Pr oject Summar y. Reduction of foodborne pathogens on cattle during loading through control of dust generation

Transcription

1 Pr oject Summar y Reduction of foodborne pathogens on cattle during loading through control of dust generation Principal Investigator: Mark Miller, Mindy Brashears, Chance Brooks, Department of Animal and Food Sciences and Guy Loneragan Texas Tech University Study Completed May 2007 Funded by The Beef Checkoff 1

2 Reduction of foodborne pathogens on cattle during loading through control of dust generation Background We and others have clearly demonstrated that pathogen loads carried on the hides of cattle increase from the time they leave their so-called home pens at the feedlot to the time they arrive at the slaughter facility (Barnham, 2002; Beach, 2002; Fluckey et al. 2004). However, the cause of this increase is not fully known. Initially, we hypothesized that cross-contamination was occurring while cattle were on the trucks and might be related to the cleanliness of the trucks. In the summer of 2004, we conducted a study that examined prevalence of pathogens on cattle hides after they were removed from their home pen and after transportation to a commercial abattoir in either a clean or dirty truck. Despite our initial hypothesis, the cleanliness of the truck (clean or dirty) did not impact pathogen loads on hides during shipping. At both before and after shipping, pathogen prevalence on hides did not vary between cattle transported on clean or dirty trucks. Despite not detecting an effect of the truck, pathogen carriage on hides of cattle in both groups increased to a similar extent between the time they were sampled at the feed yard and the time they arrived at the abattoir. As stated, this increase was not related to the cleanliness of the truck. We observed significant air-borne particulate matter (dust) while collecting samples for this study. This dust was generated by the movement of cattle from the surface of the holding/loading pens prior to being loaded onto the truck. Importantly, we sampled the cattle prior to passing through these dusty holding pens. Moreover, most large commercial feedlots in the U.S. will use similar and frequently dusty holding pens prior to shipping. Because this dust is essentially dried and pulverized fecal material and presumably could contain high numbers of pathogens from the loading process, we hypothesized that generation of air-borne dust may increase pathogen loads carried on hides of cattle that come into contact with this dust. This in turn, might explain at least a portion of the increase in pathogen carriage observed from the feedlot to the abattoir. In other words, this dust may be an important source cross-contamination from the environment to the surface of in-contact animals and ultimately to carcasses. In the summer of 2005, we conducted another study to test our hypotheses 1) that the air-borne particulate matter contained pathogens; and 2) that exposure to this dust resulted in an increase in the prevalence and quantity of pathogens. We determined that the exposure to the air-borne particulate matter generated from the surface of loading pens significantly increased the prevalence and quantity of E. coli O157 and Salmonella recovered from the animals hides. Furthermore, we isolated these important food-borne pathogens from the air-borne dust which is an observation that has not been previously reported in the literature. Moreover, when pathogens were recovered from the dust, we observed the greatest increases in total numbers of pathogens isolated from cattle hides. We hypothesize that reducing dust generation from holding pens during loading will reduce pathogen load carried on the hides of cattle delivered to packing plants for slaughter. Because group-level hide contamination is a strong predictor for carcass contamination, effective dust control in pens used to hold cattle immediately prior to transport to packing plants will reduce carcass contamination and improve the safety of our food supply. Over the past two years, we have conducted studies examining the increase of pathogen numbers and prevalence in beef feedlot cattle during loading and transportation. Our data indicate that loading might be a critical control point to prevent cross-contamination in the farm-to-table system. Additionally, despite the reductions on the hides and in the feces of the cattle using preharvest interventions, there appears to be a consistent increase in the pathogen loads in the feed yard and the loads found on the hides at the plant. We initially attributed the increase in pathogen loads 2

3 to something that occurs during shipping. However our preliminary data indicate that the increase in pathogen loads occurs before the animals are put onto the trucks and appears to be related to the exposure to the cattle to the dust in the loading area. Dust production is common in cattle feedlots (Sweeten, 1988). Dust production in general peaks in the evening hours and it is associated with dry, warm conditions. However, a significant amount of dust is also generated during cattle loading. Dust levels generated around the loading chute are more than are typically observed in the evening hours and are limited to the loading chute area. It is possible that pathogens could attach to dust particles and be re-distributed onto other animals thus resulting in cross-contamination from animal to animal during loading. Dust potential in feedlots is inversely associated with pen-surface moisture. In much of the High Plains region of the United States rainfall and relative humidity is very low resulting in a significant dust potential from pen surfaces. Since particulate matter generated from pen surfaces is essentially dried and pulverized fecal material, dust clouds may contain pathogens and be an important environmental source of contamination for animal surfaces (i.e., hides). A limited amount of data exists on the presence of airborne pathogens in feed yards. Wilson et al (2002) collected dust from 8 commercial feed yards during dusk when the dust level was the highest in the feed yard. They reported that no foodborne pathogens were detected in the dust. They did not even recover any gram negative bacteria in their samples. They detected small numbers of gram positive bacteria as well as fungi. Seedort et al (1998) measured the concentrations of microorganisms in livestock building in Northern Europe. They reported that the average total bacteria in cattle buildings were 4.3 log10 cfu/m3 of air. They did not measure any foodborne pathogens in the air samples. To date, these are the only published data on the microbial condition of dust or air in cattle settings. The stated objectives for this work were: The objective of this study will be to determine if controlling dust created during the loading of cattle will result in a reduction of pathogen loads on the hides before and after shipping. Methodology We investigated the control of dust during loading at the feed yard and its impact on the prevalence and total numbers of E. coli O157 on cattle hides. We evaluated the following conditions to determine if preventing dust generation would result in a decrease in pathogen numbers. 1) The complete covering of the load out area to eliminate dust during loading using a concrete loading area and 2) Dirt loading area (control). We measured the pathogen loads on hides before and after loading and after shipping we sampled hides at the harvest facility during bleeding. Cattle hides were sampled before and after loading and the air was sampled as well. E. coli O157 was isolated from the hides and air. Details of sampling and microbiological analysis are described below. Experimental Design During the months of July, August, and September in the summer of 2006 cattle were sampled in 5 separate sampling intervals in a commercial feedlot. All samples were collected in the same feedlot as follows: 1. Load out area #1 Dirt Floor CONTROL 2. Load out area #2 Concrete Floor Cleaned before cattle are loaded 3

4 The collaborating feed yard was the Cargill beef yard in Lockney, Texas and they allowed us to use two load out areas, one with dirt and one clean. On each day of sampling the pathogen cloud dust cattle were loaded through the dirt load out pen. Another group of animals were subjected to sampling in the covered and clean load out area. Approximately 30 animals (load size varied between 30 and 34 animals depending on size) were sampled before loading which represented one truck load of cattle for each treatment (two trucks/sampling time). These animals were all loaded onto a truck and each group was kept on separate trucks (two trucks/sampling day). At total of 300 animals were sampled before at the feed yard and after exposure to the pathogen cloud dust at the packing plant for a total of 300 samples collected at the feed yard. After arrival at the slaughter facility, the cattle were sampled again at the commercial slaughter facility. This resulted in a total of 300 samples collected at the slaughter facility. The total number of samples collected for the entire study was 600 at the feed yard and 300 at the packing plant for a total of 900 total samples. Treatments: Control A typical rectangular loading pen was used to hold animals prior to loading. The area was 25 ft x 15 ft. Animals were placed into this loading area prior to loading onto the truck after they went through the squeeze chute for weighing. The floor of this pen was composed of dirt and feces from all animals that have been previously loaded and was the standard type of holding pen used across the US although size may vary somewhat. In typical feedlot management practices, this area is not cleaned. Animals were sampled in the chutes after being removed from their home pens, exposed to the dust in the loading pen for 3 minutes (typical trucks are loaded in 3-5 min) and loaded on trucks and shipped to Cargill harvest facility in Plainview, Texas. They were harvested at the beginning of the shift and samples were obtained after stunning at the harvest facility. Concrete Floor The dirt in the loading pen was removed and the concrete cleaned. Animals were sampled in the same manner as the control at both the feed lot and harvest facility. On each day of sampling, a total of 30 animals/treatment group were tested resulting in 60 animals sampled per day for a total of 5 days of sampling; this provided a total of 300 hide samples collected before loading, and 300 samples after shipping at the harvest facility. A total of 600 samples were collected 300 at the feed yard before loading and 300 at the packing plant. Cattle for the study were the first ones loaded for the day and were sampled before loading of any other cattle. Sampling Hides were sampled using a sterile spongesicle hydrated with sterile butterfields diluent. Animals were tagged and the left side of the animal was swabbed in a 500 cm2 area on the dorsal midline and withers resulting in a composite sample from 1000 cm2 of the hide prior to loading. After the animal passed through the loading area where the cloud of dust has been generated, it was sampled again on the right side on the dorsal midline and withers at the harvest plant. A consistent sampling area was used to facilitate both quantification and enumeration of E. coli O157 in the samples. Quantification is essential to determine not only if the pathogens are present, but also to determine if the pathogen load increased due to passing through the dust area. Air and dirt samples were collected at the feed yard in the home pens as well as in the load out area and analyzed for E. coli O157. An Andersen air sampler was used to collect air/dust samples for periods of 15 sec every minute for 3 min as described by Wilson et al (2002). Samples were collected before the animals enter the loading area, during dust production and 10 minutes after dust production for a total of 9 samples collected/treatment/day. A known quantity of air passed through the sampler each min. Samples were collected on a membrane filter and separate filters were collected for E. coli O157 determination. Filters were subjected to MPN protocols 4

5 described below to determine the amount of pathogen/liter of air. Additionally, the amount of dust particles passing through the filter was determined automatically by the filter giving us a quantitative measurement of the amount of dust particles in the air before, during and after dust generation. A total of 202 air/dust samples were collected in the study. We also sampled the load out area in 9 areas before loading and after all cattle were sampled. Dirt from the control areas was collected and a sponge sample in the concrete areas was collected. A total of 405 samples from the load out areas were collected during the course of this study. Microbiological Analysis E. coli O157 Isolation A sensitive assay that includes immunomagnetic separation was used to isolate E. coli O157:H7 (Elder et al, 200) from all samples. Ninety milliliters of GN-VCC (GN broth with 8 µg/ml of vancomycin, 50 ng/ml of cefixime, and 10 µg/ml of cefsulodin) broth was inoculated with 10 g of feces or with 10 ml of fluid from the sponge/raj samples and incubated for 6 h at 370C. E. coli O157:H7 cells was subjected to immunomagnetic separation by mixing 1 ml of the culture above with 20 µl of anti-o157 beads (Dynal, Lake Success, NY) for 30 min at room temperature. Beads were washed three times in PBS-Tween 20, and 50 µl of the bead-bacteria mixture was spread onto Chromagar plates. Typical colonies were streaked for isolation of individual colonies. Isolated colonies were tested for the O157 antigen using a latex agglutination kit (Remel, Lenexa, KS). Final confirmation was performed by PCR analysis for the O157 antigen on a Dupont BAX system (Dupont Qualicon, Wilmington,DE) and shiga toxin production using PCR. E. coli O157 Quantification A 3 X 5 MPN was conducted by adding 1 ml of the original sample into three parallel 9 ml tubes of gram negative (GN) broth making 3 tubes inoculated at a 10-0 MPN dilution. This procedure was repeated for four subsequent sets of three tubes for dilutions. All MPN tubes were incubated for 6 h at 370C. Immunomagnetic separation (IMS) was conducted on all of the MPN tubes after incubation. After IMS, 50 μl of the bead-bacteria complex was plated onto Chromagar and then typical colonies were streaked for isolation. After incubation, 5 typical E. coli O157 colonies were selected and agglutinated using commercial agglutination kits. Final confirmation was determined using PCR analysis for shiga toxins. Both typical and 2-3 non-typical colonies were confirmed positive by PCR analysis. The corresponding MPN tube to the plates that contain E. coli O157 were considered positive for the MPN calculations. We determined the MPN/g of feces using the positive tube combinations and compared them to published MPN charts in the FDA Bacteriological Analytical Manual. Findings Figure 1 illustrates the data collected on the hides of the animals sampled in the clean areas (no dust) and dirty (dust generation) areas in the feedyard. Animals were sampled before they were exposed to the loadout area, after exposure and then upon arrival at the plant. Total numbers of E. coli O157 were significantly lower on the hides of animals before and after sampling in the clean loadout area compared to the hides of animals exposed to the dirty areas at all sampling locations. Before exposure, the animals in the clean areas had an average of 1.21 log10 mpn/cm 2 E. coli O157 on the hides while after exposure to the loadout area, the total amount of E. coli O157 detected was 1.02 log10 mpn/cm 2. Upon arrival at the plant, the total numbers remained the same at 1.33 log10 mpn/cm 2. For animals subjected to the dirty loadout area, there was more than a 1 log cycle increase after exposure to the dust in the loadout area. The total numbers increased significantly from 2.48 log10 mpn/cm 2 to 3.66 log10 mpn/cm 2. Comparing the total numbers of E. coli isolated from animals in the clean area and in the dirty areas, there were significantly more E. coli O157 5

6 isolated from animals in the dirty areas compared to those in the clean areas at all sampling intervals. The total E. coli O157 on the hides at the plants on the animals subjected to the dust was at 3.26 log10 mpn/cm 2 on the hides which was significantly less than that on the hides at the plant which was 1.33 log10 mpn/cm 2. However, due to logistical restraints, a smaller sampling area was collected in the plant which could explain the differences. Figure 2 illustrates the percentage of animals that tested positive for the presence of E. coli O157 before and after loading in clean and dirty loadout areas and upon arrival at the harvest facility. There were no significant differences in the total percentage of animals testing positive that were exposed to the clean areas with values ranging from 31.2% before exposure to the loadout area, 37% after exposure, and 33% at the plant. For the animals exposed to the dirty/dusty loadout areas, there were no differences before and after exposure to the loadout area. However, upon arrival at the plant, the total amount significantly increased to 47% of the animals testing positive. E. coli O157 was present in the air at a similar rate that it was present in soil samples in the dirty/dusty loadout areas (Figure 3). The soil samples in the clean loadout area still contained similar amounts of E. coli O157 compared to the dirty areas. However, because most of the dirt/soil was removed, only 12% of the air samples became positive during loading compared to more than 50% of the air samples being positive in the dirty loadout areas. It is apparent that E. coli O157 present in the soil can become airborne. However on cleaned loadout areas, the pathogen did not become airborn. Implications This study illustrates that during loading, E. coli O157 can be transferred from the soil to the air if dust is not controlled in the loadout area. Subsequently, the hides of the cattle exposed to this dust had an increased amount of E. coli O157 on them after dust exposure. There were no increases in the total number of E. coli present on the hides of cattle when they were loaded through a clean loadout area. Additionally, we observed that while the soil in the loadout area did contain E. coli in the areas that were cleaned, it was not getting transferred to the air. Controlling the dust in a feedyard may be an effective intervention in reducing the amount of E. coli O157 that enters the harvest facility. 6

7 Figure 1. Total numbers of E. coli O157 detected on the hides of beef feedlot cattle after exposure to feedlot dust LOG 10 mpn/cm clean dirty 0 before after plant 7

8 Figure 2. Percentage of Animals that Tested Positive for the Presence of E. coli O157 Before and After Loading and at the Plant 50 % positive Before After Plant 0 clean dirty/dusty Condition of the Loadout Area 8

9 Figure 3. Presence of E. coli O157 in air and soil samples collected in loadout areas of the feedyard during sample collection % positive clean dirty 0 air soil For more information contact: National Cattlemen's Beef Association A Contractor to the Beef Checkoff 9110 East Nichols Avenue Centennial, Colorado (303)