Campylobacter jejuni

Authors


William John Snelling, School of Biomedical Sciences, University of Ulster, Coleraine, Co., Londonderry BT52 1SA, UK
(e-mail: b.snelling@ulster.ac.uk).

Summary

This review describes characteristics of the family Campylobacteraceae and traits of Campylobacter jejuni. The review then focuses on the worldwide problem of C. jejuni antimicrobial resistance and mechanisms of pathogenesis and virulence. Unravelling these areas will help with the development of new therapeutic agents and ultimately decrease illness caused by this important human pathogen.

Introduction

Campylobacters may have been discovered in 1886 by Theodor Esherich, from the colons of infants who had died of what he called ‘cholera infantum’ (Skirrow and Butzler 2000). However, it took until 1972 for Dekyser and Butzler to isolate Campylobacter from the blood and faeces of a previously healthy young woman with acute febrile haemorrhagic enteritis (Skirrow and Butzler 2000). The family Campylobacteraceae consists of Campylobacter, Arcobacter and Bacteroides ureolyticus and occur primarily as commensals in humans and domestic animals (Vandamme 2000). Members are typically motile with a characteristic corkscrew-like motion via a single polar unsheathed flagellum at one or both ends of their cells (Vandamme 2000). Members of the family Campylobacteraceae neither ferment nor oxidize carbohydrates, instead they obtain energy from amino acids, or tricarboxylic acid cycle intermediates and the only respiratory quinones that have been detected are menaquinones, with menaquinone-6 and menaquinone-5 being the major components (Vandamme 2000). Oxidase activity is present in all 14 Campylobacter spp. (Table 1), except Campylobacter gracilis (Fields and Swerdlow 1999; Vandamme 2000). Campylobacters are generally microaerophilic and may be cultured in atmospheres with 3–15% oxygen, supplemented with 2–10% CO2 (Forsythe 2000). The thermophilic campylobacters C. jejuni, Campylobacter coli, Campylobacter lardis and Campylobacter upsaliensis are more associated with human gastrointestinal disease, especially C. coli and C. jejuni ssp. jejuni which accounts for 95% of all clinical isolates in the UK (Thomas et al. 1999; Matsuda and Moore 2004). Each species of Campylobacter has a favoured reservoir, C. jejuni, the most common species associated with human illness is prominently associated with poultry and has evolved to preferentially colonize the chicken gut given its optimal growth conditions, e.g. 42°C (Vispo and Karasov 1997; Altekruse et al. 1998; Newell 2001; van Vliet and Ketley 2001). Once excreted into the environment, C. jejuni usually does not multiply, because of its relatively high minimal growth temperature (>30°C) (van de Giessen et al. 1996). Campylobacters, unlike Salmonella, are generally unable to multiply in foods and are normally not linked to large outbreaks of campylobacteriosis. Over 90% of cases are sporadic and occur mainly in the summer (Jones et al. 1991; Ketley 1997). Nonetheless campylobacters are responsible for the majority of intestinal infectious diseases world-wide, annually affecting 1·1% and 1% of populations in the UK and USA respectively (Jones et al. 1991; Ketley 1997; Pebody et al. 1997; Altekruse et al. 1998; On et al. 1998; Thomas et al. 1999; Hong et al. 2004). The true public incidence, due to under-reporting, is estimated to be up to 10 times higher than documented case numbers (Griffiths and Park 1990; Allos 2001). As a result of campylobacteriosis, substantial worldwide losses are accumulated annually because of clinical costs and lost working hours, e.g. $1·3–6·2 billion annually in the USA (Forsythe 2000).

Table 1.  Differential characteristics between Campylobacter species, Vandamme (2000)
SpeciesAlpha- haemolysisCatalaseHippurate hydrolysisUreaseNitrate reductionSelenite reductionH2S/TSI (trace amounts)Indoxyl acetate hydrolysisGrowthResistance to
25°C42°CMinimal mediumMacConkeyGlycine (1%)NaCl (4%)Cefoperazone (64 mg l−1)Nalidixic acidCephalothin
  1. +, Characteristic present in >90% of examined strains; −, characteristic present in <11% of examined strains; V, strain-dependent reaction; *, at least 80% of examined strains contained this characteristics.

Campylobacter coliV+++V+++V+++
C. concisusVVVVVV
C. curvusVV+VVVV*V*+V*+
C. fetus ssp. fetus++V*+V*VV+++
C. fetus ssp. venerealisVV*++V*VV
C. gracilisVV*VVVV*+V
C. helveticus++++VV
C. hyointestinalis ssp. hyointestinalisV++++V+VV++V
C. hyointestinalis ssp. lawsoniiV++++++VVVV+
C. jejuni ssp. doylei+V++V
C. jejuni ssp. jejuni++++V+++++
C. lariV+V+V+++V+
C. mucosalis++V*VVV*
C. rectus+V++V+V*
C. showae+++VVVV+V
C. sputorum+VV+V++VV+VV
C. upsaliensis++++V*+VV

Campylobacter jejuni

Within the species C. jejuni two subspecies, ssp. doylei and ssp. jejuni, can be distinguished on the basis of nitrate reduction and cephalothin susceptibility (Table 1) (Allos 2001). A weak catalase reaction may be observed for C. jejuni ssp. doylei (Table 1) (Vandamme 2000). The pathogenic role of C. jejuni ssp. doylei is unknown (Vandamme 2000) and subsequently in this review where C. jejuni is mentioned it refers to C. jejuni ssp. jejuni. Campylobacter jejuni are S-shaped rods (0·2- to 0·8-μm wide and 0·5- to 5·0-μm long), Gram-negative, nonsaccharolytic, nonspore forming, catalase, oxidase and hippurate hydrolysis positive (Griffiths and Park 1990; Vandamme 2000; Ottosson and Stenstrom 2003). They are motile with a characteristic corkscrew-like motion, via a polar unsheathed flagellum and occur as commensals in warm-blooded animals, especially poultry (Fields and Swerdlow 1999; Song et al. 2004). Campylobacter jejuni has a relatively small genome of c. 1·6–1·7 Mbp of adenine and thymine rich DNA; the guanine and cytosine ratio ranges from 29 to 47 mol% (Ketley 1997; Vandamme 2000). The small size of the genome is perhaps reflected in a requirement for complex growth media, no oxidation or fermentation of carbohydrates, no lipase or lecithinase activity and a lack of growth below pH 4·9 (Collins and Lyne 1985; Ketley 1997; Vandamme 2000). The hippuricase gene is only found in C. jejuni (Rautelin et al. 1999). However, some C. jejuni isolates are hippuricase-negative, making it impossible to differentiate C. coli from hippuricase-negative C. jejuni using purely biochemical tests (Fields and Swerdlow 1999).

C. jejuni has an infective dose of between 500 and 10 000 organisms and, after an incubation period of 1–7 days, the resulting campylobacteriosis in humans usually involves a self-limited gastrointestinal illness lasting up to 7 days (Pebody et al. 1997; Altekruse et al. 1998). Clinical manifestations of campylobacteriosis are extremely diverse, ranging from a complete absence of symptoms to fulminating sepsis and rarely death mainly in immunosusceptible hosts (Altekruse et al. 1998; Thomas et al. 1999; Hong et al. 2004). Campylobacter infections are also associated with postinfectious complications including arthritis, Reiter syndrome and Guillain–Barré syndrome (Altekruse et al. 1998; Thomas et al. 1999; Hong et al. 2004).

Antimicrobial resistance

Most Campylobacter infections are self-limiting, therefore the majority of patients require no more than supportive treatment, e.g. maintenance of hydration and electrolyte balance (Peterson 1994; Koenraad et al. 1997; Allos 2001). However, there are specific clinical manifestations where, when combined with patient circumstances, antibiotics are used, e.g. infection with HIV, or pregnant women with symptoms lasting more than 1 week (Allos 2001; Nachamkin et al. 2002). Many factors have contributed to the emergence of antimicrobial resistant Campylobacter (Altekruse et al. 1998). The increasing number of human infections by antimicrobial resistant strains of C. jejuni makes the clinical management of campylobacteriosis increasingly difficult (Fields and Swerdlow 1999). Resistance to antimicrobial drugs can prolong illness and compromise the treatment of patients with bacteraemia (Fields and Swerdlow 1999).

Worldwide in the past few years a rapidly increasing proportion of Campylobacter strains have been found to be fluoroquinolone (FQ) resistant (Allos 2001). Higher case numbers of antimicrobial-resistant campylobacters are in developing countries, where antibiotic usage in humans and animals is relatively unrestricted (Fields and Swerdlow 1999). Primary resistance to quinolone therapy in humans was first noted in the early 1990s in Asia and in European countries, e.g. Sweden, the Netherlands, Finland, Spain and the UK (Allos 2001). The resistance coincided with initiation of the administration of the FQ antibiotic enrofloxacin, to food animals in those countries, e.g. broiler chicken feed (Altekruse et al. 1998; Allos 2001). The majority of C. jejuni FQ-resistant isolates have a mutation in gyrA that results in the substitution Thr-86 by Ile (Griggs et al. 2005). Despite poultry companies halting the usage of FQs for treating flocks, FQ resistance persists in the commercial poultry environment in the absence of FQ-selective pressure (Price et al. 2005). Campylobacter jejuni has become increasingly resistant to FQ antimicrobials and the rapid emergence of FQ-resistant Campylobacter may be attributable partly to the enhanced fitness of the FQ-resistant isolates (Luo et al. 2005). When FQ-susceptible and FQ-resistant Campylobacter isolates are co-inoculated into chickens, the FQ-resistant strains outcompete the majority of the FQ-susceptible strains, indicating that the resistant Campylobacter are biologically fitter in the chickens (Price et al. 2005).

The newer macrolides, azithromycin and clarithromycin, are effective against C. jejuni infections, but they are more expensive than erythromycin and provide no clinical advantage (Allos 2001). High rates of resistance make tetracycline, amoxicillin, ampicillin, metronidazole and cephalosporins poor choices for the treatment of infections with C. jejuni (Allos 2001). Despite decades of usage, the rate of resistance of Campylobacter to erythromycin remains relatively low (Allos 2001), making erythromycin considered to be the optimal drug for treatment of Campylobacter infection (Koenraad et al. 1997; Allos 2001).

Virulence factors and pathogenesis

In association with food or water ingestion, campylobacters enter the host intestine via the stomach acid barrier and colonize the mucous blanket overlying the epithelium of the distal ileum and colon (Ketley 1997). When colonizing intestines, C. jejuni expresses several virulence factors (Fig. 1), e.g. chemotaxis (Takata et al. 1992; van Vliet and Ketley 2001). Heat shock proteins are associated with the thermal stress–response of bacteria and are important virulence factors, e.g. DnaJ mutants are unable to colonize chickens (Konkel et al. 1998; van Vliet and Ketley 2001). Campylobacter penetrate the mucous layer covering the intestinal cells using their polar flagella and ‘cork-screw’ motion (Szymanski et al. 1995). Mutants of flaA, the primary structural gene for flagella, are unable to colonize 3-day-old chicks and cannot invade human intestinal epithelial cells in vitro (Fields and Swerdlow 1999). Adhesion and invasion are dependent on both motility and flagellar expression, as C. jejuni mutants with decreased motility because of paralysed flagella show reduced adhesion and no invasion, indicating that while flagella are involved in adherence, other adhesins are involved in subsequent internalization (Yao et al. 1994; van Vliet and Ketley 2001).

Figure 1.

The different phases of Campylobacter invasion of the intestine (van Vliet and Ketley 2001). 1, motility; 2, chemotaxis; 3, oxidative stress defence; 4, adhesion; 5, invasion; 6, toxin production; 7, iron acquisition; 8, temperature stress response; 9, coccoid dormant stage. inline image, viable Campylobacter cell; inline image, coccoid dormant Campylobacter cell; inline image, epithelial cell

The outer-membrane adhesion proteins CadF and PEB1 are involved in adherence and invasion. The PEB1 protein is encoded by the peb1A locus, with significantly decreased adherence, invasion of HeLa cells and reduced colonization of mouse models associated with peb1A mutants, and cadF mutants are unable to bind fibronectin and colonize newly hatched leghorn chickens (Konkel et al. 1997; Pei et al. 1998; Ziprin et al. 1999; van Vliet and Ketley 2001). The major outer membrane constituents lipo-oligosaccharide and lipopolysaccharides are involved in serum resistance, endotoxicity and adhesion (Jin et al. 2001; van Vliet and Ketley 2001). The superoxide dimutase protein SodB is the main component of the C. jejuni superoxide stress defence and sodB mutants show significantly decreased intracellular survival in human embryonic intestinal (INT-407) cells (Pesci et al. 1994; Purdy and Park 1994). In host tissues bacteria complex iron into haem compounds, to transferrin in serum and lactoferrin at mucosal surfaces (van Vliet and Ketley 2001). The C. jejuni genome encodes a putative ferrous ion transport system, a homologue of the Escherichia coliFeoB protein (van Vliet and Ketley 2001; Raphael and Joens 2003). However, while Escherichia coli and Helicobacter pylorifeoB mutants were unable to colonize the intestine and stomach respectively in a mouse model, ferrous iron uptake could not be attributed to feoB in C. jejuni (van Vliet and Ketley 2001; Raphael and Joens 2003).

The internalization mechanism triggered by C. jejuni is associated with the combined effect of microfilaments (MF) and microtubules (MT) of host intestinal epithelium cells (Biswas et al. 2000; Kopecko et al. 2001; Biswas et al. 2003). Invasive bacterial pathogens generally interact with their host via biochemical cross-talk, stimulating signalling cascades in both the bacterium and the host that ultimately trigger rearrangements of the host cytoskeleton and cause internalization of the pathogen (Kopecko et al. 2001). Following uptake into a vacuole, internalized C. jejuni are incapable of escaping from their membrane-bound compartment, only one round of intracellular replication has been observed during the first 5 h postinfection of human epithelial cells, and intracellular survival within the host cell may be aided by the production of catalase to protect against oxidative stress from lysosomes (Forsythe 2000; Kopecko et al. 2001). Campylobacter jejuni translocates within vacuoles to the basolateral surface where exocytosis can occur, the organism may also enter a viable, but nonculturable state, which may be of importance in the organism's virulence (Forsythe 2000; Kopecko et al. 2001).

The exact mechanisms by which C. jejuni induces disease in humans remain unknown, symptoms could be a result of cytolethal-distending toxin (CDT)-induced host cell death and the ensuing inflammatory responses (Kopecko et al. 2001; Newell 2001). CDT triggers cell-cycle arrest and, ultimately, death and together with bacterial adherence or invasion, mediates interleukin-8 production and release from the epithelium (Kopecko et al. 2001).

Future prospects

Campylobacter is currently the leading cause of bacterial diarrhoea in the developed world and therefore presents a significant challenge to public health. Campylobacter jejuni has a ubiquitous nature and is mainly linked with sporadic cases of food poisoning. Disappointingly, despite the steady improving techniques in molecular biology, the epidemiology of many infections remains unclear. This confusing epidemiological evidence is partly because of the lack of standard global typing methods and communication between laboratories (Wassenaar and Newell 2000). However, these problems are being addressed via initiatives like CAMPYNET (Wassenaar and Newell 2000) and PulseNet for standard molecular typing. Despite the array of virulence factors identified, so far, it has not been possible to develop a vaccine. Thus, research could focus on ways of reducing Campylobacter in livestock and domestic animals and in generating new antimicrobial agents. Further pathogenicity and virulence studies are required immediately. This helps to provide new antimicrobial agents, e.g. antibiotics and bacteriocins. These agents then need to be administered and used responsibly, to decrease the human illness in the long term and to ultimately reduce the economic loss because of lost working hours and clinical testing costs.

Acknowledgements

WJS was supported by a CAST award from the Department of Higher and Further Education for Northern Ireland, O'Kane's Poultry and Moy Park. JEM was funded by the Research and Development Office, Department of Health, Northern Ireland [Infectious Disease – Recognised Research Group (RRG) 9.9]. Facilitation of the completion of this review between the Japanese and the UK authors has been made possible by a grant awarded by the Great Britain Sasakawa Foundation.

Ancillary