Persistence of Campylobacter jejuni on surfaces in a processing environment and on cutting boards

Authors


Mieke Uyttendaele, Laboratory of Food Microbiology and Food Preservation, Faculty of Agricultural and Applied Biological Sciences, Ghent University, Coupure links 653, B-9000 Gent, Belgium (e-mail: mieke.uyttendaele@ugent.be).

Abstract

Aims:  The objectives of the study were to determine the spread and persistence of Campylobacter in a poultry processing plant and to provide a quantitative estimate of the survival of Campylobacter jejuni on the surface of a cutting board.

Methods and Results:  Several contact surfaces in a poultry processing plant were sampled before the start of processing, after 30 min and after 120 min. Next, the survival of four C. jejuni strains was studied on a beech and polypropylene cutting board during 120 min.

Conclusions:  A rapid introduction and spread of Campylobacter in a well cleaned processing plant as well as a significant survival in time on the example of a cutting board is shown.

Significance and Impact of the Study:  The need to prevent cross-contamination in the food processing and preparation area and the importance of an integrated approach throughout the whole food chain to control transmission of Campylobacter is highlighted.

Introduction

Campylobacter spp. is recognized as the leading cause of bacterial food-borne diarrhoeal disease in industrialized countries. Campylobacter jejuni has an isolation rate which exceeds that of Salmonella and accounts for >90% of all Campylobacter infections in humans (Park 2002). The four major sources of infection are undercooked poultry meat, untreated water, raw milk and pets (Buswell et al. 1998). Poultry meat is internationally recognized as an important vehicle for Campylobacter infection (Humphrey et al. 1995). As the skin is the primary residence of Campylobacter on chicken carcasses (Berrang et al. 2002), it is surprising that 25–83% of chicken pieces, stripped of their skin, for sale on the retail market are contaminated with Campylobacter (Uyttendaele et al. 1999; Frost 2001).

Campylobacter spp. were isolated from sampled equipment along the processing line of broiler carcasses, from the chicken transport crates to the chillers, as well as from the air (Berndtson et al. 1996). The effect of the different processes from stunning and kill, scalding, defeathering, evisceration, washing and chilling on the incidence of Campylobacter is well documented (Sanchez et al. 2002; Anon. 2003). Few reports have been found in the literature regarding contamination of poultry meat during portioning but automatic portioning equipment is likely to be a potential source of contamination (Anon. 2003). In addition, other surfaces that the portions come into contact with such as conveyer belts, cutting boards and packaging material, may add to the microbial load of the poultry cuts. Hands and clothing of factory personnel and utensils such as knives are also likely to contribute to microbiological contamination. The first objective of this study was to determine the spread and the persistence of C. jejuni in the processing line environment of a poultry processing plant, as cross-contamination can contribute to the maintenance of C. jejuni, resulting in the repeated contamination of processed chicken cuts. After sale, these contaminated chicken cuts are introduced in the institutional large-scale or consumer's kitchen. Cross-contamination from raw poultry meat is believed to be important in the transmission of campylobacters to man. Cogan et al. (1999) demonstrated the easy spread of Campylobacter spp. from naturally contaminated chickens to kitchen surfaces. In a Dutch investigation, Campylobacter spp. were found on cutting boards in 38 of 76 trials (50%) after handling raw chicken. In the same study Campylobacter spp. could be detected in two of 21 samples of heat-treated chicken handled on a cutting board previously used for raw chicken (De Boer and Hahne 1990). Current studies primarily report on the presence of contamination on various surfaces following processing or preparation activities. However, as surfaces become contaminated, it is important to know if and for how long Campylobacter can persist on these surfaces and thus can contribute to infection by cross-contamination. The second objective of the study concentrated on providing a quantitative estimate of the survival of C. jejuni on a beech and a polypropylene cutting board, determined by its culturability on a nonselective medium.

Materials and methods

The poultry processing plant

The plant was organized as shown in Fig. 1. Raw poultry carcasses coming from different slaughterhouses were mechanically stripped of skin and bones. The produced poultry breasts were stripped of remaining fat by hand. Afterwards, poultry breasts of the same weight were collected in a box and packaged for transport to the clients. Another part was cut into smaller pieces for the production of poultry brochettes.

Figure 1.

Organization of the poultry processing plant (a: white conveyer for transport of poultry carcasses; b: stripping of skin and bones of carcasses; c: stripping of remaining fat of poultry breasts; d: collection of poultry breasts; e: cutting poultry breasts for brochettes) and places tested in the environment of a processing plant at days 1, 2 and 3. (1: poultry carcasses; 2: white conveyer; 3: glove personnel handling poultry carcasses; 4: workbench 1; 5: blue conveyer 1; 6: knife; 7: glove personnel working with knife; 8: workbench 2; 9: blue conveyer 2; 10: workbench for cutting meat for brochettes; 11: poultry breast meat; 12: poultry meat for brochettes)

Sampling in the environment of a poultry processing plant

On three different sampling dates, parts of 25 cm2 of different surfaces were swabbed using a cotton swab moistened in peptone saline solution [PPS; 1 g l−1 peptone (Oxoid) and 8·5 g l−1 NaCl (VWR, Leuven, Belgium)]. On the first sampling date, 18 surfaces and three poultry meat products were tested at a single time interval for the presence of campylobacters. Eight of 18 surfaces were positive for Campylobacter. These eight surfaces and one additional surface were examined on the second and third sampling date. Only the surfaces that were sampled on all three dates are presented in Fig. 1 (1–12). On the second and third sampling date, swabbing of each surface in triplicate was performed before processing started (0 min), after 30 min and after 120 min. Furthermore, on the second sampling date at each time interval 10 portions of freshly produced poultry breast meat were taken. At 0 min and after 30 min, also poultry meat used to prepare brochettes (a wooden stick with small pieces of chicken breast meat) was taken (after 120 min this meat was not available anymore as production of brochettes had finished). On the third sampling date, in addition to freshly produced poultry breast meat, 10 broiler carcasses (starting material) were taken before processing (t = 0) and after 30 min. Broiler carcasses could not be taken after 120 min because the batch was already finished. No meat for brochettes could be taken as production was not performed that day.

Treatment of samples

The swab, submerged in Bolton broth containing 5% lysed horse blood and the antibiotic supplement, was kept cool during the sampling procedure and the transfer to the laboratory. On arrival, the inoculated Bolton broth was immediately incubated for 4 h at 37°C and for 44 h at 42°C under a micro-aerobic atmosphere. Subsequently, 100 μl was spread on karmali agar (Oxoid) and incubated for 48 h at 42°C under a micro-aerobic atmosphere. Suspected colonies were picked and streaked for purity on karmali agar. Confirmation of Campylobacter spp. was performed by microscopic examination of typical spiral morphology and motility and by the absence of growth upon aerobic incubation at 37°C and micro-aerobic incubation at 22°C. Of each poultry meat sample, 10 g was added to 90 g Bolton broth with the selective supplement. The suspension was homogenized for 30 s. Incubation was performed at 37°C for 4 h and at 42°C for 44 h under a micro-aerobic atmosphere. Subsequently, 100 μl was spread on karmali agar (Oxoid) and incubated at 42°C for 48 h under a micro-aerobic atmosphere. Isolation and confirmation of suspected colonies was performed as described above.

Bacterial strains and maintenance

Four strains of C. jejuni were selected based on their persistence in drinking water (Cools et al. 2003). The reference strain (LMG 6629) was obtained from the BCCMTM/LMG bacteria collection of the Laboratory of Microbiology, Ghent University (Ghent, Belgium). The poultry isolate (KC 592) was a gift from M. Heyndrickx of the Department for Animal Product Quality and Transformation Technology, Centre for Agricultural Research (Melle, Belgium). The water isolate (RIZA 488) was obtained from Wilma Jacobs-Reitsma, Animal Sciences Group, Wageningen University (the Netherlands). The human isolate (79) came from M. Federighi, Unit of Food Hygiene Research, Veterinary School of Nantes (France). A stock culture of each strain was kept at −70°C in plain horse blood without cryo-protection (E&O Laboratories, Bonnybrigde, UK). A working culture of each strain was kept at 4°C in Bolton broth (Oxoid, Basingstoke, UK) (Baylis et al. 2000) containing 5% lysed horse blood (E&O Laboratories) and the Bolton antibiotic supplement (Oxoid) under a micro-aerobic atmosphere using Campygen packs in anaerobic jars (Oxoid). The working culture was transferred monthly. Campylobacter jejuni strains were activated by inoculation of a loopful in 10-ml Bolton broth containing 5% lysed horse blood and the Bolton antibiotic supplement and incubated under a micro-aerobic atmosphere for 24 h at 37°C (24-h culture).

Survival of C. jejuni on beech and polypropylene cutting boards

A beech cutting board and a polyethylene cutting board were used. Both cutting boards were scratched to imitate the situation in the domestic kitchen as well as possible. On both surfaces 18 small test surfaces of 25 cm2 were marked. To accomplish sterile starting conditions the cutting boards were rinsed with boiling water and disinfected with 70% ethanol (Merck, Darmstadt, Germany).

One millilitre of a 24-h culture was centrifuged for 5 min at 13 600 g and resuspended in 10-ml Bolton broth containing 5% lysed horse blood but without the selective supplement added. A 200-μl volume of this suspension was spread on 15 of the test surfaces by means of a cotton swab and subsequently left at room temperature. Three noninoculated 25 cm2 surfaces were included as controls. Immediately after the application (t = 0), three 25 cm2 surfaces were swabbed using each time a new cotton swab moistened with PSS. Each swab was submerged in 1 ml of Bolton broth containing 5% lysed horse blood but without the selective supplement added. Appropriate dilutions of the latter were spread on Columbia Blood Agar [CBA: Columbia agar (Oxoid), supplemented with 5% lysed horse blood]. After 48 h of incubation at 37°C under a micro-aerobic atmosphere, colonies were enumerated. The same was performed after 30, 60, 90 and 120 min. The noninoculated surfaces were swabbed after 120 min. For every C. jejuni strain the experiment was performed in triplicate and repeated at three different days.

Results

The spread and persistence of campylobacters in the processing environment is shown in Table 1. At the first sampling date, 18 surfaces were tested. However, only eight were positive for Campylobacter. These positive surfaces and an additional one were included in testing on the second and third sampling date and results are presented in Table 1.

Table 1.  The presence of campylobacters at different sites in the processing environment and the poultry products produced at the plant at the three sampling dates. Results are displayed as numbers of positive samples on total samples tested
Sites and products tested in the processing environment*Date 1 (120 min)Date 2Date 3
0 min30 min120 min0 min30 min120 min
  1. ND, not determined.

  2. *Numbers in agreement with the numbers in Fig. 1

 1 Poultry carcasses0/5NDNDND0/100/10ND
 2 White conveyer1/30/33/30/30/30/31/3
 3 Glove personnel handling poultry carcasses1/10/32/30/30/31/32/3
 4 Workbench 11/10/31/31/30/30/32/3
 5 Blue conveyer 11/20/33/30/30/30/31/3
 6 Knife2/20/33/30/30/30/31/3
 7 Glove personnel working with knife0/10/33/30/30/30/32/3
 8 Workbench 21/10/33/31/30/30/30/3
 9 Blue conveyer 21/10/31/30/30/30/30/3
10 Workbench for cutting meat for brochettes1/10/32/30/30/31/30/3
11 Poultry breast meat1/28/104/102/104/101/101/10
12 Poultry meat for brochettes0/19/109/10NDNDNDND

On the second and the third sampling date, all the sampled surfaces were negative for Campylobacter before the start of the production. On the second sampling date, all surfaces became contaminated after 30 min. After 120 min only sampling points 4 and 8 remained contaminated. On the third sampling date, sampling points 3 and 10 were positive for Campylobacter after 30 min. After 120 min, an additional contamination of sampling points 2, 4, 5, 6 and 7 were noticed whereas for sampling point 10, contamination was no longer noticed. Hence, contamination with Campylobacter decreased in time on the second date of sampling but increased on the third date of sampling.

On the second sampling date, immediately after production started, eight of 10 of the freshly produced chicken breasts were contaminated, after 30 min the contamination had decreased to four of 10 and after 120 min to two of 10. On the third sampling date four of 10 poultry breasts were contaminated immediately after production started. A decrease to one of 10 was noticed after 30 min and contamination level remained at one of 10 after 120 min. At 0 min as well as after 30 min, 90% of the poultry meat used to prepare brochettes was contaminated. None of the examined carcasses taking on the third sampling date was Campylobacter positive.

The survival of the four C. jejuni strains on a beech and polypropylene cutting board is shown in Fig. 2. The results are the mean of nine individual experiments (three replicates for three separate experiments). The same trend, namely a three log reduction during the first 30 min of incubation of the C. jejuni strains on the surface at room temperature can be noticed. After 30 min the numbers of campylobacters recovered from the surface of the polypropylene cutting board remained rather constant although variation occurred indicated by the rather large 95% confidence intervals. Survival of C. jejuni on the beech cutting board was characterized after 30 min by a trend to gradual decrease of the numbers of C. jejuni. Campylobacter jejuni strain 79 was not recovered after 90 min but this may be as a result of the lower inoculum levels used. Overall a mean log reduction of 3·99 log ±0·56 was established on a polypropylene cutting board and 4·61 log ±0·41 on a beech cutting board. In general, no difference in survival of C. jejuni was obtained by type of strain or by type of cutting board. However, it can be noticed from Fig. 2b that the poultry isolate (KC 592) was after a initial die-off persistent at the surface of both cutting boards.

Figure 2.

Survival of Campylobacter jejuni strains LMG 6629 (a), KC 592 (b), RIZA 488 (c) and 79 (d) on a beech (bsl00046) and a polypropylene (bsl00001) cutting board. The results, expressed as CFU per 25 cm2 are the mean of nine individual experiments (three replicates for three separate experiments). The 95% confidence interval is represented by the vertical bars

Discussion

Campylobacters have the reputation to be extremely sensitive to the extra-intestinal environment. The organisms are micro-aerophilic, and render nonculturable upon exposure to air and light. Furthermore, they are highly susceptible to desiccation and do not survive well on dry surfaces (Fernandez et al. 1985). But in Belgium still 40% of the chicken pieces for sale on the retail market are contaminated with Campylobacter (Uyttendaele et al. 1999). In the processing plant, no campylobacters could be recovered at each of the sampling dates before the processing started, suggesting that the cleaning and disinfection procedure is effective. However, on the second sampling date, 80% of the poultry breasts, produced at the onset of the processing were contaminated with C. jejuni. An inquiry with the production manager revealed that the first batch of processed carcasses derived from a slaughterhouse that is known to occasionally deliver carcasses heavily contaminated with C. jejuni. The initial contamination of these carcasses gave rise to the C. jejuni contamination on all sampled surfaces after 30 min. Although the second batch of carcasses was not contaminated with C. jejuni, cross-contamination from the contaminated surfaces still gave rise to contamination of 40% of the poultry breasts taken at the end of the processing line after 30 min. Finally, after 120 min of processing, only two of nine surfaces remained contaminated with C. jejuni and 20% of the chicken breasts were Campylobacter positive. It seems that campylobacters initially introduced to the surfaces can contribute to cross-contamination of the poultry breasts. However, they also seem to be sensitive to the adverse environmental conditions as after 120 min, they are not longer detected on all the initially contaminated surfaces, at least not with the traditional culture methods. In contrast, on the third sampling date, no carcasses were positive for Campylobacter and only a moderate level of poultry breasts were contaminated with C. jejuni at the onset of processing but six of nine surfaces were Campylobacter positive after 120 min of processing. The presence of Campylobacter on the last batch of chicken carcasses could not be determined, as stated in Materials and methods. From the results of the second sampling date, it could be assumed that this batch was one with a higher contamination level giving rise to the contamination of these six surfaces after 120 min.

These results can explain the high numbers of contaminated poultry meat on the retail market. If the initial carcasses are contaminated with Campylobacter, also the contact surfaces in the processing environment become contaminated. These surfaces contribute to cross-contamination of noncontaminated poultry meat. Therefore, starting with good quality poultry carcasses should diminish the contamination level of processed poultry products.

Furthermore, it has been established that the handling of contaminated raw poultry meat in the food preparation area will cause the bacteria to become widely disseminated (Cogan et al. 1999). The question arises during which time period they can survive and persist on these surfaces and thus play a role in cross-contamination and subsequent infection. Kusumaningrum et al. (2002) inoculated stainless steel surfaces with a test suspension of C. jejuni and noticed a three log reduction in the first 30 min. After 4 h of incubation, no campylobacters could be recovered by this author. When droplets of blood containing C. jejuni cells were placed on squares of formica, no campylobacters could be isolated after 2 h, i.e. the time needed for the droplets to dry (Humphrey et al. 1995). In the present study, the same three log reduction in the first 30 min was noted and in general, a low level (≤103 CFU/25 cm2) of Campylobacter was recovered after 2 h of incubation on a beech as well as on a polypropylene cutting board. The discrepancy with the results from other authors cited above could be due to the differences in experimental conditions, e.g. origin of C. jejuni strains, method of surface contamination, sampling method, enumeration methods, etc. Polypropylene surfaces are considered as nonporous materials, which make them the most appropriate surfaces to use in food preparation as they give the bacteria less chance to escape cleaning and disinfection. Although C. jejuni can persist on polypropylene surfaces cross-contamination can be reduced when regularly subjected to cleaning and disinfection. In contrast, wood has a porous character and can absorb microorganisms and fluid. Raloff (1993) established experimentally that microorganisms, once in the interior of the wood, may persist in the inner structure and unless they are thoroughly cleaned and disinfected may continue to contaminate foodstuff. Probably, there is a higher risk for cross-contamination from C. jejuni surviving and persisting at the surfaces of a polypropylene cutting board than from C. jejuni being present in the deep interior of the wood, as the latter have less direct contact with foodstuffs. However, C. jejuni is removed more easily from the upper nonporous surface of the polypropylene than from the interior of the deep crevices in the wood. Therefore, both wood and polypropylene have their (dis)advantages to be used as material for cutting boards used in the food preparation area. Our results did not show a significant effect on the survival and persistence of C. jejuni by the type of cutting board or the type of isolate.

In conclusion, campylobacters, introduced in a poultry processing environment by contaminated poultry carcasses, can persist on these surfaces and also contribute to cross-contamination of the final products. The poultry breasts, although stripped of skin which is known as the primary residence of Campylobacter on poultry, showed incidences of 10–80%. The prolonged survival (up to 120 min), after an initial three log decrease of the C. jejuni strains on the surface of for example a cutting board further confirms the need to prevent cross-contamination in the food processing and preparation area. These results clearly show the responsibility of the whole food chain in solving the Campylobacter problem because the processing plant or preparation area can only have an efficient control over the spread and contamination by Campylobacter if Campylobacter-free carcasses and poultry meat is used as starting material.

Acknowledgements

Ils Cools is indebted to the Special Research Fund of the Ghent University for a doctoral fellowship. Mieke Uyttendaele is indebted to the Fund for Scientific Research (FWO-flanders, Brussels, Belgium) for a position as Postdoctoral Research Fellow. The authors want to thank Wilma Jacobs-Reitsma, Marc Heyndrickx and Michel Federighi for supplying the C. jejuni strains.

Ancillary