Enhancing Food Safety Through Control of Food-Borne Disease Agents

NON-TECHNICAL SUMMARY: Microbial contamination of raw food commodities is an undesirable but generally unavoidable occurrence during processing. These microorganisms exist naturally in the environment in which the food is produced. This project examines the survival and growth of bacteria of public health significance in the environment during the production of foods of animal origins. In addition, it examines decontamination methods in an attempt to reduce both the numbers and the occurance of these organisms. <P>

APPROACH: 1) Research will be conducted to determine the survival of bacteria of public health significance in the production livestock environment. Since recent evidence suggests that Escherichia coli O157:H7 occurs more frequently on live cattle (rather than in the intestinal tract), survival on the bacterium in the environment may be an important determinant in the transmission of the bacterium into the human food chain. 2) Research will be conducted on the effectiveness of various antimicrobial intervention strategies on raw food commodities, specifically those of animal origin. The strategies will be developed with the fundamental concept of using these as potential critical control points in a HACCP inspection system. Interventions will include but not be limited to food grade additive such as organic acid and approved alkalis, as well as ozone.

<P>

PROGRESS: 2000/10 TO 2006/09
Salmonella Typhimurium DT 104 strains were inoculated into ground pork to determine the effect of fat content on the D-values after irradiation. The ground products were 90 % lean, 50:50 fat:lean and 100% fat. Two different sources of ionizing radiation were also compared. There was no significant difference in D-values for the three products for the individual process. Therefore, fat content had no effect. The average D-value for the e-beam and gamma radiation was 0.43 kGy and 0.60 kGy, respectively, which is significantly different at P = 0.001. The results of this study do show that these D-values are within the reported range for irradiation destruction of Salmonella contaminated meat products. Thermal tolerance of acid-adapted Escherichia coli O157:H7 or Salmonella in ground beef was evaluated during storage at 4C or -20C. Both pathogens were adapted to acidic conditions (pH 4.6) by growing in tryptic soy broth supplemented with 1% glucose. Decimal reduction time (D values) of the pathogens was determined as an indicator of thermal tolerance. Significantly higher D62 values were observed on days 21 and 28 for non adapted E. coli O157:H7 stored at 4C and on days 90 and 120 for non-adapted E. coli O157:H7 stored at -20C. Higher D62 values were observed on days 21 and 28 among non-adapted Salmonella strains stored at 4C and on day 28 for acid-adapted strains of Salmonella stored at 4C. Higher D62 values for acid adapted strains of Salmonella stored at -20C were observed on days 30, 60, and 90, while no differences were observed in the D65 values of acid adapted and non-adapted strains of E. coli O157:H7 and Salmonella throughout storage at both temperatures. This suggests that acid adaptation of food borne pathogens provides a certain level of protection against heat treatment at lower cooking temperatures, while at higher temperatures there were no observed differences between the sensitivity of acid adapted and non-adapted strains in an actual food system during an extended period of refrigerated and frozen storage. Reductions in the incidence of food animal illnesses might reduce bacterial contamination on meat, thereby reducing human illness. Antibiotic use in food animals might benefit human health by reducing the incidence of animal illness, but this use also can select for antibiotic resistant bacteria that can threaten human treatment options. The model is designed so that potential human health risks and benefits from continued antibiotic use in animal agriculture can be evaluated simultaneously. We applied the model to a hypothetical example of Campylobacter from chicken, and the model predicts that small increases in non-Campylobacter animal illness might lead to increased Campylobacter contamination on chicken meat that then could cause large increases in human campylobacteriosis. In general, the model suggests that very minor perturbations in microbial loads on meat products could have relatively large negative impacts on human health, and consequently, small improvements in food animal health might result in significant reductions in human illness.
<BR> <BR>
IMPACT: 2000/10 TO 2006/09<BR>
The research has resulted in the development of a predictive model to estimate the growth of salmonellae during the cooling of beef carcasses. This model was used during an evaluation of proposed USDA regulations to determine the relative probability of salmonellae growth under a variety of cooling conditions. The research work has been applied to reduce bacterial attachment by the use of carcass washing and sanitizing in animal processing environments, resulting in the development of an inexpensive, technology neutral process that is effective in controlling enteric pathogens on animal carcasses. The results of this project may result in practical methods to reduce microbiological contamination on foods of animal origins. In addition, the results of this project may increase the understanding of the presence of antimicrobial resistant bacteria on animal carcasses. This project also covered issues relating to the security and defense of food systems in the event of an intentional contamination event.
<BR><BR>
PROGRESS: 2005/01/01 TO 2005/12/31<BR>
Fresh meat products can become contaminated with the pathogen, E. coli O157:H7, during the slaughter process, so the identification of an E. coli O157:H7 indicator to verify the effectiveness of process controls in slaughter establishments would be extremely useful. Pre rigor lean and adipose beef carcass tissue was artificially contaminated with stationary phase cultures of the five generic E. coli beef cattle isolates or a cocktail of five E. coli O157:H7 strains in a fecal inoculum. Each tissue sample was processed with the following microbial interventions: 90C water; 90C water followed by 55C, 2% lactic acid; 90C water followed by 20C, 2% lactic acid; 20C water followed by 20C, 2% lactic acid; 20C water followed by 20C, 20ppm chlorine; 20C water followed by 20C, 10% trisodium phosphate. The appropriateness of the E. coli isolates as potential E. coli O157:H7 indicator was found to be dependent upon the microbial intervention utilized. For all microbial intervention methods applied irrespective of tissue type, the mean log10 reductions of at least two E. coli isolates were not significantly different (P>0.05) from the mean log10 reduction of the E. coli O157:H7 cocktail. Due to the frequent employment of multiple microbial intervention methods in industry no single isolate can realistically represent the effectiveness of all microbial interventions of E. coli O157:H7. Thus, the use of a combination of E. coli isolates tested here may be required to accurately predict the effectiveness of microbial intervention methods on the reduction of E. coli O157:H7 from beef carcass tissue. Approximately 100 colony forming units/cm2 of a five strain mixture of Listeria monocytogenes was co-inoculated on to frankfurters with three different populations (102, 104 and 106 cfu/cm2)of an undefined spoilage microflora, which was derived from commercial frankfurters. The frankfurters were vacuum packaged and stored at 10C for up to 48 days. The populations of L. monocytogenes, aerobic mesophilic bacteria, lactic acid bacteria and Enterobacteriaceae were determined at various time intervals during storage. After 14 d, the population of L. monocytogenes was highest with an inoculated spoilage microflora population of 102 cfu/cm2, and this trend continued until 48 d. Throughout the entire storage period, the populations of L. monocytogenes at any level of inoculated spoilage microflora rarely differed by more than 0.5 log10 per cm2, and the maximum observed difference as 1.1 log10 per cm2 at 40 d. The growth rate of L. monocytogenes was approximately the same at all levels of the inoculated spoilage microflora. These results suggest that the level of spoilage microflora present on the on the original processed meat may have a slight impact on the growth of L. monocytogenes in the package.
<BR><BR>
IMPACT: 2005/01/01 TO 2005/12/31<BR>
This project is intended to advance the knowledge of food microbiology with particular emphasis on food safety. The safety of foods of animal origins are emphasized, although any food may be included as necessary and appropriate.
<BR><BR>
PROGRESS: 2004/01/01 TO 2004/12/31<BR>
Five individual strains of Listeria monocytogenes and a mixed cocktail of all five were studied for attachment on frankfurters, ham, bologna, and roast beef relative to their cell surface characteristics. The ratio of strongly attached (sessile) L. monocytogenes cells compared to total (sessile and planktonic) attached cells on RTE meats was also determined. Since bacterial cell surfaces were characterized by net negative charge and hydrophobicity, electrostatic interaction chromatography (ESIC) and cationized ferritin (CF) were methods chosen to study net negative charge distribution on the bacterial cell surface while hydrophobic interaction chromatography (HIC) and contact angle measurement (CAM) were used to examine the cell surface hydrophobicity. No differences (P > 0.05) were observed in cell surface charge or cell surface hydrophobicity among strains. Approximately 84 to 87% L. monocytogenes were found to attach strongly to RTE meats within 5 min. No differences (P > 0.05) were found among strains or among meats. Micrographs observed from scanning electron microscope (SEM) showed no difference among the strains but a difference in age of cells (mixed culture) in terms of surface negative charge distribution. Since no difference in strains was observed, the CAM was carried out using the five-strain mixed culture. The surface hydrophobicity increased in frankfurters, decreased in roast beef, while it was unchanged in ham and bologna, due to inoculation. Survival of a desiccated five-strain Listeria monocytogenes mixture during storage in sand at 4C for 2 months was determined using the acridine orange direct count (AODC) method using novobiocin and plate count. Samples of inoculated sand were taken every two weeks. Samples were incubated at 37C for 6 h, stained with acridine orange, and then examined in the fluorescent microscope. Elongated viable but nonculturable (VBNC) cells were most frequently observed during weeks 2 and 4. At weeks 6 and 8, most of the cells either remained viable or were dead. In each microscopic field, only one or two VBNC cell was observed among hundreds of other viable culturable cells, indicating that L. monocytogenes does not generally become VBNC. Therefore, VBNC cells are not a concern when plating environmental samples or desiccated L. monocytogenes cells on nonselective media. The effects of aerobic or anaerobic incubation and media supplementation with 0.1% or 1% (wt/vol) sodium pyruvate were tested to optimize recovery of desiccated cells. Nonselective media showed better recovery when TSAYE and Columbia agar contained 0.1% (wt/vol) pyruvate and were incubated aerobically. These two culture methods were comparable (P > 0.05) in recovering desiccated L. monocytogenes cells.
<BR><BR>
IMPACT: 2004/01/01 TO 2004/12/31<BR>
This project has demonstrated that Listeria monocytogenes do not survive prolonged periods of drying in a state which is more difficult to detect by conventional detection methods (viable but non-culturable). This demonstrates that conventional microbiologicalmethods are adequate to detect Listeria in food processing environments.

<BR><BR>

PROGRESS: 2003/01/01 TO 2003/12/31<BR>
Survival of a desiccated five-strain Listeria monocytogenes mixture during storage in sand at 4C for 2 months was determined using the acridine orange direct count (AODC) method using novobiocin and plate count. Samples of inoculated sand were taken every two weeks. Samples were incubated at 37C for 6 h, stained with acridine orange, and then examined in the fluorescent microscope. Elongated viable but nonculturable (VBNC) cells were most frequently observed during weeks 2 and 4. At weeks 6 and 8, most of the cells either remained viable or were dead. In each microscopic field, only one or two VBNC cell was observed among hundreds of other viable culturable cells, indicating that L. monocytogenes does not generally become VBNC. Therefore, VBNC cells are not a concern when plating environmental samples or desiccated L. monocytogenes cells on nonselective media. Tryptic soy agar with 0.6% (wt/vol) yeast extract (TSAYE) and Columbia agar were used as nonselective plate count media. Modified Oxford (MOX) agar and TSAYE + 5% (wt/vol) sodium chloride were used as the selective plate count media. The effects of aerobic or anaerobic incubation and media supplementation with 0.1% or 1% (wt/vol) sodium pyruvate were tested to optimize recovery of desiccated cells. Nonselective media showed better recovery when TSAYE and Columbia agar contained 0.1% (wt/vol) pyruvate and were incubated aerobically. These two culture methods were comparable (P > 0.05) in recovering desiccated L. monocytogenes cells.
<BR><BR>
IMPACT: 2003/01/01 TO 2003/12/31<BR>
This project has demonstrated that Listeria monocytogenes do not survive prolonged periods of drying in a state which is more difficult to detect by conventional detection methods (viable but non-culturable). This demonstrates that conventional microbiologicalmethods are adequate to detect Listeria in food processing environments.
<BR><BR>
PROGRESS: 2002/01/01 TO 2002/12/31<BR>
Cell surface attachment is characterized by net negative charge and hydrophobicity. Electrostatic interaction chromatography (ESIC) and cationized ferritin (CF) are methods chosen to study net negative charge distribution on the bacterial cell surface. Hydrophobic interaction chromatography (HIC) and contact angle measurement (CAM) were used to examine the cell surface hydrophobicity. The ratio of strongly attached (sessile) L. monocytogenes cells compared to total (sessile and planktonic) attached cells on RTE meats was also be determined. Five individual strains of Listeria monocytogenes and a mixed cocktail of all five were studied on frankfurters, ham, bologna, and roast beef. No evidence of a difference (P>0.05) was observed in cell surface charge and cell surface hydrophobicity among strains. Approximately 84 to 87% L. monocytogenes was found to attach strongly to RTE meats in 5 min. No evidence of a difference (P>0.05) was found among strains and among meats. Micrographs from scanning electron microscope (SEM) showed no visual difference among the strains but a difference in age of cells (mixed culture) in terms of surface negative charge distribution. More negative sites were observed at day 0 and day 7 and much fewer at day 3, indicating possibly a change in cell surface properties. Since no difference in strains was established, the CAM was carried out using the five-strain mixed culture. Inoculated (108 CFU/ml) and non-inoculated RTE meats with L. monocytogenes showed evidence of a significant difference (P<0.05). Survival of a five-strain Listeria monocytogenes culture, including serotype 4b, during starvation in sand at 4,aC for 2 months was determined using the acridine orange direct count (AODC) method and plate count. Sand samples inoculated with L. monocytogenes were taken every two weeks. The inoculated sand was then subjected to a 6-h incubation at 37,aC and fixed with acridine orange. Elongated VBNC cells were observed more often during weeks 2 and 4 under a fluorescent microscope. At weeks 6 and 8, some of the cells either remained viable or dead. In each microscopic field, only one or two VBNC cells were observed with hundreds of other viable culturable cells, indicating that L. monocytogenes does not readily become VBNC; the bacterium is either viable and culturable or dead. Therefore, when plating environmental samples or stressed L. monocytogenes cells on non-selective media, VBNC cells are not a significant concern. Media used for plate count included tryptic soy agar with 0.6% yeast extract (TSBYE) and Columbia agar as non-selective media and Modified Oxford agar (MOX) as the selective medium. Variations within these media included aerobic or anaerobic incubation, with 0.1% or 1% sodium pyruvate to determine the best condition to recover the highest number of injured cells. Results from the non-selective media showed better recovery observed on TSBYE incubated aerobically and with 0.1% pyruvate and Columbia agar incubated aerobically and with 0.1% pyruvate. However, these methods were statistically indistinguishable from each other (P>0.05).
<BR><BR>
IMPACT: 2002/01/01 TO 2002/12/31<BR>
This project focuses on the control of bacterial pathogens of human health significance in foods of animal origins. It examines the basis for contamination and the potential for growth of these pathogens in these foods. It also determines optimum control and decontamination strategies for these pathogens. It is expected that, in the long term, this research will impact the levels of bacterial pathogens in human foods.

<BR><BR>
PROGRESS: 2001/01/01 TO 2001/12/31<BR>
Lactobacilli organisms were discovered with probiotic effect inhibitory towards Salmonella when orally inoculated into pigs in some but not all experiments (see detailed progress report attached). The Lactobacilli used as probiotics did not stimulate the immune response in the pig as had been reported to occur in the mouse. Thus, attempts to utilize Lactobacilli as carriers of Salmonella antigens were not deemed worthwhile. During the last 6 months, a different approach was taken to establish a model in gnotobiotic pigs for the discovery of other probiotic organisms as well as to study the rapid extraintestinal dissemination (RED) of Salmonella. Preliminary experiments have been conducted which indicate that germ-free pig can be used to reproduce RED caused by Salmonella. In addition, a cocktail of additional microbes (Lactobacillus, Streptococcus, and Bacterioides) is inhibitory to RED of Salmonella in the gnotobiotic pig model. Listeria monocytogenes cultures, including serotype 4b, were inoculated onto four different types of ready-to-eat (RTE) meats, frankfurters, ham, roast beef, and bologna. The meats were vacuum-packed and stored at 40C for 24 h prior to irradiation (targeted levels ranging from 0.5 to 4 kGy). Populations of L. monocytogenes were recovered by surface plating on non-selective and selective media. The margins of safety studied include 3D and 5D kill to achieve an optimal level of reduction while retaining organoleptic qualities of the meats. A 3-log reduction of L. monocytogenes on the non-selective medium was obtained at 1.5 kGy. On selective medium, the dosages were 1.5 kGy for bologna and roast beef and 2.0 kGy for frankfurters and ham. The D-values ranged from 0.42 to 0.44 kGy. A 5-log reduction of L. monocytogenes on the non-selective medium was obtained at 2.5 kGy. On selective medium, the dosages were 2.5 kGy for bologna and roast beef and 3.0 kGy for frankfurters and ham. Survival of Listeria monocytogenes in the same RTE meat types after irradiation was also studied. The meats were irradiated at doses of 2.0 and 4.0 kGy. Recovery of the surviving organisms was observed during storage at temperatures of 40C and 100C for 12 weeks. Preliminary results showed no growth in meats irradiated at 4.0 kGy. Survivors were observed for irradiated meats at 2.0 kGy stored at 100C after the second week. No growth was observed in samples irradiated at 2.0 kGy stored at 40C until the fifth week.
<BR><BR>
IMPACT: 2001/01/01 TO 2001/12/31<BR>
The outcomes of the previous years' research have addressed biological hazards in both the animal production phase and processing phase of foods of animal origins. Research within this Plan of Work has led to knowledge which may provide a new method for control of foodborne pathogens, specifically salmonellae, at the animal production site. The use of specific lactobacilli in feed could be a readily available user friendly technique for animal production sites. Additional research has documented the survival of a serious food borne pathogen, Listeria monocytogenes, in a food processing plant environment. A more comprehensive understanding of the contamination issue will lead to research on more successful interventions.