L. M. Tam and D. E. Conner
Poultry Science Department,
Auburn University,
Alabama, U.S.A
Research was conducted to define the microbial ecology of poultry products and equipment surfaces in commercial broiler deboning operations, and to determine effects of processing on predominant bacterial types. Sampling of whole carcasses, breast meat, and equipment surfaces yielded 600 isolates. Among these isolates, 34 different genera representing 98 different species were found. Staphylococcus, Pseudomonas, Flavobacterium and Acinetobacter were the predominant genera; however, the overall bacterial community varies by sample type.
Clear patterns of cross contamination were observed. Isolates were assayed to determine their ability to inhibit growth of Campylobacter jejuni, and 62 were found to be antagonistic to C. jejuni. Isolates of Pseudomonas aeruginosa were highly inhibitory to growth of the pathogen. Microbial ecology and changes that occur with commercial deboning procedures may impact the incidence of C. jejuni on processed.
Introduction
Poultry processing methods continue to evolve to meet consumer demand for a wide variety of convenient ready-to-cook products. While there are abundant data on factors affecting bacterial contamination of poultry during initial slaughter and processing, there are little data on contamination during further processing steps such as cut-up and deboning. Moreover, poultry is a known reservoir for the pathogen C. jejuni, and there is a high prevalence of this pathogen on whole broiler carcasses, while recent reports indicate a significantly lower prevalence of C. jejuni on further processed poultry (Davis et al., 2002). The reason for this difference is unknown, however competition by other bacteria may, in part, account for the observed lower prevalence. Thus, research is needed to further define the microbial ecology of poultry products in the modern market place, and to determine the potential of this ecology to influence overall safety of these poultry products.
Material and methods
In experiment 1, samples were taken at random at the postchill-debone lines of the same commercial processing facility over three visits. Collection at each sampling time consisted of 20 whole carcasses (W), 20 breast skins (SK), 20 breast meats (BM), and 20 equipment (E) samples.
Whole carcasses were collected immediately after immersion chilling. Skin samples were collected at the skin transfer belt. Breast samples were gathered immediately after their removal from carcasses. Equipment surfaces were sampled by swabbing, and included debone belt, splash plate behind debone belt, splash plate to the side of debone line, overhead drip pan, knife sharpener, drip pan under cone line, skin transfer belt, employee apron, employee glove, and cone line.
Samples (W, SK, BM) were individually placed in sterile plastic bags, with 400 ml buffered peptone water (BPW), and shaken for one minute. Equipment surfaces (E) were swabbed using templates (160 cm2). Sterilized gauzes moistened with buffered peptone water were used to swab the areas inside the templates. After swabbing, gauzes were placed in 400 ml BPW. Each rinse fluid was serially diluted using Butterfield's phosphate diluent and spiral plated (50µ) onto tryptose soy agar (TSA). Plates were incubated at 37°C for 24 h. Isolated colonies exhibiting different morphology were selected, and streaked for isolation on fresh TSA. The total number isolates obtained from the various sample types are as follows:
Isolates were identified based on fatty acid composition using the Sherlock™ system (MIDI, Inc., Newark, DE, USA). Isolates were streaked onto TSA with 5% sheep blood (Remel) and incubated at 35°C for 24 hours. After incubation, there were five steps in the preparation of GC-ready extracts from cell cultures for fatty acid composition analysis (MIDI, Inc.), namely harvesting, saponification, methylation, extraction, and base wash. Harvested cells were removed from culture medium. One loopful (approximately 40 mg) of live wet cells harvested from the most dilute quadrant exhibiting confluent growth (late log phase) along the streaking exit was used for identification procedure. Harvested cells were placed into a clean dry 13 mm Ν 100 mm screw cap culture tube. The next step was saponification, in which cells were lyzed to liberate fatty acids from cellular lipids, which were subsequently methylated to convert fatty acids to fatty acid methyl esters to increase the volatility of the fatty acids for GC analysis. Following methylation, fatty acid methyl esters were removed from the acidic aqueous phase and transferred to an organic phase.
A mild base solution was then added to the sample preparation tubes to remove free fatty acids and residual reagents from the organic extract prior to chromatographic analysis. Fatty acids extracted from the isolates were automatically quantified and identified by the Sherlock™ software to determine the fatty acid composition. The fatty acid profile was then compared to a library of reference organisms to determine the identity of the unknown isolate. Isolates were identified to the species level.
In experiment 2, the amount of inhibition that other poultry bacterial isolates have on C. jejuni was determined.
In this experiment, two assays were used: agar and broth procedures. For the agar assay, 600 bacterial isolates collected from a commercial deboning line (from experiment 1) were used. TSA (30ml) was allowed to solidify in an agar plate. After solidification, five wells were punched into the agar. Three of these wells received 50µl of TSA containing 108 cfu/ml of the isolate being tested for inhibition of Campylobacter. One well received 45µl of TSA plus 5µl of acetic acid (3N). The final well received 50µl of TSA. These plated were incubated at 35°C for 24 hours. After incubation, the plates were subjected to UV light for 2 hours. After this exposure, the plates were overlayed with 10 ml of Campy-cefex agar containing 107 cfu/ml of C. jejuni. The plates were then incubated microaerophilically at 42°C for 48 hours. After incubation, the plates were examined for zones of inhibition. 62 isolates representing 26 species produced zones of inhibition and were subjected to the broth assay. For the broth assay, each target isolate was inoculated (109 cfu/ml) into 9 ml Brucella broth and incubated microaerophilically at 42°C for 24 hours. These isolates were then plated onto both TSA and Campy-Line media. These plates were incubated at 42°C for 48 hours. This was to ensure that the target organism did not grow on the Campy-Line agar. To determine the amount of inhibition of each isolate, 1ml of each isolate (109 cfu/ml) and 1 ml of C. jejuni (108 cfu/ml) were inoculated into 8 ml Brucella Broth. This broth was incubated at 42°C for 24 hours. After incubation, the mixture was plated onto Campy-Line agar and incubated at 42°C for 48 hours. For control, 1ml of C. jejuni was inoculated into 9 ml Brucella Broth and incubated at 42°C for 24 hours. This broth was then plated onto Campy-Line agar and incubated at 42°C for 48 hours. To measure inhibition, the population of the mixture of C. jejuni and the target organism was subtracted from the population of pure C. jejuni.
Results and Discussion
In experiment 1, a total of 670 isolates (210 W, 112 SK, 244 BM, 104E) were obtained based on different morphology on TSA. Of these, 600 isolates (193 W, 87 SK, 222 BM, 98 E) were identifiable using the Sherlock™ system. Among the 600 isolates identified, there were 34 different genera, representing 98 different species. Staphylococcus, Pseudomonas, Flavobacterium, and Acinetobacter were the most predominant genera (Table 2).
Processing environment affected bacterial population as evidenced by observed increases in the proportion of psychrotrophic bacteria during sequential broiler deboning operations (Table 3). Based on this shift in bacterial types it appears that equipment surfaces serve as the primary source of psychrotrophic bacteria on breast skin and meat. The high prevalence of staphylococci on whole carcasses (42.49%) and breast skin (19.54) and the low prevalence of these bacteria on equipment surfaces (10.20%) and breast meat (7.66%) indicate that the plant environment was not the main source of the staphylococci contamination on broiler products. Enterobacteriacea were found from all the four types of samples, however, cold tolerant genera of the family Enterobacteriacea such as Serratia, Enterobacter, Citrobacter were found at high proportion on equipment surfaces, and thus also found at high proportion of breast meat.
The large number of genera found in breast meat (data not shown) indicated that significant cross contamination occurs during the deboning operation. Thus, steps to minimize this cross contamination would likely lead to product of improved microbiological quality.
In experiment 2, sixty-two of the initial 600 isolates produced zones of inhibition in the agar assay of the experiment (Table 4). These 62 isolates represented 26 different species of bacteria. In the broth assay, the amount of inhibition was measured. Of the 62 isolates, 26 (representing 15 species) reduced the growth of C. jejuni by 1-2 log10 cfu/ml. Twenty six isolates (representing 16 species) reduced the growth of C. jejuni by 2-3 log10 cfu/ml and 10 isolates (representing 1 species) reduced the growth of C. jejuni by > 3 log10 cfu/ml.
Chemical substances released by these microorganisms may account for the inhibitory effects on C. jejuni, including antibiotics, bacteriocins, hydrogen peroxide, and some organic acids. Nonetheless, presence of other poultry borne bacteria influence the survival of C. jejuni, which may affect the ability of this pathogen to persist on raw poultry meat.
References
Barbut, S. 2002. Poultry products processing - an industry guide. CRC Press, Boca Raton, Florida
From Proceedings of the "XVI European Symposium on the Quality of Poultry Meat" and the "X European Symposium on the Quality of Eggs and Egg Products", Saint-Brieuc Ploufragan, France.
