Campylobacter is a notoriously difficult organism to culture and maintain in the laboratory (Solomon and Hoover, 1999) and as a result there have been many methods developed, and method modifications proposed, for its detection in foods. As with the development of methods for other pathogens, media originally used for the isolation of the organism from faeces were used. Subsequent modifications have been required to enable the detection of low numbers of sublethally injured cells in the presence of higher numbers of competitor organisms and this has led to methods based on liquid enrichment prior to selective agar plating with colony identification. The history of the development of selective media for isolation of campylobacters, including the rationale for the choice of selective agents has been described by Corry et al. (1995). Most of the media include ingredients intended to protect campylobacters from the toxic effect of oxygen derivatives. Most commonly used are lysed or defibrinated blood; charcoal; a combination of ferrous sulphate, sodium metabisulphite and sodium pyruvate; and haemin or haematin.
A number of approaches have been taken to avoid the inhibitory effects of the toxic components in the media on sublethally injured cells including a preliminary period of incubation at reduced temperature and a delay in the addition of antibiotics. A 4h pre-enrichment time at 37 °C was found to be optimal for allowing the recovery of low levels of C. jejuni while preventing competitive inhibition due to the outgrowth of the accompanying flora (Uyttendaele and Debevere, 1996). A delay of 4-8h before adding antibiotics to broth was found to significantly increase the Campylobacter isolation rate from naturally contaminated river water compared with direct culture in selective broth. However, with chicken samples, significantly better results were obtained with selective broth as the primary medium (Mason et al., 1999). These findings probably reflect the varying degrees of sublethal injury of the Campylobacter cells in the different environments.
A number of enrichment broths for Campylobacter have been developed. In a recent study three of these broths - Bolton broth (BB), Campylobacter enrichment broth (CEB) and Preston broth (PB) - were compared for the isolation of Campylobacter from foods (Baylis et al., 2000). Both BB and CEB were better than PB for the isolation of Campylobacter from naturally contaminated foods, although BB yielded more confirmed Campylobacter growth than CEB. The use of selective enrichment broths supplemented with the enzyme Oxyrase®, a membrane-bound enzyme derived from E. coli, has been used and shown to be effective in order to provide an oxygen-reduced atmosphere for optimal Campy-lobacter recovery (Abeyta et al., 1997). In contrast, a blood-free enrichment broth (BFEB) has been investigated under aerobic conditions and compared favourably with the US Food and Drug Administration's Bacteriological Analytical Manual (BAM) method for the recovery of C. jejuni from inoculated foods. The BFEB method had the advantage of not requiring the use of blood, Oxyrase® or special equipment (Tran, 1998).
Several agars have also developed for the isolation of Campylobacter. Three of the most commonly used agars are Butzler agar, charcoal cefoperazone des-oxycholate agar (CCDA) and Preston agar and in a comparison for the detection of campylobacters in manually shelled egg samples and raw meat samples no substantial difference was seen (Zanetti et al., 1996) although CCDA was preferred. In this study the variable that had the most influence was the incubation temperature with a higher number of strains isolated at 42 °C than at 37 °C. Three isolation media (Karmali, Butzler and Skirrow agar) were compared for the detection of Campylobacter in 1500 samples of oysters, ready-to-eat vegetables, poultry products and raw milk cheeses following enrichment in Preston or Park and Sanders broths (Federighi et al., 1999). Park and Sanders enrichment and isolation on Karmali agar appeared to be the most efficient combination.
Cefoperazone amphotericin teicoplanin (CAT) agar was developed from charcoal cefoperazone deoxycholate (mCCD) agar by modification of the selective antibiotics in order to permit growth of strains of Campylobacter upsaliensis (Corry and Atabay, 1997). CAT agar supported the growth of a wider variety of Campylobacter species than mCCD agar, which was attributed to the level of cef-operazone in mCCD agar being inhibitory to some Campylobacter strains.
There is an international standard method for the detection of Campylobacter in food and animal feeding stuffs (Anon., 1995). This method involves enrichment in either Preston broth or Park and Sanders broth. If Preston broth is used, incubation is at 42 °C in a micro-aerophilic atmosphere for 18h. If Park and Sanders broth is used the initial suspension is incubated in a micro-aerophilic atmosphere at 32 °C for 4h, then antibiotic solution is added and incubated at 37 °C for 2h followed by transfer to 42°C for 40-42h. The enrichment broth cultures are streaked out onto Karmali agar and a second selective agar from the following: modified Butzler agar, Skirrow agar, CCDA and Preston agar. Agar plates are incubated at 42 °C in a micro-aerophilic atmosphere for up to 5 days prior to a series of confirmatory tests on characteristic colonies.
In recent years, numerous rapid methods have been developed for the detection of Campylobacter, some of which have been evaluated for application with foods. However, very few of these methods have been commercialised, which probably reflects the fact that the food industry is not doing a lot of testing for Campylobacter and the market for rapid method test kits is therefore small. The reason for the lack of testing by the food industry is probably due to a number of factors including: the organism does not grow in food under most normal storage conditions; it does not survive well and is relatively easily controlled in processed foods; it is prevalent in raw foods where the ultimate critical control is in the hands of the consumer; and the organism is fastidious and its detection and maintenance in the laboratory are not easy.
Latex agglutination tests for Campylobacter have been available for several years (Wilma et al., 1992). These tests enable convenient and easy visualisation of serological agglutination of Campylobacter and they are intended for confirmation of presumptive isolates. The concentration of cells needed for agglutination ranged from 106 to 108cfu/ml and similar sensitivities were seen with the non-culturable coccoid forms of Campylobacter. In an attempt to improve the speed of the Campylobacter detection method, latex tests have been applied to enrichment broth cultures (Wilma et al., 1992). The Microscreen® Campylobacter test has been used after incubation in CCD broth (42 °C, 8h), filtration (0.45 mm) and incubation in blood-free modified CCD broth (42°C, 16-40h) for the detection of Campylobacter in fresh and frozen raw meat (Baggerman and Koster, 1992). The Microscreen® Campylobacter latex kit has also been used for the testing of water samples following a physical enrichment (filtration and cen-trifugation) rather than a cultural enrichment (Baggerman and Koster, 1992).
More recently developed rapid methods for Campylobacter have been based on the polymerase chain reaction (PCR) technique (Nogva et al., 2000; O'Sulli-van et al., 2000). A magnetic immuno-PCR assay (MIPA) was developed for the detection of C. jejuni in milk and chicken products (Docherty et al., 1996). Target bacteria were captured from the food sample by magnetic particles coated with a specific antibody and the bound bacteria then lysed and subjected to PCR. The MIPA could detect 420 cfu/g of chicken after 18 h enrichment, 42 cfu/g after 24 h and 4.2cfu/g after 36h. For artificially contaminated milk, 63cfu/ml could be detected after 18 h and 6.3cfu/ml after 36 h. Immunocapture has also been combined with PCR for the detection of Campylobacter in foods and this enabled detection of the pathogen without an enrichment step and took about 8 h to perform (Waller and Ogata, 2000). The assays were quantitative and the limit of detection was 1 cell/ml and this was not affected when tested in spiked milk samples and chicken skin washes.
Many food samples and enrichment media, e.g. the presence of charcoal and iron (Thunberg et al., 2000), are inhibitory to the PCR, thereby lowering its detection capacity. Sample preparation methods using buoyant density centrifuga-tion (Uyttendaele et al., 1999; Wang et al., 1999) and heat treatment at 96 °C (Uyttendaele et al., 1999) have been used to overcome this inhibition. While OxyraseTM has been shown to significantly enhance the growth of C. jejuni, it appeared not to improve the PCR detection of C. jejuni in naturally contaminated chickens (Wang et al., 1999).
A commercial kit, the API Campy, is available for the differentiation of Campylobacter spp., although identification of species within the family Campy-lobacteriaceae using standard biochemical tests can be problematical because of the variability and atypical reactions of some strains (Phillips, 2001). The heat-stable serotyping system (the 'Penner' Scheme) has been used for Campylobacter. However, in a study in Denmark it was found that this system had limitations in that 16% of the stains were untypeable (Nielsen and Nielsen, 1999).
For epidemiological studies, molecular typing techniques have been preferred (Rautelin and Hanninen, 2000; Phillips, 2001). Pulsed-field gel electrophoresis was used for the first time in a Campylobacter outbreak in the USA, where a cafeteria worker was the probable source, and was critical in determining that community cases were not linked (Olsen et al., 2001). The use of PCR with a RAPD protocol has also proved a useful tool for the epidemiological analysis of Campylobacter (Hilton et al., 1997).
Was this article helpful?