Comparison of Poultry Processing Equipment Surfaces for Susceptibility to Bacterial Attachment and Biofilm Formation

MATERIALS AND METHODS
Sample Preparation
New York-dressed carcasses were collected from a commercial broiler processing plant, placed on ice, and transported to the laboratory. The time for transport until processing was about 30 min. Each carcass was bagged, weighed, and then rinsed with 100 ml, phosphate-buffered saline. Over three trials, aerobic plate counts showed that the average number of bacteria present in the suspensions was 2.12 x 106/mL, rinse. Aliquots of the rinse were diluted with nutrient broth1 in duplicate, 10- fold series for further testing. Test surfaces (1x4 cm) were materials, not previously used, which included stainless steel, conveyor belting/ polyethylene,C and picker-finger rubber. The surfaces were washed briefly in 1% Micro cleaning solution,rinsed irt distilled water, and sonicated for 30 min before being added to test tubes containing the carcass rinse. Surfaces were added to tubes containing diluent for negative controls, and tubes with carcass rinse and no surfaces were used as positive controls. The pi I was measured for all samples before and after assay. The test intervals and length of time that the test materials were held in the rinse suspensions varied for each experiment as shown in the Results section and Figures 1 to 4. Changes in turbidity were measured by spectrophotometry during these incubation times, and then the surfaces were removed and prepared for microscopy as described below.
Turbidimetric Assays
Assays were performed with mixed microbial populations obtained from the whole carcass rinse. Initially, the rinses were diluted with nutrient broth for absorbance readings. Absorbance values of the carcass rinse suspensions were measured by a Beckman DU640 spectrophotometer equipped with Peltier temperature controller and auto cell holder.8 Samples were incubated at 37 C, 80 to 85% RH, during the experiments. The temperature and humidity chosen represent conditions during early processing, for example, the picker room. Prior to the experiments, wavelength scans from 350 to 750 nm at 1,200 nm/min were made of the carcass rinse suspensions to determine maximum response. During the experiments, spectrophotometric readings of the suspensions were taken at 412 nm (maximum response during the scans). Absorbance values of carcass rinse suspensions containing test surfaces were compared with values of carcass rinse suspensions without test surfaces (controls). During the course of the experiments, loss of turbidity in the suspensions was monitored to estimate reduced numbers of bacteria relative to controls. Numerical data were evaluated by the pooled-variance t-test (Shott, 1990) and AN- OVA to compare the differences observed between sample and control. Probability values are given where applicable for each experiment. Differences were determined between surfaces within individual trials and for all trials combined.
Scanning Electron Microscopy
The test surfaces used in the assay described above were examined by scanning electron microscopy (SF.M). The surfaces were removed from the test tubes, rinsed with 0.1 M sodium cacodylate buffer to remove unattached bacteria, and fixed for 2 h at room temperature with a mixed fixative of 1% glutaraldehyde and 1% paraformaldehyde1 in 0.1 M sodium cacodylate buffer. Then, samples were rinsed in 0.1 M sodium cacodylate buffer, dehydrated in ascending concentrations of ethanol, critical-point dried in a Samdri 814 critical point dryer, mounted on aluminum stubs, and sputter-coated with gold-palladium in a Samsputter 2-A." The samples were examined in a JEOL 6400V SEM, at an accelerating voltage of 5 kV.
Energy Dispersive X-Ray Analysis

Test materials were analyzed for elemental composition using energy-dispersive X-rav analysis with (SIS v.3.0 software on a JEOL 6400V SEM. These samples were mounted on carbon stubs and were carbon-coated. Initial analyses at accelerating voltages of 10, 15, and 20 kV were performed to determine elements present and the best accelerating voltage for resolving the elemental peaks and minimizing charging effects. Elemental X-rays were detected by a Si(Li) spectrometer detector with a 7.5 mm thick beryllium window, an active area of 10 mm2, with the specimen at a 0 degree tilt angle. Subsequent elemental analyses were at 15 kV. Spectra were also collected using the thin window position of the Oxford detector to determine if any significant light elements could be found in trace amounts.