How a sugary bug gets through the day

Abstract

Campylobacter jejuni is a highly prevalent yet fastidious bacterial pathogen that poses a significant health burden worldwide. Lacking many hallmark virulence factors, it is becoming increasingly clear that C. jejuni pathogenesis involves different strategies compared with other well-characterized enteric organisms. This includes the involvement of basic biological processes and cell envelope glycans in a number of aspects related to pathogenesis. The past few years have seen significant progress in the understanding of these pathways and how they relate to C. jejuni fundamental biology, stress survival, colonization, and virulence attributes. This review focuses on recent studies in three general areas where “pathogenesis” and “basic biology” overlap: physiology, stress responses and glycobiology.

Introduction

Campylobacter jejuni, a Gram-negative, helical ε-proteobacterium, is the leading cause of bacterial gastroenteritis in the developed world, outnumbering E. coli and Salmonella spp combined and infecting up to 1% of North American, European and Australian populations each year. Infection of humans with C. jejuni causes severe, watery to bloody diarrhea, fever, nausea and vomiting and can be lethal in very young, old and immunocompromised individuals. Serious sequelae include irritable bowel syndrome and arthritis; C. jejuni infection is also the leading antecedent to development of the debilitating, sometimes-fatal neuropathy Guillain-Barre´ Syndrome (GBS) (for general reviews on C. jejuni pathogenesis, please see refs. 1–4).

Despite causing severe human disease, C. jejuni is a zoonotic organism and resides harmlessly in the intestinal tracts of birds and other animal species. It is also fastidious, requiring lower O₂ and higher CO₂ concentrations than ambient air, rich growth medium and temperatures ranging from 37°C (humans) to 42°C (birds). The primary mode of C. jejuni transmission to humans is the consumption of contaminated poultry products, or the cross-contamination of other food with raw poultry juice. C. jejuni can also be transmitted via consumption of contaminated water and raw milk.

Although thought to be primarily an extracellular organism, C. jejuni has been shown to adhere to, invade, survive inside, and transcytose through epithelial cells in susceptible hosts. As small animal models of disease are in nascent stages of development, the ability of the bacteria to execute these processes in vitro is frequently used as a marker for virulence.

To date, the only C. jejuni factor thought to be dedicated to virulence has been a cytolethal distending toxin (CDT) with DNase properties, and that causes cell cycle arrest.^5,6 However, recent work showing that campylobacteriosis caused by CDT-negative strains is clinically identical to that of patients infected with a strain producing high levels of CDT has called into question the importance of the CDT in disease.⁷ Several genome sequences^8-10 have likewise identified little in the way of specific virulence mechanisms, instead showing that C. jejuni has a relatively small (~1.6 Mb) genome lacking canonical virulence factors such as type III secretion systems known to be important in pathogenesis of other enteric organisms. C. jejuni also lacks hallmark stress response elements such as the stationary phase sigma factor RpoS, yet nonetheless survives and thrives in numerous in vivo and environmental niches. However, genome sequences did reveal that a significant proportion of the genome encodes enzymes involved in carbohydrate biosynthesis. Likely differences in metabolic potential between strains were also identified. Collectively, it is becoming increasingly clear that C. jejuni uses different virulence mechanisms from other bacteria, and that “basic” processes play critical roles in pathogenic properties such as colonization, host cell interactions and stress survival.

This review focuses on several new developments made over the past few years in three areas related to the fundamental biology of C. jejuni: (1) physiology, (2) stress responses and (3) glycobiology (Fig. 1). Other aspects of C. jejuni biology such as iron homeostasis, motility and chemotaxis, DNA uptake, nitrogen metabolism and nitrosative stress responses and polyphosphate are covered extensively in other recent reviews^1,11-14 and are thus not discussed here, nor are older findings regarding C. jejuni pathogenesis which are also well-covered elsewhere.^3,4

Figure 1.

Graphical summary of topics highlighted in this review on fundamental biology underlying Campylobacter jejuni survival and pathogenesis. Word cloud generated with wordle.net; graphic of C. jejuni by Thuan Nguyen.

Physiology

Amino acid (AA) utilization and uptake

C. jejuni lacks phosphofructokinase, and as such cannot use glucose as a carbon source for growth. The identification of a fucose utilization pathway is described below; however, most strains of C. jejuni primarily use AAs as carbon and energy sources.¹⁵ Older work demonstrated that serine, aspartate, glutamate and proline are preferentially depleted from media and used for growth.¹⁶ However, recent work has shown that AA utilization and uptake also impact unexpected aspects of C. jejuni pathogenesis. For instance, some strains of C. jejuni were also found to utilize glutamine and asparagine, and the ability to utilize these AAs is dependent on periplasmic GGT and AnsB enzymes which deaminate glutamine and asparagine to glutamate and aspartate, respectively.¹⁷ GGT is important for efficient intestinal colonization^17,18 while AnsB is required for dissemination into deeper tissues,¹⁷ implicating AA utilization in tissue tropism. Genes encoding GGT, AnsB and an AA chemotaxis receptor are now being used as metabolic markers to help characterize strains involved in C. jejuni outbreaks.¹⁹

It has also now been shown that C. jejuni can import the hydrophobic AAs leucine, isoleucine and valine (LIV) via an AA-ABC transport system.²⁰ While import per se was not required for chicken colonization, mutants deleted for LIV periplasmic binding proteins were colonization-defective, suggesting more complicated roles for these proteins than simply LIV import. Another AA-ABC transporter involved in glutamine and glutamate uptake was likewise shown to play unexpected roles in pathogenesis, with mutants exhibiting diminished toxicity to host cells during cell infection.²¹ AAs are abundant in the host gastrointestinal (GI) tract and are also involved in GI homeostasis. Together with these new findings, this suggests that the differential ability of C. jejuni to import AAs may impact host responses to C. jejuni infection even beyond their utilization as carbon sources.

Electron transport chain (ETC) dynamics

For an organism with a relatively small genome, C. jejuni has a complex, highly-branched respiratory chain. Previous work demonstrated that C. jejuni could utilize electron donors such as formate, hydrogen, lactate, succinate, gluconate and 2-oxoglutarate, and that electrons were passed through menaquinone (MQ) en route to one of several terminal electron acceptors including oxygen, fumarate, nitrate, nitrite, TMAO and DMSO.¹⁵ Significant recent advances have been made in identifying and characterizing new electron donor oxidoreductase enzymes, exploring mechanisms of localization and regulation of key proteins involved in these pathways, clarifying the MQ biosynthetic pathway and the nature of enzymes required for transfer of electrons to terminal acceptors, and elucidating roles for some of these genes and pathways in pathogenesis.

It was known that C. jejuni can respire using lactate and also use lactate as a carbon source.²² Recent work has now identified an L-lactate-specific transporter as well as two novel lactate dehydrogenase (LDH) enzymes in C. jejuni.²³ It is unusual for one organism to harbor two LDH enzymes, which is consistent with C. jejuni being well-suited to use multiple electron donors as it traverses distinct niches, and particularly the GI tract which houses numerous commensal species excreting lactate. Another unusual finding, although one hinted at by older work looking at electron donor utilization, was the identification of a sulfite:cytochrome c oxidoreductase (SOR) allowing C. jejuni to respire using sulfite.²⁴ These systems are typically found in chemolithotrophs, which derive energy from oxidation of inorganic compounds. This was the first identification of a sulfite respiration system in a chemoheterotrophic pathogenic bacterium such as C. jejuni, which derives energy from organic compounds. Kelly and colleagues hypothesized that the SOR system might contribute to C. jejuni survival under low-oxygen conditions and/or in food processed with sulfite as a preservative.²⁴ A subsequent study showed that a mutant deleted for sorA exhibited motility and cell invasion defects, indicating a role in pathogenesis.²⁵ This is further supported by recent findings suggesting that growth of C. jejuni following host cell infection is enhanced when sulfite is present in the recovery medium (Pryjma, Gaynor, et al., submitted). Although an in vivo requirement for LDH has not yet been identified, it has now also been shown that a C. jejuni mutant defective for formate utilization was modestly attenuated for chicken colonization, with the inability to utilize 2-oxoglutarate or both hydrogen and formate in combination yielding severe defects.²⁶ An inability to chemotax toward formate likewise caused motility and invasion defects,²⁷ and mutants defective for both formate chemotaxis and formate utilization were very recently shown be diminished for immunopathology in a new gnotobiotic mouse model harboring “humanized” gut normal flora.²⁸ While new animal models of virulence are not a focal point of this review, development of models such as this together with increased understanding of C. jejuni biology are important advances likely to yield valuable new information in the future.

Progress has also been made in understanding modulation and localization of ETC proteins. Many enzymes involved in the ETC utilize either molybdenum or tungsten as cofactors. A single protein ModE was found to regulate transporter systems specific to each metal (MOD or TUP), repressing mod genes in the presence of either metal, and tup genes only in the presence of tungsten.²⁹ Additional work showed that formate dehydrogenase (FDH), the only known tungsto-enzyme in C. jejuni, requires another protein, FdhM, for maturation, possibly via an impact on tungsten cofactor incorporation.^30,31 FdhM and a number of other ETC enzymes were also found to be transported to the periplasm via the twin-arginine translocase (TAT) pathway: a functional TAT pathway was required for C. jejuni to utilize formate, hydrogen, sulfite or gluconate as electron donors and TMAO or nitrate as alternative electron acceptors.³⁰ A two-gene operon designated FdhTU was also recently found to be required for fdh gene expression and FDH activity (Pryjma M et al., submitted; Shaw et al., submitted, personal communication) and recovery of C. jejuni following host cell infection (Pryjma M et al., submitted).

Recent insight into later events in the ETC has also been achieved. Genome mining suggested that MQ biosynthesis in C. jejuni might utilize an alternative futalosine pathway recently identified in Streptomyces. Genetic and biochemical analyses demonstrated not only that the MQ biosynthetic pathway is essential in C. jejuni, but also that C. jejuni synthesizes a novel, adenine-modified futalosine as an MQ biosynthetic intermediate, and that an enzyme also involved in autoinducer biosynthesis (MTAN) converts this compound into the next pathway intermediate.³² Fumarate and succinate are important tricarboxcylic acid (TCA) cycle intermediates and also serve as ETC electron acceptor and donor, respectively. C. jejuni harbors genes predicted to encode both succinate dehydrogenase (sdh) and fumarate reductase (frd). Recent work has shown that the frdABC operon encodes a bifunctional complex that both reduces fumarate and oxidizes succinate, helping to correct a mis-annotation and providing further insight into ETC function.³³ An additional reductase, encoded by mfrABE (originally annotated as sdhABE), was also shown to reduce not only fumarate, but also mesaconate and crotonate, both of which are fermentation products of gut commensal bacteria.³⁴ A frdA mutant was significantly defective for chicken colonization (compared with an sdhA mutant which was not),³³ and AspA and AspB required for fumarate production were likewise found important for host cell adherence, intracellular viability and colonization in a MyD88-deficient mouse model.³⁵ Two newly identified likely terminal cytochrome c peroxidases were also shown to be required for optimal chicken colonization, further defining roles for the ETC in host-related attributes.³⁶

Other physiology-related developments

Also defining roles for C. jejuni physiology in biology and pathogenesis are two studies exploring essential gene identification,^37,38 one of which also provided a predicted metabolic model which may identify new antimicrobial targets.³⁷ Many predicted essential genes with annotated functions are involved in metabolism. Consistent with this was the recent identification of an essential but alternative spermidine biosynthetic pathway in C. jejuni.³⁹ This pathway was also identified in several other pathogens and resident commensal organisms, indicating conserved means of synthesizing this important cellular metabolite. A non-essential zinc transport system (ZnuABC) has also been identified and shown to be critical for chicken colonization, suggesting in vivo roles for one or more enzyme(s) using zinc as a cofactor.⁴⁰ Another study implicated metabolic shifts, and especially that to an anaerobic lifestyle, as important in intracellular survival and for C. jejuni existence in a newly-defined “campylobacter-containing vacuole (CCV).”⁴¹ Finally, a newly-identified peptidoglycan (PG) stem peptide-modifying enzyme, Pgp1, was recently shown to be required for maintenance of C. jejuni’s helical shape, with a pgp1 mutant displaying a completely straight morphology (Frirdich E et al., in press). The pgp1 mutant was also defective for chick colonization and elicited enhanced activation of the host proinflammatory molecules, suggesting roles for PG and Pgp1 in pathogen-host interactions.

Stress Responses

Temperature stress and the heat-shock response

As noted, the optimal growth temperature for C. jejuni is 37–42°C. When exposed to higher temperatures and other stresses that induce accumulation of misfolded proteins, C. jejuni elicits a classic heat shock response in which heat-shock proteins (HSPs) are induced to promote proper protein folding and degrade misfolded polypeptides. Recent work on the periplasmic HSP HtrA, which has both protease and chaperone activities, showed that protease activity was dispensable for growth at high temperatures but was necessary to prevent induction of the cytoplasmic heat shock response under normal conditions.⁴² However, HtrA chaperone activity was required for growth under high temperature stress, and also for adherence to both epithelial cells and macrophages.^42,43 Analysis of the two C. jejuni heat shock response negative regulators HspR and HrcA showed that HspR negatively regulates groES and groEL while HspR negatively regulates dnaK-hrcA, cbpA-hspR and clpB.⁴⁴ Penn and colleagues proposed an attractive model to explain precise timing during and following heat shock: each operon is rapidly derepressed upon heat shock, and the ensuing enhanced production of HrcA rapidly downregulates groESL when the heat shock is over, thereby restoring homeostasis. The RacRS two-component signal transduction system⁴⁵ was also recently found to both directly and indirectly modulate expression of heat-shock genes, and to be required for growth during temperatures above 42°C and conditions of high osmolarity (Apel D et al., in press).

Although neither 42°C nor 37°C are “stressful,” C. jejuni must be able to adapt to both temperatures as it moves between commensal chicken and susceptible human hosts. The lipooligosaccharide (LOS), a surface polysaccharide also discussed below, was recently found to occur in different amounts and in different forms at 37°C compared with 42°C,⁴⁶ indicating host-related temperature-dependent variation in a key surface structure. Chemotaxis toward certain AAs was also stronger at the human temperature of 37°C.⁴⁷ Another periplasmic protein with chaperone (SurA)-like and peptidyl-prolyl cis/trans isomerase domains was upregulated at 37°C rather than 42°C and was termed Cj0596/Peb4.^48,49 One group showed that a mutant deleted for this gene caused hyper-motility, enhanced invasion, altered outer-membrane protein profiles, and a mouse colonization defect.^49,50 Another group using a different C. jejuni strain instead found a peb4 mutant to be defective for cell adhesion and motility.⁵¹ This may reflect strain-strain differences but nonetheless show a role for Peb4 in pathogenic properties.

C. jejuni must also survive lower-than-optimal temperatures in the environment and on/in refrigerated food and water. While much less is known about cold stress responses than heat stress responses, one recent study found that exposure to 6°C for 24 h reduced ETC activity compared with 37°C but, unlike E. coli and Salmonella spp, did not enhance sensitivity to heat stress.⁵² Another group, which previously showed enhanced survival of C. jejuni in 5°C chicken juice compared with laboratory medium,⁵³ has now identified genes differentially expressed in these two conditions.⁵⁴ This includes upregulation of luxS in 5°C chicken juice, potentially implicating quorum sensing in C. jejuni cold survival. A polynucleotide phosphorylase has also been shown to be required for prolonged (up to 23 d) survival at both 10°C and 4°C.⁵⁵

Oxidative stress

C. jejuni lacks SoxRS and OxyR, regulators which modulate oxidative stress responses in many Gram-negative pathogens, but does harbor a peroxide-sensing PerR regulator. The PerR regulon was recently identified and compared with general oxidative stress stimulons, revealing both overlapping and unique profiles.⁵⁶ That study also led to mutagenic analyses of several oxidative stress genes (sodB, ahpC, katA) and an uncharacterized gene cj1386, identifying roles for each in animal (chicken or pig) colonization and a role for Cj1386 in heme trafficking to catalase.^56,57 PerR was also shown to autoregulate by binding to its own promoter in an iron-dependent manner.⁵⁸

Other C. jejuni factors recently identified as important for countering oxidative stress include the HtrA protease (described above)⁴³ and CmeG, a putative efflux transporter also required for antibiotic resistance.⁵⁹ Simultaneous disruption of two dsb oxidoreductase genes significantly diminished chicken colonization and abrogated host cell invasion and/or intracellular survival,⁶⁰ implicating disulfide bond formation and the oxidative state of extracytoplasmic proteins as critical for these pathogenic properties. A large-scale screen for new invasion and/or intracellular survival factors also identified a sodB mutant as defective for both animal colonization and host cell adherence and intracellular survival,³⁵ as well as a virK mutant as defective for antimicrobial peptide resistance and intracellular survival following invasion.⁶¹ Another newly identified oxidase, RdxA, conferred metronidazole resistance but was dispensable for chicken colonization, demonstrating the utility of rdxA as a selectable locus for introduction of complementing clones.⁶²

From the host side, a eukaryotic cell factor that releases carbon monoxide (CORM-3) was shown to inhibit C. jejuni formate-dependent respiration and induce production of H₂O₂, but unlike its effect on other bacteria, did not inhibit microaerobic growth of C. jejuni.⁶³ Interestingly, although exposing C. jejuni to nutrient limitation, heat shock and prolonged atmospheric oxygen conditions had deleterious effects on adherence, invasion, and/or intracellular survival, brief exposure of C. jejuni to oxygen enhanced invasion and intracellular survival.⁶⁴ Consistent with this, another study found that pre-incubation of C. jejuni with paraquat enhanced host cell adherence.⁶⁵ It is interesting to consider these findings regarding the pathogen-host cell dynamic in the context of C. jejuni’s fastidious growth requirements and relatively sparse contingent of oxidative stress response regulators.

Biofilms

Biofilm formation has been implicated as an important C. jejuni survival mechanism, particularly in the ex vivo environment where conditions are sub-optimal for growth. A number of recent studies suggest that biofilms form when C. jejuni are stressed, whether from an environmental condition or in mutant strains deleted for stress survival genes. Aerobiosis, for instance, was shown to enhance biofilm formation, whereas viable organisms were released back into a planktonic state independent of oxygen concentration.⁶⁶ C. jejuni mutants defective for general stress response factors such as the stringent response and polyphosphate likewise formed enhanced biofilms,^67,68 as did mutants with severely truncated LOS chains.⁶⁹ Analysis of a mutant deleted for the CprS two-component system sensor kinase, which modulates essential functions and survival of certain stresses, also showed enhanced biofilms and suggested that DNA is a likely component of the biofilm matrix.⁷⁰ Flagellar mutants, on the other hand, are biofilm-defective,^71,72 with other strains such as the straight pgp1 mutant exhibiting intermediate defects (Frirdich E et al., in press). Clarifying a role for biofilm formation in vivo has been confounded by the fact that each of the above mutants, whether hyper- or hypo-biofilm, also exhibit chicken or mouse colonization defects, likely because they are also defective for stress survival.^67,69,71,72 Recently, however, microcolonies and biofilms were shown to form on in vitro-infected human intestinal tissue,⁷³ suggesting in vivo relevance for the biofilm community lifestyle. Moreover, demonstration that bacteriophages can disperse C. jejuni biofilms⁷⁴ both lend insight into this understudied aspect of the biofilm-planktonic dynamic and provides a potential new strategy for reducing the environmental burden of C. jejuni.

General regulators

As alluded to above, many regulatory systems in C. jejuni are pleiotropic, both modulating basic processes and impacting aspects of pathogenesis. This is perhaps not surprising given the relatively small C. jejuni genome, which also contains a disproportionately low number of canonical regulators such as sigma factors (only 3 in total) and two-component signal transduction system genes (less than 1% of the genome compared with ~2% for Salmonella spp). Four general transcriptional regulators affecting multiple aspects of C. jejuni stress survival were either recently identified or shown to have additional functions than previously described. Expanded characterization of the bile-responsive TetR-family CmeR regulator showed that in addition to repressing transcription of an efflux pump required for bile stress survival,⁷⁵ CmeR also activates expression of a lipoprotein-encoding operon, at least one gene of which is important for host cell adherence and chicken colonization.⁷⁶ Crystal structures of CmeR also revealed that the regulator binds two similar bile acids in distinct manners,⁷⁷ lending insight into potential means by which the same regulator controls different bacterial processes. This theme was extended by demonstrating that the non-steroidal anti-inflammatory salicylate inhibits binding of CmeR to target DNA, enhancing expression of the cmeABC efflux pump, decreasing susceptibility to antibiotics, and potentially providing a mechanism for enhanced emergence of fluoroquinolone-resistant strains.⁷⁸ The flagellar regulator RpoN was also recently implicated in survival of acid stress and susceptibility to H₂O₂,⁷⁹ although mechanisms by which RpoN modulates both flagellar biosynthesis and stress survival were not proposed. A new essential regulator, CosR, was also shown via overexpression and a new antisense knock-down strategy to both positively (ahpC) and negatively (sodB, dps, luxS) regulate oxidative stress genes, consistent with the antisense knock-down strain displaying enhanced resistance to oxidative stress.⁸⁰ Finally, a new MarR-family regulator, Cj1556, was identified and shown to be important for oxidative and aerobic stress tolerance, intracellular survival in epithelial cells and macrophages, biofilms and virulence in an insect infection model.⁸¹ As with the gnotobiotic “humanized” mice described above, this invertebrate model may also yield new avenues for testing roles of specific genes in virulence in vivo.

Glycobiology

Metabolic diversity among C. jejuni strains: Relevance of fucose metabolism

There is an increasing body of evidence demonstrating that complex glycans, either present in the human diet, or on cell and mucosal surfaces of the intestine and the associated microflora, are involved in proper gut development and pathogen interactions.^82,83 Fucosylated structures are predominant in intestinal mucin,⁸⁴ milk oligosaccharides⁸⁵ and on the capsular polysaccharides and glycoproteins of gut flora such as Bacteroides species.^86,87 In the case of B. thetaiotaomicron, L-fucose induces the fucose utilization pathway which enables the organism to metabolize fucose⁸⁸ and also to incorporate the sugar into its surface glycoconjugates.⁸⁹ In contrast, when fucose is limiting, the fucose sensor of B. thetaiotaomicron is capable of signaling the host to upregulate fucosylated glycan expression on intestinal epithelial cells.⁸⁸ Another common gut commensal, Bifidobacterium longum subsp Infantis,⁹⁰ the predominant organism found in the intestine of breast-fed infants, possesses an extensive repertoire of enzymes targeting milk oligosaccharide utilization, including fucosidases, which allow foraging of dietary and host oligosaccharides.⁹¹ And Helicobacter pylori, a close relative of C. jejuni, is capable of inducing human α-L-fucosidase 2 which releases host L-fucose for H. pylori to incorporate into its Lewis antigen-containing lipopolysaccharides (to mimic the host) and to ensure proper binding to the host cell through these structures.⁹² Thus, it is not completely surprising that the common gut pathogen C. jejuni also has a propensity toward fucose displayed through chemotaxis⁹³ and binding to glycan structures containing this sugar.⁹⁴ Newburg and colleagues demonstrated that human milk oligosaccharides were able to inhibit C. jejuni colonization of mice in vivo.⁹⁵ They concluded that milk fucosyl-oligosaccharides and specifically α-(1→2)-fucosylated carbohydrate moieties containing the H(O) blood group epitope inhibit this binding and may represent a novel class of antimicrobial agents against C. jejuni.

As noted, C. jejuni lacks many of the pathways for nutrient metabolism used by other pathogens and the normal intestinal microbiota. For example, in addition to phosphofructokinase, several key enzymes within the glycolytic pathway, the Entner–Doudoroff pathway, and the pentose phosphate pathway are absent. As such, it was believed that C. jejuni was asaccharolytic, unable to use any form of carbohydrate as a substrate for growth and instead relying on the use of AAs and TCA cycle intermediates as carbon sources. However, work by Stahl et al. and Muraoka and Zhang recently showed that several strains of C. jejuni possess a gene operon (cj0481–cj0490 in strain 11168) that is upregulated in the presence of L-fucose (and mucin) and allows for the utilization of L-fucose for growth.^96,97 The metabolic pathway for fucose utilization in C. jejuni is unusual and does not mirror those described for other bacteria such as E. coli or Bacteroides where either L-fucose phosphates (indicating fuculose kinase activity) or GDP-modified intermediates (indicating pyrophosphorylase activity) are used. Instead, enzymes encoded by the cj0481–cj0490 operon in C. jejuni 11168 resemble those first described in Xanthomonas that are involved in converting L-fucose first to fuconate and then to pyruvate and lactate, consistent with the use of the TCA cycle. Confirmation of the metabolic activities of the enzymes remains to be determined. However, C. jejuni mutants in the putative fucose permease (fucP, cj0486) are deficient in fucose uptake and demonstrate a competitive disadvantage when colonizing the piglet model of human disease, while the same trend is not observed in the colonization of poultry. This identifies a previously unrecorded metabolic pathway in select strains of C. jejuni that is associated with a virulent lifestyle—and suggests that fucosylated milk oligosaccharides and fucose rich diets may actually exacerbate C. jejuni-induced diarrheal disease in certain cases. Indeed, when fucose was fed to chickens, the C. jejuni wildtype strain displayed a competitive colonization advantage compared with the mutant deficient in fucose uptake.⁹⁷ Interestingly, structural studies of chicken mucins showed that the glycans contain significant amounts of sulfate modifications which might block endogenous fucosidases from releasing fucose.⁹⁷ Consistent with lack of fucose accessibility, other groups have been unable to detect fucose from chicken intestinal mucins using the fucose lectin from Ulex europaeus, UEA.^98,99 It will be interesting to determine if these observations have any link to studies published by Bourke’s group^100,101 demonstrating that chicken mucous, but not human mucous, inhibited C. jejuni invasion of intestinal epithelial cells.¹⁰¹ Furthermore, periodate oxidation of purified chicken mucins reversed the inhibitory affect on C. jejuni invasion suggesting that the inhibitory factor is carbohydrate-related.¹⁰⁰

Influence of C. jejuni glycoconjugates on host innate and adaptive immune responses

Although C. jejuni has a limited ability to metabolize sugars, the organism possesses a large number of enzymes involved in the biosynthesis of carbohydrates which are then incorporated into its peptidoglycan (PG), lipooligosaccharides (LOS), capsular polysaccharides (CPS) and both N- and O-linked glycoproteins. Therefore, C. jejuni has become an excellent model system for bacterial glycomics studies. Recent work demonstrating the importance of C. jejuni PG changes and the influence on cell shape and host-related properties was discussed above, so PG will not be mentioned further in this section. There have also been several reviews on C. jejuni LOS, CPS and protein glycosylation pathways which are recommended reading to obtain a more comprehensive background on these glycoconjugates.^102-105 This section will focus on how these carbohydrate structures influence the host immune system.

Since C. jejuni infections are typically cleared within a week after infection, the host innate immune system, which includes cationic antimicrobial peptides, plays a key role in the control of this prominent pathogen. Antigen presenting cells such as macrophages and dendritic cells (DCs) also play an important role in the innate and adaptive immune responses during bacterial infection. Both macrophages and DCs express several pattern recognition factors such as Toll-like receptors (TLR) and carbohydrate recognizing C-type lectins. It has been demonstrated that live and heat-killed C. jejuni are capable of inducing DC maturation.^106,107 This results in the activation of NFκB and the cytokines: interleukin- (IL-)1beta, IL-6, IL-8, IL-10, IL-12, gamma interferon and tumor necrosis factor α.¹⁰⁶ Interestingly, purified C. jejuni LOS appears to play a major role in the increased production of cytokines by DCs and in the initiation of a Th1 (pro-inflammatory) adaptive immune response.¹⁰⁶ Rathinam et al. also observed similar cytokine profiles and Th1 responses during C. jejuni infection.¹⁰⁷

There have been numerous studies demonstrating that many isolates of C. jejuni are capable of expressing LOS that mimic human ganglioside structures. It is believed that individuals demonstrating a breakdown in self-tolerance will develop Guillain-Barre´ Syndrome (GBS) through the generation of cross-reactive ganglioside antibodies.¹⁰⁸ Although GBS can result after infection by several agents, C. jejuni has become the most common antecedent to the development of this peripheral neuropathy. Since C. jejuni LOS sialylation is the single known factor associated with the development of GBS, several studies have been directed at determining whether sialic-acid binding immunoglobulin-like lectins (Siglecs) that are present on the surfaces of many immune-modulating cell types, are capable of recognizing C. jejuni LOS. Of the 10 known human Siglecs, Siglec-7 showed specific binding to sialylated LOS among the strains examined.¹⁰⁹ Using another strain set, it was demonstrated that GBS-associated C. jejuni strains bind to sialoadhesin (Sn, Siglec-1).¹¹⁰ Since Siglecs are also expressed on dendritic cells in high levels, Bax et al. examined whether C. jejuni LOS could induce DC maturation.¹¹¹ The authors found that α2,8-linked sialylated LOS bound to Siglec-7 and resulted in a Th1 response, while α2,3-linked sialylated LOS bound to Sn and induced a Th2 (anti-inflammatory) response. The authors demonstrated that the sialic acid composition of C. jejuni LOS could induce different Th responses potentially through targeting of different DC-expressed Siglecs. LOS induction was shown in earlier studies to be TLR4-dependent.¹¹²

Interestingly, it has been shown that some C. jejuni isolates are capable of producing a broader range of LOS structures that mimic other human glycoconjugates such as P blood group antigens (P1 and P^k), paraglobosides (LNnT), lacto-N-biose (LNB, common oligosaccharide found in human milk) and sialyl-Lewis c units.¹¹³ The biological consequences of infection with C. jejuni strains expressing these novel LOS mimics remains to be determined.

Similar to other Gram-negative bacteria, C. jejuni LOS also protects the organism from the action of cationic antimicrobial peptides,¹¹⁴ and this in part is due to the phosphoethanolamine (PEtn) modification of the LOS lipid A.¹¹⁵ Studies by Cullen and Trent demonstrated that mutation of the C. jejuni PEtn transferase resulted in 20-fold greater sensitivity to polymyxin B and also reduced motility which was caused by the loss of the PEtn modification on the flagellar rod protein, FlgG.¹¹⁵ The significance of this finding is that an enzyme previously known to modify carbohydrates, is also able to recognize proteins, and influence the phenotypes of both biomolecules. Another mechanism for C. jejuni resistance to antimicrobial peptides is through the addition of amide-linked acyl chains to the lipid A which also reduces TLR4 activation.¹¹⁶

C. jejuni is capable of expressing a diverse range of CPS structures which are important for serum resistance and divide the species into 47 different serotypes.^114,117 Although the use of C. jejuni CPS as part of a glycoconjugate vaccine may appear challenging when considering the variety of structures the isolates can produce, many CPS structures have common elements such as complex heptose derivatives and O-methyl phosphoramidate (MeOPN) modifications. The CPS from C. jejuni strains 81–176 and CG8486 were chemically conjugated to the carrier protein CRM197, a mutated diphtheria toxin.¹¹⁸ Vaccination resulted in strong immune responses and significantly reduced the disease following challenge with the homologous strains in mice. The 81–176 CPS-CRM197 vaccine was then tested in the New World monkey model and showed 100% protection against diarrhea following challenge with the homologous strain.¹¹⁸ In contrast, Rose et al. recently demonstrated that the CPS of C. jejuni 11168 attenuates cytokine production by dendritic cells which was not exclusively dependent on TLR4 signaling.¹¹⁹ This effect was also observed with MeOPN modification of CPS since mutants lacking this residue showed increased cytokine production compared with the parent.

Although C. jejuni isolates show very diverse LOS and CPS structures, it is believed that all strains of this species post-translationally modify at least 60 periplasmic and membrane proteins with a conserved N-linked heptasaccharide terminating in N-acetylgalactosamine (GalNAc).¹²⁰ Recently, van Sorge et al. showed that the heptasaccharides derived from the N-glycosylation pathway are recognized by the human macrophage galactose-type lectin (MGL),¹²¹ a C-type lectin commonly found on immature dendritic cells with specificity for terminal GalNAc residues. Similarly, C. jejuni isolates with LOS containing terminal GalNAc residues were also recognized by MGL. Interestingly, IL-6 production by the DCs was enhanced in N-glycosylation mutants compared with the parent suggesting that protein modification with this conserved heptasaccharide allows C. jejuni to limit the immune response.

C. jejuni also expresses an O-linked protein glycosylation system which exclusively modifies flagellin proteins with nonulosonic acid sugars related to sialic acid. Most bacterial flagellins are bound by TLR5, but C. jejuni evades TLR5 recognition. Therefore, de Zoete et al. tested whether the O-glycans on C. jejuni flagellin were masking the flagellin recognition sites. The authors demonstrated that the O-glycans did not play a role in TLR5 evasion, but instead identified a novel β-hairpin structure on the protein that was required for activation of human and mouse TLR5, but not for chicken TLR5.¹²² The ability of 4 C. jejuni isolates to activate human TLR1/2/6, TLR4, TLR5 and TLR9 and chicken TLR2t2/16, TLR4, TLR5 and TLR21 was also examined.¹²³ Live bacteria showed low levels of TLR activation, but lysed bacteria induced strong NFκB activation through human TLR1/2/6 and TLR4 and chicken TLR2t2/16 and TLR4, but not through TLR5 of either species. Furthermore, C. jejuni induced TLR4-mediated β interferon in human, but not chicken cells.¹²³ Broader studies comparing the innate and adaptive immune responses against C. jejuni in chickens and human models may provide some clues to how this ubiquitous organism is capable of both commensal and pathogenic lifestyles.

Concluding Perspectives

Rapid advances in our understanding of C. jejuni make this an exciting time to be studying this important human pathogen. A comprehensive picture of C. jejuni cytoplasmic, periplasmic and cell envelope biology is beginning to emerge, as are bacterial attributes contributing to its success as a pathogen. Together with advances in animal models and new technologies not detailed here, the next few years should see even greater strides in our understanding of how this organism traverses so many distinct niches and ideally will lead to the development of new antimicrobial strategies targeted for use in commensal animal and susceptible human hosts.