase variation of Campylobacter jejuni 81-176 lipooligosaccharide affects ganglioside mimicry and invasiveness in vitro. Infect. Immun. 70, 787–793.

389

[36] Linton, D., Karlyshev, A.V., Hitchen, P.G., Morris, H.R., Dell, A., Gregson, N.A. and Wren, B.W. (2000) Multiple N-acetyl neuraminic acid synthetase (neuB) genes in Campylobacter jejuni: identification and characterization of the gene involved in sialylation of lipo-oligosaccharide. Mol. Microbiol. 35, 1120– 1134. [37] Gilbert, M., Brisson, J.R., Karwaski, M.F., Michniewicz, J., Cunningham, A.M., Wu, Y., Young, N.M. and Wakarchuk, W.W. (2000) 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. 275, 3896–3906. [38] Guerry, P., Alm, R., Szymanski, C. and Trust, T.J. (2000) Structure, function and antigenicity of Campylobacter flagella In: Campylobacter (Nachamkin, I. and Blaser, M.J., Eds.), pp. 405–421. American Society for Microbiology, Washington, DC. [39] Linton, D., Gilbert, M., Hitchen, P.G., Dell, A., Morris, H.R., Wakarchuk, W.W., Gregson, N.A. and Wren, B.W. (2000) Phase variation of a beta-1,3 galactosyltransferase involved in generation of the ganglioside GM1-like lipo-oligosaccharide of Campylobacter jejuni. Mol. Microbiol. 37, 501–514. [40] Moran, A.P., Penner, J.L. and Aspinall, G.O. (2000) Campylobacter polysaccharides In: Campylobacter (Nachamkin, I. and Blaser, M.J., Eds.), pp. 241–257. ASM Press, Washington. [41] Henderson, I.R., Owen, P. and Nataro, J.P. (1999) Molecular switches – the ON and OFF of bacterial phase variation. Mol. Microbiol. 33, 919–932. [42] Hood, D.W., Deadman, M.E., Jennings, M.P., Bisercic, M., Fleischmann, R.D., Venter, J.C. and Moxon, E.R. (1996) DNA repeats identify novel virulence genes in Haemophilus influenzae. Proc. Natl. Acad. Sci. USA 93, 11121–11125. [43] Kahler, C.M. and Stephens, D.S. (1998) Genetic basis for biosynthesis, structure, and function of meningococcal lipooligosaccharide (endotoxin). Crit. Rev. Microbiol. 24, 281–334. [44] Saunders, N.J., Peden, J.F., Hood, D.W. and Moxon, E.R. (1998) Simple sequence repeats in the Helicobacter pylori genome. Mol. Microbiol. 27, 1091–1098. [45] Hallet, B. (2001) Playing Dr Jekyll and Mr Hyde: combined mechanisms of phase variation in bacteria. Curr. Opin. Microbiol. 4, 570–581. [46] Heinrichs, D.E., Yethon, J.A. and Whitfield, C. (1998) Molecular basis for structural diversity in the core regions of the lipopolysaccharides of Escherichia coli and Salmonella enterica. Mol. Microbiol. 30, 221–232. [47] Gilbert, M., Godschalk, P.C.R., Parker, C.T., Endtz, H.P. and Wakarchuk, W.W. (2004) Genetic basis for the variation in the lipopolysaccharide outer core of Campylobacter jejuni and possible association of glycosyltransferase genes with post-infectious neuropathies In: Campylobacter: New Perspectives in Molecular and Cellular Biology (Ketley, J.M. and Konkel, M.E., Eds.), pp. 219–248. Horizon Bioscience. [48] Wassenaar, T.M., Wagenaar, J.A., Rigter, A., Fearnley, C., Newell, D.G. and Duim, B. (2002) Homonucleotide stretches in chromosomal DNA of Campylobacter jejuni display high frequency polymorphism as detected by direct PCR analysis. FEMS Microbiol. Lett. 212, 77–85. [49] McSweegan, E. and Walker, R.I. (1986) Identification and characterization of two Campylobacter jejuni adhesins for cellular and mucous substrates. Infect. Immun. 53, 141–148. [50] Yuki, N., Susuki, K., Koga, M., Nishimoto, Y., Odaka, M., Hirata, K., Taguchi, K., Miyatake, T., Furukawa, K., Kobata, T. and Yamada, M. (2004) Carbohydrate mimicry between human ganglioside GM1 and Campylobacter jejuni lipooligosaccharide causes Guillain-Barre syndrome. Proc. Natl. Acad. Sci. USA 101, 11404–11409.

390


[51] Nachamkin, I., Allos, B.M. and Ho, T. (1998) Campylobacter species and Guillain-Barre syndrome. Clin. Microbiol. Rev. 11, 555–567. [52] Nachamkin, I., Ung, H., Moran, A.P., Yoo, D., Prendergast, M.M., Nicholson, M.A., Sheikh, K., Ho, T., Asbury, A.K., McKhann, G.M. and Griffin, J.W. (1999) Ganglioside GM1 mimicry in Campylobacter Strains from Sporadic Infections in the United States. J. Infect. Dis. 179, 1183–1189. [53] Chart, H., Frost, J.A., Oza, A., Thwaites, R., Gillanders, S. and Rowe, B. (1996) Heat-stable serotyping antigens expressed by strains of Campylobacter jejuni are probably capsular and not long-chain lipopolysaccharide. J. Appl. Bacteriol. 81, 635–640. [54] Karlyshev, A.V. and Wren, B.W. (2001) Detection and initial characterization of novel capsular polysaccharide among diverse Campylobacter jejuni strains using alcian blue dye. J. Clin. Microbiol. 39, 279–284. [55] Moran, A.P. and Penner, J.L. (1999) Serotyping of Campylobacter jejuni based on heat-stable antigens: relevance, molecular basis and implications in pathogenesis. J. Appl. Microbiol. 86, 361–377. [56] Aspinall, G.O., McDonald, A.G. and Pang, H. (1992) Structures of the O chains from lipopolysaccharides of Campylobacter jejuni serotypes O:23 and O:36. Carbohydr. Res. 231, 13–30. [57] Aspinall, G.O., McDonald, A.G. and Pang, H. (1994) Lipopolysaccharides of Campylobacter jejuni serotype O:19: structures of O antigen chains from the serostrain and two bacterial isolates from patients with the Guillain-Barre´ syndrome. Biochemistry 33, 250– 255. [58] Aspinall, G.O., Lynch, C.M., Pang, H., Shaver, R.T. and Moran, A.P. (1995) Chemical structures of the core region of Campylobacter jejuni O:3 lipopolysaccharide and an associated polysaccharide. Eur. J. Biochem. 231, 570–578. [59] Bacon, D.J., Szymanski, C.M., Burr, D.H., Silver, R.P., Alm, R.A. and Guerry, P. (2001) A phase-variable capsule is involved in virulence of Campylobacter jejuni 81-176. Mol. Microbiol. 40, 769–777. [60] Benz, I. and Schmidt, M.A. (2001) Glycosylation with heptose residues mediated by the aah gene product is essential for adherence of the AIDA-I adhesin. Mol. Microbiol. 40, 1403– 1413. [61] Logan, S.M., Trust, T.J. and Guerry, P. (1989) Evidence for posttranslational modification and gene duplication of Campylobacter flagellin. J. Bacteriol. 171, 3031–3038. [62] Doig, P., Kinsella, N., Guerry, P. and Trust, T.J. (1996) Characterization of a post-translational modification of Campylobacter flagellin: identification of a sero-specific glycosyl moiety. Mol. Microbiol. 19, 379–387. [63] Guerry, P., Doig, P., Alm, R.A., Burr, D.H., Kinsella, N. and Trust, T.J. (1996) Identification and characterization of genes required for post-translational modification of Campylobacter coli VC167 flagellin. Mol. Microbiol. 19, 369– 378. [64] Thibault, P., Logan, S.M., Kelly, J.F., Brisson, J.R., Ewing, C.P., Trust, T.J. and Guerry, P. (2001) Identification of the carbohydrate moieties and glycosylation motifs in Campylobacter jejuni flagellin. J. Biol. Chem. 276, 34862–34870. [65] Creuzenet, C. (2004) Characterization of CJ1293, a new UDPGlcNAc C6 dehydratase from Campylobacter jejuni. FEBS Lett. 559, 136–140. [66] Zufferey, R., Knauer, R., Burda, P., Stagljar, I., te Heesen, S., Lehle, L. and Aebi, M. (1995) STT3, a highly conserved protein required for yeast oligosaccharyl transferase activity in vivo. EMBO J. 14, 4949–4960. [67] Knauer, R. and Lehle, L. (1999) The oligosaccharyltransferase complex from yeast. Biochim. Biophys. Acta 1426, 259– 273.

[68] Yan, Q. and Lennarz, W.J. (2002) Studies on the function of oligosaccharyl transferase subunits. Stt3p is directly involved in the glycosylation process. J. Biol. Chem. 277, 47692–47700. [69] Virji, M. (2000) (Doyle, R.J., Ed.), Glycomicrobiology, pp. 31–65. Kluwer Academic/Plenum Publishers, New York. [70] D. Linton, N. Dorrell, P.G. Hitchen, S. Amber, A.V. Karlyshev, H.R. Morris, A. Dell, M.A. Valvano, M. Aebi, Functional analysis of a novel bacterial N-linked protein glycosylation pathway cloned in Escherichia coli, Mol Microbiology (in press). [71] Stimson, E., Virji, M., Makepeace, K., Dell, A., Morris, H.R., Payne, G., Saunders, J.R., Jennings, M.P., Barker, S. and Panico, M., et al. (1995) Meningococcal pilin: a glycoprotein substituted with digalactosyl 2,4-diacetamido-2,4,6-trideoxyhexose. Mol. Microbiol. 17, 1201–1214. [72] Hegge, F.T., Hitchen, P.G., Aas, F.E., Kristiansen, H., Lovold, C., Egge-Jacobsen, W., Panico, M., Leong, W.Y., Bull, V., Virji, M., Morris, H.R., Dell, A. and Koomey, M. (2004) Unique modifications with phosphocholine and phosphoethanolamine define alternate antigenic forms of Neisseria gonorrhoeae type IV pili. Proc. Natl. Acad. Sci. USA 101, 10798–10803. [73] T. Endo, S. Kakita, S. Koitsumi, A. Ozaki, Alpha-1,4-Nacetylgalactosamine transferase gene and process for producing this enzyme and N-acetylgalactosamine-containing complex carbohydrate, in: PCT WO 01/77337, 2002. [74] Aebi, M. and Hennet, T. (2001) Congenital disorders of glycosylation: genetic model systems lead the way. Trends Cell Biol. 11, 136–141. [75] Szymanski, C.M., Burr, D.H. and Guerry, P. (2002) Campylobacter protein glycosylation affects host cell interactions. Infect. Immun. 70, 2242–2244. [76] Hendrixson, D.R. and DiRita, V.J. (2004) Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52, 471–484. [77] C.M. Szymanski, S. Goon, B. Allan, P. Guerry, Protein glycosylation in Campylobacter. Campylobacter: Molecular and Cellular Biology (in press). [78] Jones, M.A., Marston, K.L., Woodall, C.A., Maskell, D.J., Linton, D., Karlyshev, A.V., Dorrell, N., Wren, B.W. and Barrow, P.A. (2004) Adaptation of Campylobacter jejuni NCTC11168 to high-level colonization of the avian gastrointestinal tract. Infect. Immun. 72, 3769–3776. [79] Larsen, J.C., Szymanski, C. and Guerry, P. (2004) N-linked protein glycosylation is required for full competence in Campylobacter jejuni 81-176. J. Bacteriol. 186, 6508–6514. [80] Herrmann, J.L., OÕGaora, P., Gallagher, A., Thole, J.E. and Young, D.B. (1996) Bacterial glycoproteins: a link between glycosylation and proteolytic cleavage of a 19 kDa antigen from Mycobacterium tuberculosis. EMBO J. 15, 3547–3554. [81] Helenius, A. and Aebi, M. (2001) Intracellular functions of Nlinked glycans. Science 291, 2364–2369. [82] Fouts, D.E., Mongodin, E.F., Mandrell, R.E., Miller, W.G., Rasko, D.A., Ravel, J., Brinkac, L.M., DeBoy, R.T., Parker, C.T., Daugherty, S.C., Dodson, R.J., Durkin, A.S., Madupu, R., Sullivan, S.A., Shetty, J.U., Ayodeji, M.A., Shvartsbeyn, A., Schatz, M.C., Badger, J.H., Fraser, C.M. and Nelson, K.E. (2005) Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3, 1–14. [83] Linton, D., Allan, E., Karlyshev, A.V., Cronshaw, A.D. and Wren, B.W. (2002) Identification of N-acetylgalactosamine-containing glycoproteins PEB3 and CgpA in Campylobacter jejuni. Mol. Microbiol. 43, 497–508. [84] Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M.A. and Barrell, B. (2000) Artemis: sequence visualization and annotation. Bioinformatics 16, 944–945.