Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

Legacy Department


Committee Chair/Advisor

Jiang, Xiuping

Committee Member

Dawson , Paul L

Committee Member

Henson , John M

Committee Member

Tzeng , Jeremy


Although the rendering process serves as invaluable means for the disposal of inedible animal by-products, the finished products often harbor pathogenic and opportunistic microorganisms such as Salmonella and enterococci, respectively. The temperatures used during the rendering process far exceed the heat tolerance threshold of most bacterial species, so cross-contamination from the environment and/or from the incoming raw material is the proposed source of the contamination. Research has demonstrated that the raw material coming into the rendering facility is highly contaminated with pathogenic bacteria including Salmonella. While not in a rendering facility, studies have also demonstrated that bacteria such as Salmonella can persist on food processing equipment and be transferred into the product upon contact. The objectives of this study were to: 1) isolate and characterize Salmonella and enterococci from finished animal by-products, 2) produce and optimize a bacteriophage cocktail against Salmonella, 3) apply the bacteriophages to reduce Salmonella levels on environmental surfaces found in a rendering facility and in raw offal, and 4) use the bacteriophage as a feed additive to reduce or prevent Salmonella infection in mice.
To determine the prevalence of Salmonella and enterococci, two hundred finished meals provided by various rendering facilities across the U.S. were analyzed. While the animal meals were shown to not be a suitable environment for bacterial growth (moisture content 1.9 to 11.5%), these products did contain enterococci and Salmonella. Enterococci were detected in 83% of the samples and accounted for up to 54% of the total bacterial count, which ranged from 1.7 to 6.8 log CFU/g. Characterization of the enterococci isolates revealed that only 3 isolates were resistant to vancomycin (32 µg/ml). PCR analysis revealed that none of these 3 VRE isolates were E. faecalis or E. faecium. In addition, no VRE isolates were of the VanA or VanB-type, which confer the highest levels of resistance to vancomycin. Salmonella (n = ? ) was isolated from 8.7% of the finished meal samples. There were 13 Salmonella serotypes identified among the isolates with 16 pulse field gel electrophoresis (PFGE) patterns. Thermal tolerance studies revealed that these Salmonella isolates had D-values of 9.27-9.99, 2.07-2.28, and 0.35-0.40 min at 55°C, 60°C, and 65°C, respectively.
As a means to prevent or reduce Salmonella contamination, bacteriophages were isolated from raw chicken offal that would be used for rendering. Bacteriophages were isolated by using the Salmonella spp. isolated previously from the finished meals as the host bacteria. For further study, five of the isolated bacteriophages (n = ? ) were selected to produce a cocktail for bacteriophage treatment studies. The selection was based on which bacteriophage had the highest lytic activity against 5 pre-determined Salmonella serotypes, i.e., Enteritidis, Idikan, Johannesburg, Mbandaka, and Typhimurium. Prior to bacteriophage treatment studies, the optimal bacteriophage cocktail concentration was determined by multiplicity of infection (MOI) optimization. Initial results indicated that an actively growing culture is needed for lytic activity of bacteriophages. When using an actively growing Salmonella cocktail, the effectiveness of the bacteriophage cocktail was shown to increase by raising the MOI from 1 to 10, whereas MOI of 50 didn't enhance the lytic activity further. MOI optimization also revealed that resistance strains of the Salmonella spp. are selected for quickly (12 h), but that the lytic activity of the bacteriophage treatment is easily extended through the addition of a different bacteriophage cocktail at the 12 h mark.
The optimized bacteriophage cocktail was able to successfully reduce Salmonella levels on all tested environmental surfaces (HDPE plastic, cement, rubber, stainless steel). Treatment of Salmonella cells attached to the environmental surfaces resulted in a 2 log CFU/cm2 reduction at 40° and 30°C, and ca. 1 log CFU/cm2 reduction at 20°C on all surface materials. The presence of an organic layer on the surface had identical levels of reduction, indicating the organic material does not interfere with the bacteriophage's lytic activity. Treatment of the single species biofilm resulted in ca. 1-2, 2-3, and 1 log CFU/cm2 reduction in S. Enteritidis H4717 populations on all surface materials at 20, 30, and 40°C, respectively, as compared with ca. 0.5, 1.5-2, and 0.5 log CFU/cm2 reduction in S. Enteritidis populations of the double species biofilm under the same experimental condition.
In addition to the surface materials, the bacteriophage cocktail was shown to reduce Salmonella levels in raw chicken offal. When treating irradiated raw offal that was artificially contaminated with the Salmonella cocktail (105 CFU/g), the bacteriophage cocktail reduced Salmonella levels by ca. 2.0, 2.7, and 2.5 log CFU/g at 20°, 30°, and 40°C, respectively. The bacteriophage was also capable of reducing Salmonella levels to the same degree, i.e., ca. 2.0, 2.2, and 2.2 log CFU/g at 20°, 30°, and 40°C, respectively, in the non-sterile raw chicken offal at the presence of background microorganisms.
The bacteriophage cocktail was also evaluated as a feed additive. The bacteriophage cocktail was lyophilized and added into the animal meals (blood, feather, and poultry), which were artificially contaminated with a five-strain Salmonella cocktail at initial level of ca. 105 CFU/g. A series of dehydration studies revealed that the addition of dehydrated bacteriophages to finished meals does not reduce the level of Salmonella present upon rehydration; however, there was an observable difference between those samples containing the bacteriophages and those that did not after 12 h of rehydration, with those containing the bacteriophages having lower levels of Salmonella (ca. 2 log CFU/g difference ??). Our results also revealed that the bacteriophage cocktail's stability was reduced quickly when applied in a dehydrated form. The concentration of the dehydrated bacteriophage decreased by 1.5 log PFU/g within the rendered meal over a 4 week period at 30°C. However, the stability of the bacteriophage was maintained well when the bacteriophage was added to the animal feed in liquid form and stored under refrigeration conditions (4°C). Under these conditions, the bacteriophage cocktail's concentration decreased by 0.23 log PFU/g over a 4 week period.
The liquid bacteriophage was then supplemented into animal feed and given to mice during an animal trial. The mice that had been given a diet containing the bacteriophage for a period of a week prior to Salmonella inoculation were not infected as evidenced through fecal sampling. These mice shed no Salmonella in their feces over a 4 week period. Mice that had not been given the diet supplemented with the bacteriophage shed Salmonella in their feces for a period of 2.5 weeks. Histological analysis of the liver and intestine also indicated no observable signs of inflammation in those mice given the bacteriophages. In mice not receiving the bacteriophage treatment, venous dilation, cholangiohepatitis, and monocytes in the portal areas were observed in the liver.
Our results indicate that the Salmonella contamination of finished rendered meals is likely the source of cross-contamination between the environment or the incoming raw material and the finished products. The risk for cross-contamination may be reduced through the use of bacteriophage treatment which was found to reduce ca. 99 to 99.9% of Salmonella levels on both the environmental surfaces and raw material. The bacteriophage cocktail was also found to have the potential to be used as a feed additive to reduce pathogen levels within an animal host when added to the feed in liquid form just prior to consumption.

Included in

Microbiology Commons



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.