Foodborne Pathogen Persistence in the Food Processing Environment

FOODBORNE PATHOGEN PERSISTENCE IN THE FOOD PROCESSING ENVIRONMENT

Alex L. Brandt, M.S.

A Dissertation

ANIMAL SCIENCE

Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

Approved

Kendra K. Nightingale, Ph.D. Chair of Committee

Mindy M. Brashears, Ph.D.

J. Chance Brooks, Ph.D.

Margarget D. Hardin, Ph.D.

Haley F. Oliver, Ph.D.

Mark Sheridan, Ph.D. Dean of the Graduate School

May, 2014

Texas Tech University, Alex Brandt, May 2014

ACKNOWLEDGEMENTS There are countless people who have been supportive of me in my endeavor to obtain my Ph.D., and I owe them all a great deal of gratitude. Mentioning all of them here would make up an entire dissertation in itself, so even though all are not mentioned by name, they still hold a special place in my heart. First and foremost, I thank my parents, who have always been by my side in both times of joyous accomplishment and in times when I was down and desperately needed someone to talk to. The love that constantly flows from their open hearts, and the concern that is constantly in their open ears, are gifts that I can only hope to give to others in my life. Also, the lessons they taught me about working hard, and carrying on in the midst of life’s trials, are something I will always carry in my heart and mind. My extended family, including all of my grandparents, aunts, uncles, and cousins have also been supportive of all my aspirations, and I also will be forever grateful to them for all the unwavering support and encouragement they have always given me. I also thank my dear girlfriend Samantha whose prayers, understanding, and love were sometimes the only things keeping me afloat through this process. I could not have carried on if not for her encouragement, love, and willingness to be there for me. I also have been blessed with many mentors who have helped me through this journey to whom I owe much gratitude. Dr. Kendra Nightingale, my Ph.D. advisor, is at the top of this list. She has been a caring mentor and wonderful teacher to me, and I thank her immensely for giving me this opportunity to grow professionally and to acquire at least a small portion of her wealth of knowledge. I also thank my committee members Dr. Margaret Hardin, Dr. Haley Oliver, Dr. Mindy Brashears, and Dr. Chance Brooks for all of their mentorship, intellectual input, and moral support over the past several years. I would also be remiss if I did not thank my past advisors, teachers, and colleagues Dr. Matthew Taylor, Mrs. Liz Treptow, Mrs. Nancy Jurecka, Dr. Jenna Anding, and Dr. Jay Neal for always believing in me and for aiding in my educational and professional formation. I also thank Drs. Dale Woerner, Martin Wiedmann, and Guy Loneragan for all of their help enrolling facilities, providing project guidance, and answering the occasional methodological and statistical questions that arose.

Texas Tech University, Alex Brandt, May 2014

I also would be remiss if I did not thank all of my colleagues, friends, and labmates who were always willing to jump in and lend a helping hand. I especially thank Anna Van Stelten, Eva Borjas, Chip Manuel, Jessica Chen, Alex Tudor, Alma Perez- Mendez, Ashleigh Willems, Jake Elder, Gina Geornaras, Curtis Pittman, Tanya Jackson, Logan Jackson, Kathy Fermin, Miles Harris, Samantha Stewart, Peter Cook, Jessica Heiden, Sarah Navratil, Sara Gragg, Amy Parks, Jessie Vipham, Byron Chaves, Diana Ayala, Lacey Guillen, Marie Bugarel, Josh and Sarah Ison, Devin Hanson, Hattie Webb, Katelyn Malin, and Jennifer Martin for their constant support, willingness to help, and uplifting and stimulating discussions throughout the completion of my degree. Also, thanks to all the other unnamed graduate students, undergraduate students, and laboratory staff who always found time to be there when I needed it. I also thank all the facility managers and employees for their cooperation to accomplish this study. It literally could have not have been done without their willingness to participate, their flexibility, and heed of our thoughts, concerns, and suggestions. I forged relationships with them that helped me grow extensively from a professional standpoint, and I have no doubt that those relationships will last a lifetime. I also thank the United States Department of Agriculture National Institute of Food and Agriculture for providing the funding for this important study. I also thank all those at Cornell University who were so willing to help with subtyping, analysis, and interpretation of results. Sherry Roof, Emily Wright, Laura Strawn, Lorraine Rodriguez-Rivera, Tom Malley, and Matthew Stasiewicz were always willing to go the extra mile, and I am truly grateful for all of their support. Lastly, and most importantly, I thank God for His constant love and the blessings He has bestowed on me, which I am most unworthy of. I am so thankful for where He has taken me in my life. Though I count my blessings every day, I still cannot ever give Him all the gratitude He deserves. Without His comforting hand, I would not have made it through some of the trials I faced over the past few years. Thus, I am forever indebted to Him. Ad Majorem Dei Gloriam.

iii

Texas Tech University, Alex Brandt, May 2014

TABLE OF CONTENTS ACKNOWLEDGEMENTS ...... ii ABSTRACT ...... viii LIST OF TABLES ...... xi LIST OF FIGURES ...... xiii I. REVIEW OF THE LITERATURE ...... 1 Foodborne Illness ...... 1 Listeria monocytogenes ...... 3 Salmonella enterica ...... 9 Escherichia coli O157:H7...... 15 Bacterial Subtyping ...... 20 Molecular Ecology and Persistence of Foodborne Pathogens ...... 24 References ...... 42 II. INTEGRATING LONGITUDINAL SAMPLING, MOLECULAR SUBTYPING, AND STATISTICAL ANALYSIS TO INVESTIGATE PERSISTENCE OF LISTERIA MONOCYTOGENES, SALMONELLA ENTERICA, AND ESCHERICHIA COLI O157:H7 IN MEAT PROCESSING FACILITY ENVIRONMENTS ...... 92 Introduction ...... 92 Materials and Methods ...... 93 Results ...... 103 Discussion ...... 110 Conclusions ...... 117 References ...... 118 III. ASSESSING FOODBORNE PATHOGEN PERSISTENCE MITIGATION IN MEAT PROCESSING ESTABLISHMENTS FOLLOWING ENACTMENT OF PHYSICAL AND PERSONNEL-BASED CONTROL MEASURES ...... 155 Introduction ...... 155 Materials and Methods ...... 156 Results ...... 162 Discussion ...... 171 Conclusions ...... 178

Texas Tech University, Alex Brandt, May 2014

References ...... 180

IV. CHARACTERIZING THE COMPOSITION OF BACTERIAL POPULATIONS AT SITES OF RECURRENT LISTERIA CONTAMINATION IN MEAT PROCESSING FACILITIES USING 16S RRNA GENE SEQUENCING ...... 217 Introduction ...... 217 Materials and Methods ...... 218 Results ...... 220 Discussion ...... 222 Conclusions ...... 224 References ...... 225 V. IMPLICATIONS OF FINDINGS FROM THIS STUDY AND FUTURE PERSPECTIVES IN FOODBORNE PATHOGEN PERSISTENCE RESEARCH...... 234 References ...... 240 APPENDICES ...... 243 Appendix A: Master Sample Site List for Initial Six-Month Sampling Period in Facility A ...... 243 Appendix B: Master Sample Site List for Initial Six-Month Sampling Period in Facility B ...... 244 Appendix C: Master Sample Site List for Initial Six-Month Sampling Period in Facility B ...... 245 Appendix D: Master Sample Site List for Initial Six-Month Sampling Period in Facility B ...... 246 Appendix E: Initial Six-Month Sampling Period Listeria Contamination Patterns in Facility A ...... 247 Appendix F: Initial Six-Month Sampling Period Listeria Contamination Patterns in Facility B ...... 249 Appendix G: Initial Six-Month Sampling Period Listeria Contamination Patterns in Facility C ...... 251 Appendix H: Initial Six-Month Sampling Period Listeria Contamination Patterns in Facility D ...... 253 Appendix I: Facility A Isolation Patterns for Listeria monocytogenes Ribotype DUP-1052A ...... 255 Appendix J: Facility A Isolation Patterns for Listeria innocua sigB Allelic Type 56 ...... 256 v

Texas Tech University, Alex Brandt, May 2014

Appendix K: Facility A Isolation Patterns for Listeria weslshimeri sigB Allelic Type 69 ...... 257 Appendix L: Facility B Isolation Patterns for Listeria monocytogenes Ribotype DUP-1053E ...... 258 Appendix M: Facility B Isolation Patterns for Listeria monocytogenes Ribotype DUP-1062B ...... 259 Appendix N: Facility B Isolation Patterns for Listeria monocytogenes Ribotype DUP-1062E ...... 260 Appendix O: Facility B Isolation Patterns for Listeria innocua sigB Allelic Type 6 ...... 261 Appendix P: Facility B Isolation Patterns for Listeria innocua sigB Allelic Type 11 ...... 262 Appendix Q: Facility B Isolation Patterns for Listeria innocua sigB Allelic Type 23 ...... 263 Appendix R: Facility B Isolation Patterns for Listeria innocua sigB Allelic Type 38 ...... 264 Appendix S: Facility B Isolation Patterns for Salmonella enterica Serotype-Pulsotype Heidelberg ...... 265 Appendix T: Facility C Isolation Patterns for Listeria monocytogenes Ribotype DUP-1042B ...... 266 Appendix U: Facility C Isolation Patterns for Listeria innocua sigB Allelic Type 70 ...... 267 Appendix V: Facility C Isolation Patterns for Listeria innocua sigB Allelic Type 71 ...... 268 Appendix W: Facility C Isolation Patterns for Salmonella enterica Serotype-Pulsotype III 50:R:Z ...... 269 Appendix X: Facility C Isolation Patterns for Salmonella enterica Serotype-Pulsotype Infantis ...... 270 Appendix Y: Facility C Isolation Patterns for Salmonella enterica Serotype-Pulsotype Uganda ...... 271 Appendix Z: Facility D Isolation Patterns for Listeria monocytogenes Ribotype DUP-1030B ...... 272 Appendix AA: Facility D Isolation Patterns for Listeria innocua sigB Allelic Type 6 ...... 273 Appendix AB: Facility D Isolation Patterns for Listeria welshimeri sigB Allelic Type 129 ...... 274

Texas Tech University, Alex Brandt, May 2014

Appendix AC: Facility C Isolation Patterns for Salmonella enterica Serotype-Pulsotype Typhimurium Var 5- ...... 275 Appendix AD: Enrollment Letter Mailed to Facilities at Study Outset...... 276 Appendix AE: Enrollment Form Mailed to Facilities at Study Outset ...... 277 Appendix AF: Knowledge Assessment Used at Pre- and Post- Training Time Points ...... 278 Appendix AG: Facility B 2012 Behavioral Assessment Questionnaire ...... 283 Appendix AH: Facility C 2012 Behavioral Assessment Questionnaire ...... 285 Appendix AI: Facility B 2012 Direct Observation Form ...... 287 Appendix AJ: Facility C 2012 Direct Observation Form ...... 289 Appendix AK: Facility B 2013 Behavioral Assessment Questionnaire ...... 291 Appendix AL: Facility C 2013 Behavioral Assessment Questionnaire...... 293 Appendix AM: Facility B 2013 Direct Observation Form ...... 295 Appendix AN: Facility C 2013 Direct Observation Form ...... 298

vii

Texas Tech University, Alex Brandt, May 2014

ABSTRACT Emerging evidence suggests that persistence of foodborne pathogens within food processing facilities represents a key factor in their introduction into the food chain, and that persistent strains have been associated with several multi-state foodborne illness outbreaks and expansive recalls. While Listeria monocytogenes ecology has been extensively documented in various food processing facilities, until recently, no standardized definitions of persistence had been developed and applied to these ecological data, and few studies had assessed the impact of control measures on Listeria persistence in the meat processing environment. Also, very little information has been generated on the ecology of Salmonella enterica and Escherichia coli O157:H7 in food processing facilities, and thus much remains to be understood about the extent to which they persist in these premises. Additionally, little is known about population compositions for other bacteria that are present in sites of recurrent Listeria contamination, some of which may directly influence the growth of Listeria at these sites. The first phase of this study investigated foodborne pathogen persistence in meat processing facilities using an integrated strategy of longitudinal sampling, molecular subtyping, and two statistical tests for persistence. Four meat processing facilities were enrolled, and environmental sites and food products were sampled and tested for Listeria, S. enterica, and E. coli O157:H7 for a six-month period. All isolates were subtyped using molecular and serological methods, and subtyping data were statistically analyzed. Across all facilities, six-month prevalences ranged from 1.21% to 16.97% for L. monocytogenes, 6.07% to 28.18% for non-pathogenic Listeria, 0.00% to 6.97% for S. enterica, and 0.00% to 1.21% for E. coli O157:H7. Sites in the drain and floor category had the highest prevalence for Listeria and S. enterica across all facilities. A binomial test for subtype frequency identified persistent strains of L. monocytogenes and non-pathogenic Listeria species in each facility, and persistent strains of S. enterica in three facilities. Another binomial test, based on previous isolation of a strain at a particular site as a risk factor for subsequent isolation of the same strain at the same site, identified a smaller number of persistent Listeria strains, and identified no persistent S. enterica strains. Nine and seven sampling sites that were persistently colonized with specific strains of L. monocytogenes and non-pathogenic Listeria, respectively, were identified via the binomial frequency test.

viii

Texas Tech University, Alex Brandt, May 2014

The second phase of the study assessed impacts of control measure implementation on Listeria persistence. In-plant trainings were delivered to personnel in all four facilities, and significant increases in knowledge were observed in three facilities and in the entire group of participants. Three facilities were enrolled in a follow-up intervention and sampling program for Listeria. Each facility implemented physical changes and interventions suggested during in-plant trainings, and implemented additional changes autonomously. Significant short and long-term personnel behavioral changes regarding use of facility- specific footwear and footwear sanitation were observed for one facility, while a significant change in hygienic item use (e.g. hairnets and gloves) was recorded for another. Persistent Listeria strains were still isolated in each facility during follow-up, but all three facilities had decreased mid-shift prevalences of Listeria between the initial and follow-up periods among a subset of sites that were i) positive for Listeria at least once during the initial period and ii) were followed for the entirety of the study; the decrease was significant for two facilities. Though reductions cannot be wholly attributed to a particular change, subsets of processing room sites of two facilities had significantly decreased Listeria prevalences after boot baths were installed. Directional sampling among a group of sites in the three facilities showed that contamination patterns at adjacent areas may differ from those at the original site. Also, sampling in a newly-constructed area of one facility demonstrated that persistent Listeria strains from older areas of the facility began to colonize newly-constructed areas. The last portion of the study focused on the population composition of culturable bacteria at sites of recurrent Listeria contamination. Nine sites from three facilities were sampled, and up to ten colonies with different morphologies on Brain Heart Infusion Agar were selected for characterization from each sample. DNA sequencing was peformed on a portion of the 16S rRNA gene of each isolate. Bacterial genera were assigned to each isolate using the Ribosomal Database Project SeqMatch utility. Pseudomonas predominated the isolates characterized from each facility, while Staphylococcus, Psychrobacter, Aeromonas and Enterobacteriaceae were also among the most predominant. Stenotrophomonas and Kocuria, both of which have been shown to benefit L. monocytogenes grwoth in co-culture, were isolated at certain sites. Lactococcus, Shewanella, Carnobacterium, and Serratia, which all have detrimental effects on L. monocytogenes, were also isolated. Overall, the results of this study help to fill several knowledge gaps regarding pathogen persistence in meat processing facilities especially with regard to the prevalence

Texas Tech University, Alex Brandt, May 2014 and extent of persistence of E. coli O157:H7 and S. enterica, the application of statistical measures to achieve standardized definitions of persistence, and the utility of trainings and control measure implementation to reduce prevalence of pathogens. This information has applicability for all meat processors, and will likely be of use in their pathogen control programs, but it is especially useful for those that are small and very small, have limited knowledge of pathogen persistence, and have few resources to combat it. Furthermore, this work also highlights several topics that remain to be addressed with regard to pathogen persistence, including persistence of S. enterica and E. coli O157:H7 in other food processing environments (especially those of low moisture foods), persistence of all pathogens at points downstream of the processing facility (i.e. at retail, in home kitchens, and in institution foodservice operations), the study of interactions between bacterial populations at sites of recurrent Listeria contamination, and characterization of persistent-state Listeria gene expression in sites of persistence.

Texas Tech University, Alex Brandt, May 2014

LIST OF TABLES 1.1 Summary of selected studies where L. monocytogenes persistence was inferred ...... 86 2.1 Polymerase chain reaction primer sequences and thermal cycling conditions ...... 127 2.2 Synopsis of pertinent characteristics for facilities enrolled in the study ...... 129 2.3 Prevalence of each organism by sampling site category in each facility ...... 130 2.4 Facility-level persistence metrics for L. monocytogenes, using a binomial test for ribotype frequency ...... 131 2.5 Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test for sigB AT frequency ...... 133 2.6 Metrics for facility sites demonstrating significant levels of L. monocytogenes persistence, using a binomial test for ribotype frequency...... 136

2.7 Metrics for facility sites demonstrating significant levels of non-pathogenic Listeria species persistence, using a binomial test for sigB AT frequency ...... 137 2.8 Facility-level persistence metrics for L. monocytogenes, using a binomial test based on previous positive results as a risk factor ...... 138 2.9 Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test based on previous positive results as a risk factor ...... 140 2.10 Facility-level persistence metrics for S. enterica, using a binomial test for serotype frequency ...... 143 2.11 Facility-level persistence metrics for S. enterica, using a binomial test based on previous positive results as a risk factor ...... 144 3.1 Dates of follow-up sampling occasions in Facilities A, B, and C ...... 184 3.2 Facility A physical change and intervention implementation ...... 185 3.3 Facility B physical change and intervention implementation ...... 186 3.4 Facility C physical change and intervention implementation ...... 188 3.5 Summary of self-reported behavior changes and direct observation of employees in Facility B ...... 190

Texas Tech University, Alex Brandt, May 2014

3.6 Summary of self-reported behavior changes and direct observation of employees in Facility C ...... 192 3.7 Initial and follow-up Listeria contamination patterns in Facility A ...... 194 3.8 Initial and follow-up Listeria contamination patterns in Facility B ...... 196 3.9 Initial and follow-up Listeria contamination patterns in Facility C ...... 199 3.10 Listeria contamination patterns in new smokehouse room and cooler area of Facility B ...... 202 3.11 Initial and follow-up period Listeria prevalence among certain site groups in each facility ...... 203 4.1 Descriptions and Listeria results for sampling sites included in this study ...... 230 4.2 Ribosomal Database Project identifications given to isolates collected ...... 231 4.3 Simpson’s Index of Diversity values for groups of isolates from facilities and specific sites ...... 233

xii

Texas Tech University, Alex Brandt, May 2014

LIST OF FIGURES 2.1 Sampling scheme map for Facility A...... 145 2.2 Sampling scheme map for Facility B ...... 146 2.3 Sampling scheme map for Facility C ...... 147 2.4 Sampling scheme map for Facility D...... 148 2.5 Prevalence of each organism by sampling date in ready- to-eat meat Facilities A (I) and B (II) ...... 149 2.6 Dendrogram of XbaI PFGE patterns for S. enterica in Facility B ...... 150 2.7 Prevalence of each organism by sampling date in fresh meat Facilities C (I) and D (II) ...... 151 2.8 Dendrogram of XbaI PFGE patterns for S. enterica in Facility C ...... 152 2.9 Dendrogram of XbaI PFGE patterns for E. coli O157:H7 in Facility C ...... 153 2.10 Dendrogram of XbaI PFGE patterns for S. enterica in Facility D ...... 154 3.1 Facility A sampling scheme for 2012 follow-up sampling occasions ...... 204 3.2 Facility B sampling scheme for 2012 follow-up sampling occasions ...... 205 3.3 Facility C sampling scheme for 2012 follow-up sampling occasions ...... 206 3.4 Facility A sampling scheme for 2013 follow-up sampling occasions ...... 207 3.5 Facility B sampling scheme for 2013 follow-up sampling occasions ...... 208 3.6 Facility C sampling scheme for 2013 follow-up sampling occasions ...... 209 3.7 Mean pre- and post-training knowledge assessment scores for all facilities ...... 210 3.8 Diagram of Listeria results from directional swabs taken from sites in Facility A ...... 211 3.9 Diagram of Listeria results from directional swabs taken from sites in Facility B ...... 213 3.10 Diagram of Listeria results from directional swabs taken from sites in Facility C ...... 216

xiii

Texas Tech University, Alex Brandt, May 2014

CHAPTER I REVIEW OF THE LITERATURE Foodborne Illness Current Public Health and Economic Burden Preserving the safety of the world’s food supply is an issue of the utmost public health and economic importance. Each year in the U.S., an estimated 47.8 million cases of foodborne illness occur, leading to an estimated 128,000 hospitalizations, 3037 deaths, and a $51.0 billion health-related economic burden (317, 319). The etiologies of these illnesses remain largely unspecified (80%), but a significant portion (20%) have known etiologic agents that belong to a group of 31 common foodborne pathogens (317). The bacterial pathogens Escherichia coli O157:H7, nontyphoidal Salmonella, and Listeria monocytogenes are included in this group. Collectively, these agents account for an estimated public health burden of 1.1 million annual cases of illness, nearly 23,000 hospitalizations, and 653 deaths (318). The substantial public health threat posed by these pathogens, their $7.1 billion annual economic impact (319), and the largely preventable nature of the illnesses associated with them, have prompted initiatives aimed at reducing their incidence in the U.S. population. Since the initial announcement of the Healthy People 2000 objectives in 1990, the U.S. Department of Health and Human Services (HHS) has included reductions in incidence of illness due to these pathogens among the primary objectives in all of its subsequent Healthy People campaigns (362). Establishing Goals and Tracking Progress The measures of disease incidence used by the Healthy People campaigns are largely based on data from the Foodborne Diseases Active Surveillance Network, or “FoodNet,” which is a collaborative effort between multiple U.S. agencies for surveillance of nine foodborne pathogens in areas within ten different states (386). Following the collection of the inaugural 1996-1998 FoodNet baseline data, progress was noted for L. monocytogenes and E. coli O157:H7. Incidence of illness due to L. monocytogenes sharply decreased from 0.53 per 100,000 person-years in 1998 to 0.25 per 100,000 person-years in 2002 (386). Likewise, incidence of E. coli O157:H7 illness decreased from 2.37 per 100,000 person-years in 1998 to 0.91 per 100,000 person-years

Texas Tech University, Alex Brandt, May 2014 in 2004 (386). However, since then, the trends in incidence of illness due to these pathogens have leveled off and did not meet the Healthy People 2010 incidence of illness goals of 0.24 and 1.0 per 100,000 person-years for L. monocytogenes and E. coli O157:H7, respectively, nor are they currently near the Healthy People 2020 goals of 0.20 and 0.60 per 100,000 person-years (362). The incidence of illnesses due to Salmonella has not seen the same progress as the two mentioned above, and has actually increased from the baseline values of 1996-1998. In 1998, the incidence was 13.61 per 100,000 person-years and in 2012 it was to 16.42 per 100,000 person-years (386). These values did not meet the Healthy People 2010 incidence objective of 6.40 per 100,000 person- years and are still not aligned with the Healthy People 2020 goal of 11.40 per 100,000 person-years (362). Food Product Recall Burden Not only do these three pathogens have detrimental public health impacts in terms of morbidity, mortality, and economics, but they also present significant economic burdens for food producers. Of the 1395 recall events recorded by the U.S. Food and Drug Administration (FDA) from 2003 to 2011, over 90% were tied to either L. monocytogenes, Salmonella, or E. coli O157:H7 (85). Also, for the 441 recalls of food products overseen by the U.S. Department of Agriculture Food Safety and Inspection Service (USDA-FSIS) from 2005 to 2011, nearly 47% were associated with at least one of these three pathogens (361). Because recall costs can be large, these figures provide an indication of the extensive economic burden created by these pathogens that has to be absorbed by the food industry. Thus, industry entities have a vested interest in mitigating these pathogens not only for the purposes of protecting public health, but also for avoiding the financial impact of a recall of product which they produce and distribute. Moving Forward Because these pathogens have foodborne transmission rates of 99%, 94%, and 68% (for L. monocytogenes, Salmonella enterica, and E. coli O157:H7, respectively) more attention needs to be devoted to developing strategies that mitigate these pathogens in the food chain if the current set of Healthy People objectives are to be met by the year 2020 (318). Consequently, research aimed at means to control the presence of these

Texas Tech University, Alex Brandt, May 2014 pathogens at all stages of food production from farm-to-fork is indispensable, and is currently a focus area that is of key priority to many grant funding agencies.

Listeria monocytogenes Classification and Reservoirs L. monocytogenes is a Gram positive, facultative anaerobic, non-sporeforming bacterium which belongs to the Listeriaceae family within the division Firmicutes (160). Presently, the Listeria genus is comprised of four clades with different species that comprise each clade (83). The Listeria sensu stricto clade consists of L. monocytogenes, Listeria marthii, Listeria innocua, Listeria welshimeri, Listeria seeligeri, and Listeria ivanovii. A second clade consists of Listeria fleischmannii and two newly-described species, Listeria aquatica sp. nov. and Listeria floridensis sp. nov. A third clade consists of Listeria rocourtiae, and Listeria weihenstephanensis, along with three new species, Listeria cornellensis sp. nov., L. grandensis sp. nov., and Listeria riparia sp. nov. A fourth clade consists solely of Listeria grayi (83) Although illnesses caused by L. innocua, L. seeligeri, and L. grayi, have been reported in rare occasions (285, 295, 302), only two species, L. monocytogenes and L. ivanovii, are considered true mammalian pathogens (276). L. ivanovii is more rare than L. monocytogenes and predominantly causes disease in ruminants (276). L. monocytogenes, however, is an important human pathogen of public health importance. Though it was first isolated from diseased rabbits in 1926 (256), it was later discovered that L. monocytogenes colonizes an array of domestic and wild animals (255) while also existing as a saprophyte in soil, decaying vegetation, and animal excrement (412). Morphology and Growth Characteristics Morphologically, L. monocytogenes cells appear as short rods with rounded ends that range from 1-2 µm in length and 0.5 µm in width (160). It is catalase positive, oxidase negative, and is able to grow between the temperatures -0.4 and 50°C (174). At temperatures in the range of 20 to 25°C the organism possesses peritrichous flagella, which lend it its distinctive tumbling motility. However, at temperatures near 37°C flagellin production is markedly decreased, and motility is limited (284). This is the consequence of temperature-dependent repression of flagellar gene expression by the

Texas Tech University, Alex Brandt, May 2014 transcriptional regulator mogR (137). When grown in tryptic soy broth with 0.6% yeast extract at 30°C the pH range permissible for growth has been shown to extend from 4.5 to 7.0, with minimal growth below pH 4.0 (281). Though the organism grows optimally at water activity (aw) values >0.97, it is considered to be fairly osmotolerant, can continue to multiply at NaCl concentrations as high as 10-12%, and will survive for extended periods of time at aw values as low as 0.83 (326). Serotypes and Lineages Strains of the bacterium can be sub-classified into several groups beyond the species level. At present, 13 different serotypes of the organism are recognized (323). However, the majority of human illnesses are attributed to serotypes 1/2a, 1/2b, and 4b (240). Serotype 4b is most frequently associated with illness worldwide, though illnesses due to serotype 1/2a occasionally outnumber those due to 4b in some regions (220). The species is also currently divided into four distinct genetic lineages (276). Lineage I contains strains belonging to serotypes 1/2b, 3b, 3c, and 4b, and is overrepresented among human isolates (415). Lineage II is comprised of strains from serotypes 1/2a, 1/2c, and 3a, and is overrepresented among isolates from food and food environments, as well as isolates from natural environments (415). Lineages III and IV are considerably more rare, contain strains of serotypes 4a, 4b, and 4c, and are predominantly isolated from ruminants (415). Serotype 4b spans several genetic lineages, but it has been revealed that strains that belong to the different lineages possess distinct molecular features (217). Epidemic Clones and Associated Outbreaks Molecular typing evidence from several early studies of L. monocytogenes strains associated with foodborne listerosis outbreaks demonstrated that genetic similarities were present among the outbreak strains even though they were geographically and temporally distinct. Comparing the isolates from these studies, Kathariou proposed the existence of several groups of epidemic-associated strains termed epidemic clones (ECs) (182). Four groups, ECI, ECIa, ECII, and ECIII, were initially proposed (182). Multi- virulence-locus sequence typing (MVLST) later confirmed the genetic relatedness of groups ECI, ECII, and ECIII, but suggested that ECIa be reclassified as ECIV and exclude isolates from a Boston outbreak involving pasteurized milk (62, 107). More 4

Texas Tech University, Alex Brandt, May 2014 recently, groups ECV, ECVI, and ECVII have been added (191, 218). At present, ECI describes serotype 4b isolates from outbreaks associated with coleslaw in Nova Scotia in 1981 (320), soft cheese in Switzerland from 1983 to 1987 (229), Mexican-style soft cheese in California in 1985 (216, 363), and pork tongue in France in 1992 (165). ECII includes serotype 4b isolates that caused multi-state U.S. outbreaks of listeriosis associated with frankfurters from 1998 to 1999 (241, 368) and turkey deli meat in 2002 (129, 373). ECIII is comprised of serotype 1/2a isolates from a sporadic case of listeriosis in Oklahoma in 1988 that was associated with turkey franks produced at a facility in Texas (364). ECIII also includes isolates from a multi-state outbreak in 2000 that was associated with turkey deli meat produced in the same Texas facility and sickened 29 people in 11 states (274, 372). ECIV includes serotype 4b isolates from a U.K. outbreak in 1988 associated with pâté (253), a Massachusetts outbreak involving vegetables in 1983, and a 1997 outbreak in Italy that was associated with contaminated corn salad (15). ECV is a more recently described EC (191), and consists of serotype 1/2a strains that were responsible for an outbreak of listeriosis linked to ready-to-eat (RTE) deli meats produced by a Canadian firm in 2008 (99), as well as a noteworthy amount of sporadic cases in Canada possibly reaching as far back as 1988 (191). ECVI is a newly- designated EC (218), and includes serotype 1/2b isolates from the 2011 listeriosis outbreak associated with cantaloupes produced in Colorado that sickened 147 people in 28 states, killing 33 (237). Also included in ECVI are isolates from a 1996 Canadian outbreak involving imitation crab meat (103). ECVII includes the serotype 1/2a isolates tied to the Colorado cantaloupe outbreak (237), as well as isolates from an outbreak involving whipping cream in Canada in 2000 (279). Outside of EC-associated outbreaks, L. monocytogenes has also been the etiologic agent for U.S. outbreaks associated with chocolate milk in Illinois in 1994 (78), Mexican-style cheese in North Carolina in 2007 (227), tuna salad in New York in 2008 (68), and diced celery in Texas in 2010 (117). Public Health Burden and High Risk Foods As evidenced by the vehicles implicated in the listeriosis outbreaks included above, almost all listeriosis cases (99%) are attributed to foodborne transmission (318). The HHS Centers for Disease Control and Prevention (CDC) estimates that annual numbers of L. monocytogenes-related illnesses, hospitalizations, and deaths are 1591, 5

Texas Tech University, Alex Brandt, May 2014

1455, and 255, respectively (318). The organism is also responsible for a $2.03 billion economic burden of illness, prompted 236 Class I recalls among FDA-regulated foods in the U.S. from 2003 to 2011 (85), and was the reason for 114 recalls of USDA-FSIS regulated products from 2005 to 2013 (361). A joint L. monocytogenes risk assessment completed by FDA and USDA-FSIS in 2003 listed several foods as “high risk” for listeriosis on a per serving basis (390). Included in this group are deli meats, non-reheated frankfurters, pâté and meat spreads, unpasteurized fluid milk, smoked seafood, and cooked RTE crustaceans (390). These foods are commonly handled after receiving a lethality heat treatment and before packaging (3), which may result in cross- contamination of the product from handlers or the post-lethality processing environment. With the ability of L. monocytogenes to grow at refrigeration temperatures (174) and the favorable nutrients, pH, and moisture content inherent to proteinaceous foods (126), cross-contamination of RTE products can result in consumption of a high number of bacterial cells by the consumer when the product is eaten. Furthermore, these products are commonly vacuum-packaged, and the low O2 environment that is created by vaccum- packaging provides ideal conditions for growth of the microorganism. Consequently, control of the pathogen in the food processing environment is of great importance (351). Current Regulations and Control Measures FDA maintains a strict zero-tolerance policy for detection of the pathogen in RTE foods under its jurisdiction (327) based on the provisions of Sections 402(a) (1) and 402(a) (4) of the Federal Food, Drug, and Cosmetic Act (FFDCA). USDA-FSIS, which regulates meat and poultry product safety in the U.S., also maintains a zero-tolerance policy for detecting the organism in RTE meat and poultry products under provisions of the Federal Meat Inspection Act and the Poultry Products Inspection Act (356). In 2003 FSIS finalized its Listeria Control Rule, which required facilities producing post-lethality exposed RTE meat and poultry products to establish control measures for the pathogen (356). Products produced under these conditions must receive controls for the pathogen in the form of Alternative 1 (use of a post-lethality treatment (PLT) to reduce or eliminate L. monocytogenes and an antimicrobial agent or process (AMA/AMP) to suppress or limit its growth), Alternative 2 (use of a PLT or AMA/AMP for control), or Alternative 3 (no PLT or AMA/AMP is used; control of the pathogen is 6

Texas Tech University, Alex Brandt, May 2014 achieved by sanitation only) (360). Establishments producing products under Alternative 2b (AMA/AMP only) or Alternative 3 are also required to conduct regular testing of food contact surfaces as part of a Listeria Control Program to verify environmental sanitation as defined in 9 CFR 430.4(b)(2)(iii)(A) and (3)(i)(A). To ensure compliance with the rule, FSIS conducts regular food safety assessments to determine if controls meet FSIS expectations, and the agency performs routine testing for the pathogen via its ALLRTE, RTE001, and RLm sampling programs (360). Listeriosis and High Risk Individuals In addition to describing foods of high risk for transmitting listeriosis, the 2003 FDA-FSIS risk assessment also describes populations that are at increased risk of contracting L. monocytogenes infections. Though the assessment primarily focuses on two subpopulations, the elderly and the perinatal, it also notes that anyone with a compromised immune system may be considered highly susceptible to L. monocytogenes infection (390). This includes groups such as cancer patients, transplant patients, and those with immunosuppressive diseases such as human immunodeficiency virus (257). Human perinatal infections usually involve mild influenza-like symptoms as well as diarrhea and lower back pain. However, more deleterious effects can include premature labor, stillbirth, and spontaneous abortion of the fetus (258). In utero infection of the neonate can result in production of lesions on internal organs (18), fever, respiratory distress, and neurological deformities. Listeriosis cases not associated with pregnancies are usually typified by sepsis, meningitis, and meningoencephalitis, and can result in gross lack of muscular coordination, seizures, and altered mental status (403). Classical foodborne illness symptoms such as febrile gastroenteritis, diarrhea, and vomiting have been reported in certain outbreaks (15, 78), but are less common. Though studies using non-human primate models have demonstrated that the infectious dose required for severe symptoms of listeriosis may be relatively high (106- 1010 CFU (332) or 109 CFU (102)), variation may be introduced by factors such as the virulence of the strain (271) and the health status of the infected individual (119). For instance, in the 1998 multistate U.S. listeriosis outbreak, the frankfurters consumed contained less than 0.3 CFU/g (241). Consequently, studies that employ models that more

Texas Tech University, Alex Brandt, May 2014 accurately estimate the infectious dose of L. monocytogenes are needed in order to provide better data that can be used for improving risk assessments. Pathogenesis and Virulence Factors Because foods are largely implicated as the primary means of L. monocytogenes entry into the human body (287), the intestinal tract is considered to be the primary site of L. monocytogenes pathogenesis. After reaching the intestines, L. monocytogenes employs the surface proteins InlA, ActA, Ami, p104, and FbpA to aid in adherence to the epithelial cell surface (11, 95, 245, 248, 280). Uptake into normally non-phagocytic intestinal epithelial cells is primarily induced through an interaction involving InlA (114). InlA is a membrane-anchored protein characterized by the presence of leucine-rich repeats (LRRs), whose LRR 6 binds to host cell E-cadherin at proline 16 (250). This interaction initiates a cascade involving intracellular β-catenin and α-catenin, which ultimately leads to rearrangement of the epithelial cell F-actin cytoskeleton (205) to engulf the bacterium via a “zipper” mechanism (206). Engulfed bacterial cells are subsequently encased in a vacuole (349) that is lysed by the action of the hemolysin listeriolysin O (LLO) (115) along with a phosphatidylinositol-specific phospholipase C (244) and a broad-range phospholipase C (138) that is matured by the action of a zinc metalloprotease (290); inability to lyse the vacuole prevents further L. monocytogenes replication and growth (115). Once released into the cytoplasm, L. monocytogenes is able to replicate, and begins to recruit F-actin through the action of ActA to form actin filaments (90). A characteristic “comet tail” forms (349) to propel the bacterium through the cytoplasm, toward the membrane, and into a protrusion that extends into a neighboring epithelial cell (254). The bacterium forms a double-membrane vacuole in the newly-invaded cell (349), which is lysed through the action of LLO, the zinc metalloprotease, and the two phospholipases. Upon crossing the epithelial layer, the organism travels through the blood and lymph to colonize the mesenteric lymph nodes, spleen, and liver (232). Further dissemination through the blood to the placenta of gestating individuals (4), as well as the brain and central nervous system (18, 263), subsequently occurs, and this leads to the deleterious symptoms described above. InlB, a close relative of InlA, also contributes to host cell invasion, especially in non-epithelial cells such as hepatocytes (94). InlB is encoded by inlB which is part of an 8

Texas Tech University, Alex Brandt, May 2014 operon (inlAB) that also contains inlA, the gene that codes for InlA. Transcription of this locus is controlled in part by the regulatory factor PrfA (215), which also controls the expression of those virulence factors included in the Listeria monocytogenes pathogenicity island 1 (LIPI-1) (402). Included in the LIPI-1 are prfA, plcA (encoding the phosphatidylinositol-specific phospholipase C), hly (encoding LLO), mpl (encoding the zinc metalloprotease), actA (encoding ActA), and plcB (encoding the broad-range phospholipase C). In addition to these classic virulence factors, the pathogenic contribution of other virulence factors such as other internalin proteins (130), chitinases (59), and the comK regulon (292) remains to be fully elucidated. To date, a number of premature stop codon mutations in inlA have been identified, which encode for a truncated and secreted form of InlA, and are associated with attenuated virulence in human cell lines and animal models (173, 266, 398). The distribution of these mutations is intriguing as isolates harboring them are more prevalent among foods than among clinical cases (266, 399). This seems to suggest a predisposal of this group to survival in foods and food-related environments.

Salmonella enterica Classification, Serotypes, and Reservoirs Gram negative bacteria belonging to the genus Salmonella are facultatively anaerobic and do not form spores (160). Taxonomically, the Salmonella genus is within the Enterobacteriaceae family, and, at present, includes two species: S. enterica and Salmonella bongori (6). Though cases of human infection due to S. bongori have been reported (121), the species is relatively rare, is primarily isolated from cold-blooded animals, and is less well-understood than S. enterica (108). Salmonellosis in mammals is primarily caused by S. enterica, making it the more clinically relevant of the two species. Consequently, in this review, the emphasis will be on S. enterica. S. enterica is further diversified into six subspecies: (I) subsp. enterica, (II) subsp. salamae, (IIIa) subsp. arizonae, (IIIb) subsp. diarizonae, (IV) subsp. houtenae, and (VI) subsp. indica (40). Further sub-classification of S. enterica into serotypes based on the White-Kauffmann-Le Minor scheme, which categorizes the organism based on its somatic (O), flagellar (H), and capsular (Vi) antigens, yields a total of 2587 serotypes, at 9

Texas Tech University, Alex Brandt, May 2014 present (141). Among all S. enterica subspecies, S. enterica subsp. enterica is the most diversified, accounting for nearly 60% of all currently recognized serotypes (141), and is most commonly associated with disease. Additionally, recent genomic evaluation of S. enterica subsp. enterica also suggests that two further subpopulations may exist within the subspecies that differ in their profiles regarding host and tissue tropism, metabolism, ecology, and transmission dynamics (82). Serotypes belonging to S. enterica subsp. enterica are typically assigned a name, which is generally derived from the geographic location where the isolate was first obtained; serotypes from other subspecies are designated by their associated Roman numeral and their antigenic formula (141). For abbreviation, S. enterica subsp. enterica serotypes are written with an “S.” followed by the serotype name; an example is S. Tyhpimurium. This scheme will be used henceforth in this review. S. enterica susbp. enterica serotypes can also be distinguished epidemiologically based on their host range. Serotypes such as S. Typhi, S. Paratyphi A, and S. Paratyphi C are particularly adapted to human hosts; animal-adapted serotypes (which may or may not be pathogenic to humans) include S. Gallinarum (poultry), S. Dublin (cattle), and S. Choloraesuis (swine). The remaining serotypes are not particularly host adapted, can be pathogenic to humans and animals, and comprise most foodborne isolates (168). Morphology and Growth Characteristics S. enterica cells appear as straight rods with dimensions of 0.7-1.5 µm X 2-5 µm, and commonly possess peritrichous flagella (160). Nonflagellated variants, such as S. Gallinarum and S. Pullorum, do however exist (419). Interestingly, most S. enterica are able to alternately express two forms of filamentous flagellar proteins, FljB and FliC, in a posttranscriptionally-controlled process known as flagellar phase variation (36); this variation is accounted for by multiple H types in current sertoyping schemes (141). Though the optimal growth temperature for the organism is 37°C, it is quite versatile and can maintain growth even at temperatures as high as 45.6°C (12) and as low as 2-4°C (77). The ideal pH range for salmonellae is 6.5 to 7.5, though growth has been documented at extremes of 9.50 (159) and 3.99 (14). Growth is usually inhibited at salt levels above 3-4%, but higher temperatures within its optimal range can facilitate growth at high salinity and low pH (5, 347). The organism is also resilient with regard to low aw, 10

Texas Tech University, Alex Brandt, May 2014

and some strains have been shown to survive for extended periods of time at aw values as low as 0.2 at 22°C (166). This resilience has recently prompted a greater focus on the need to develop adequate sanitation measures to counter the pathogen in low moisture food processing environments (33). Resistance to heat treatment is also increased at low aw, creating an added challenge for low moisture food processors with regard to achieving thermal lethality for the pathogen (288). Public Health and Economic Burden The current annual health burden of foodborne salmonellosis (nontyphoidal) in the U.S. is estimated to include nearly 1.03 million cases of illness (318) and a total cost of $4.4 billion (319). As such, S. enterica is the etiologic agent with the highest amount of annual illness cases among foodborne bacteria, and is the second highest among all foodborne pathogens behind Norovirus (318). Though illness estimates are large, case- hospitalization and case-fatality rates are relatively low, resulting in 19,336 annual hospitalizations and 378 deaths per annum (318). Based on surveillance from FoodNet, the top ten serotypes causing laboratory-confirmed S. enterica infections in 2011 were Enteritidis, Typhimurium, Newport, Javiana, I 4,[5],12:i:-, Muenchen, Heidelberg, Montevideo, Infantis, and I 13,23:b:- (386). Compared to the list from 2010, the only differences in 2011 were the inclusion of I 13,23:b:- and the exclusion of Saintpaul (386). Regarding recalled products, S. enterica was documented as the reason for 31 recalls of USDA-FSIS regulated foods from 2005 to 2013 (361), and instigated 951 Class I recalls of FDA regulated food products during the time span from 2003 to 2011 (85). Foodborne Outbreaks of Salmonellosis The proportion of Salmonella illness cases that originate from foodborne transmission is estimated to be 94% (318), and the pathogen prompted an estimated 996 recalls of FDA-regulated food products from 2003 to 2011 (85), over four times the amount for L. monocytogenes, the next closest pathogen. Major foodborne outbreaks have been linked to foods from various groups including eggs, meat and poultry, dairy products, raw fruits and vegetables, nuts, and spices. Though one of the first recorded U.S. outbreaks of salmonellosis linked to egg products occurred in 1974 (161), more recently a 2010 outbreak involving S. Enteritidis contamination of shell eggs produced in Iowa had 1939 laboratory-confirmed cases and 11

Texas Tech University, Alex Brandt, May 2014 sickened people in multiple states (380). Meat and poultry products are a frequent source of salmonellosis outbreaks and recent major U.S. outbreaks have involved S. Heidelberg in ground turkey causing 136 illnesses in 34 states in 2011 (382), S. Enteritidis in ground beef causing 46 illnesses in 9 states in 2012 (384), and S. Heidelberg in Foster Farms Chicken causing 430 illnesses in 23 states and Puerto Rico in 2013 (388). Though dairy products have not been implicated as the source of any recent salmonellosis outbreaks, one of the largest U.S. salmonellosis outbreaks ever recorded involved S. Typhimurium in pasteurized milk which sickened over 16,000 in 1985 (204). Fruits and vegetables have also been the food vehicles in numerous outbreaks including a large 2008 outbreak involving S. Saintpaul in raw peppers which led to 1443 illnesses in 43 states (376). Raw sprouts (377, 383) and melons (385) have also been implicated in several recent outbreaks. Low moisture foods such as nuts (378), nut products (379), cereals (375), and spices (381) have also been implicated as vehicles of salmonellosis outbreaks in recent years, and have prompted FDA to conduct risk assessments on these products (389). Current Regulations and Control Measures The current regulatory position of the FDA toward detection of S. enterica in RTE foods under its jurisdiction is similar to that for L. monocytogenes, where the agency exercises zero-tolerance based on Sections 402(a) (1) and 402(a) (4) of the FFDCA. FSIS also holds a policy of zero-tolerance for the pathogen in RTE meat and poultry products, as it does for L. monocytogenes (357), but currently does not consider raw product to be adulterated if it bears a detectable amount of the pathogen (233). Rather, based on the 1996 Pathogen Reduction; Hazard Analysis and Critical Control Point (HACCP) Systems Final Rule (353) producers of raw meat and poultry products must meet certain minimum performance standards for S. enterica in their products based on routine FSIS testing results. If the establishment fails to meet these performance standards it must (i) take immediate action to meet the standard or (ii) reassess its HACCP plan if it is unable to meet the standard on the next series of compliance tests for that standard (353). If corrective actions and the reassessed HACCP plan are not sufficient to meet the standard, FSIS reserves the right to suspend inspection based on an inadequate HACCP plan as defined by 9 CFR Part 417 (353). FSIS sampling for the pathogen in RTE meat and

Texas Tech University, Alex Brandt, May 2014 poultry products is routinely conducted under the ALLRTE and RTE001 sampling programs along with L. monocytogenes (359). Salmonellosis and Infectious Dose Clinical conditions resulting from human S. enterica infection include enteric (typhoid and paratyphoid) fevers (caused by S. Typhi, S. Paratyphi A, B, and C, and S. Sendai) (324), systemic infections by nontyphoidal organisms, and uncomplicated gastroenteritis (76, 156). Typhoidal infections can produce severe disease in the host characterized by invasion of tissues beyond the gastrointestinal tract and systemic manifestation (76). After an incubation period of one to four weeks, a prolonged period of fever ensues along with watery diarrhea, headache, abdominal pain, prostration, and cutaneous rose spots (76). Though mostly considered a traveler’s disease in the developed world, typhoidal infections are still common in developing countries (162). In the U.S. most foodborne salmonellosis is nontyphoidal in nature (318), and infections are usually characterized by self-limiting gastrointestinal illness with occasional bacteremia and focal infection (76, 156). Symptoms generally develop in the range of 12 to 14 hours after ingestion of the contaminated food, and can include nausea, abdominal pain, vomiting, headache, chills, diarrhea, and moderate fever (168). Unlike protracted typhoidal infections, which can last longer than 2 weeks (not including relapses), symptoms of non-typhoidal salmonellosis usually resolve within 2-5 days after appearance of symptoms (168), and generally only require supportive therapy such as fluid and electrolyte replacement. Though shedding of S. enterica usually ceases 5 weeks post-infection, some patients remain in a chronic asymptomatic carrier state (45). The concentration of cells required to cause salmonellosis has generally been considered to range from 107 to 109 CFU/g (168). However, it has been demonstrated that 1 11 this number can vary extensively from <10 CFU (178) up to 10 CFU (297) based on the serotype and food vehicle involved. Recent analysis of the infectious dose based on outbreak data revealed outcomes much lower than the accepted values; an infection ID50 and illness ID50 of 7 CFU and 36 CFU, respectively, were observed (343). The continued emergence of multi-drug resistant strains of S. enterica that began nearly four decades ago (348) has also created a public health concern regarding clinical treatment of salmonellosis. Emergence of these strains in human populations, and the potential 13

Texas Tech University, Alex Brandt, May 2014 contribution of antibiotic use in animal agriculture, are current topics of great concern, and have been reviewed extensively elsewhere (9, 236). Pathogenesis and Virulence Factors Like L. monocytogenes, the primary site for S. enterica infection is in the intestinal tract where it is able to enter the body by attaching to and invading epithelial cells and M-cells overlying Peyer’s Patches. A vast majority of the genes encoding S. enterica virulence factors are located in clusters known as Salmonella Pathogenecity Islands (SPIs) (309), which are primarily characterized by an altered G/C content compared to surrounding DNA (183). Among the 15 SPIs currently identified in S. Typhi (183), SPI-1 and SPI-2 are by far the best characterized (143). Genes located in SPI-1 are primarily involved in invasion, while genes from SPI-2 assist in intracellular survival (143). Prior to invasion, successful attachment is necessary (113), and is primarily mediated by interactions of glycoprotein receptors on host cell surfaces with type 1 fimbriae (encoded by the fim gene cluster) (421), gene products of the plasmid-encoded pef operon (23), and products of the lpf operon (24). Genes from the SPI-1 inv, prg, and spa islands are subsequently expressed to produce structural elements of a Type 3 Secretion System (T3SS) that allows for transport of invasion effector proteins into the host cell (424). The major effector proteins delivered through the T3SS include SipA-C, SptP, and SopA-E (69, 70), which promote Ca2+ influx to signal actin polymerization (306) and subsequent cytoskeletal rearrangement to cause membrane ruffling (239). Through ruffling, the host cell membrane and cytosol are rearranged such that the S. enterica cell is essentially surrounded by extensions of the host cell (172). The S. enterica cell is then internalized via a “trigger” mechanism (340) into a vacuole known as the Salmonella containing vacuole (SCV) (192). Within the SCV the low Mg2+ conditions are a trigger for the PhoP/PhoQ two- component regulatory system to repress hilA, which in turn suppresses activity of the SPI-1 T3SS (135). Unlike vacuoles subject to normal endocytic processing pathways, the SCV undergoes a maturation process that prevents its fusion with a lysosome (17), permitting the survival of the bacterium. The effector SopB secreted by the SPI-1 T3SS diverts the SCV trafficking from normal endosomal maturation, while also delaying the SCV-lysosomal fusion (231). The structural components of a second T3SS involved in 14

Texas Tech University, Alex Brandt, May 2014 pathogensis are encoded by the ssa operon of SPI-2, which also contains a host of genes whose products are regulators, chaperones, and effectors for the secretion system (150). As noted, the SPI-2 is key to intracellular survival. The SPI-2 effector protein SpiC is also involved in prevention of SCV-lysosomal fusion particularly for late-stage endosomes (392). Other SPI-2 effectors, SifA, SseG, and SseF cooperate to localize the SCV in a perinuclear region near the Golgi where the compartment can absorb nutrients from nearby transport vesicles (196) allowing the organism to survive intracellularly. Further dissemination of the organism beyond the intestinal epithelium involves uptake of the pathogen by phagocytic cells such as macrophages and dendritic cells (249) where it is able to spread through the lymphatic system to draining lymph nodes and cause systemic illness (171). Lymph nodes, especially those embedded deep within muscle tissue, are typically included along with beef and pork carcass trim as a component of ground meat. Within lymph nodes, S. enterica may avoid carcass surface interventions, and later be dispersed through the product when lymph nodes are ground. Even when fabricating whole cuts of meat, the presence of S. enterica in the lymph nodes is problematic, because the lymph nodes are often cut open during breaking down of carcasses and portioning. Thus, survival of S. enterica in the mammalian lymphatic system has recently garnered attention as a food safety hazard for ground meat products (133), and is likely to gain significant regulatory attention in the near future.

Escherichia coli O157:H7 Classification, Serotypes, and Pathotypes E. coli is a Gram negative rod-shaped coliform bacterium that, like S. enterica, belongs to the Enterobacteriaceae family (160). With the recent addition of the species Escherichia albertii, the Escherichia genus includes six species: E. albertii, E. coli, Escherichia fergusonii, Escherichia vulneris, Escherichia blattae, and Escherichia hermannii (2). Escherichia are facultatively anaerobic and can either be motile by peritrichous flagella or nonmotile (160). Serotyping of E. coli is similar to that of S. enterica where the somatic (O), flagellar (H), and capsular (K) antigen profiles can provide sub-classification of the organism (277); however only the O and H antigens are generally used to define E. coli serotypes.

Texas Tech University, Alex Brandt, May 2014

While most E. coli exist as commensal inhabitants of the gastrointestinal tract of humans and animals (342), certain strains have evolved to become pathogenic to hosts (259). The range in pathogenic mechanisms within the species is diverse, and at least six different pathotypes have been identified to date; these include enterotoxigenic E. coli (ETEC), enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), enteroaggregative E. coli (EAEC), enteroinvasive E. coli (EIEC), and diffusely adherent E. coli (DAEC) (259). More recently, a large outbreak in Europe involved a strain of E. coli O104:H4 that possessed elements that were characteristic of both EAEC and EHEC (387). Among the pathotypes, EHECs cause the most severe disease and produce the greatest amount of E. coli-associated illness in the U.S. (318). Though strains belonging to several O groups within the EHEC pathotype produce a substantial health burden (42), serotype O157:H7 is the predominant worldwide cause of the debilitating hemolytic uremic syndrome (HUS) associated with EHEC (341). Consequently, since its emergence as a foodborne pathogen in the early 1980s (180, 298), O157:H7 has been increasingly recognized as an organism of global public health importance. Growth Characteristics and Acid Tolerance Similar to S. enterica, E. coli O157:H7 has an optimal growth temperature of 37°C (160), though it can survive for up to 9 months at -20°C, and generally can grow well at temperatures up to 44°C (93). Compared to other foodborne pathogens the organism is particularly acid tolerant, and can survive in acidic conditions with a pH as low as 4.0 for up to 56 days; however the response seems to be dependent on the acidulant and environmental conditions (71). The acid tolerance response observed in E. coli O157:H7 is mediated by three major acid tolerance mechanisms: a glutamate- dependent system, an acid-induced arginine-dependent system, and an acid-induced oxidative system (213). The combined action of these systems is especially important for the pathogen’s transversal of the stomach during the digestive process where the organism can be exposed to pH values as low as 1.5 to 2.5. This resilience also allows the pathogen to survive in acidic foods such as fermented sausage (127) and apple cider (425). The pathogen is not particularly salt tolerant, as growth rates noticeably decrease when the pathogen is exposed to 4.5% NaCl, and have been shown to halt at a concentration of ≥8.5% NaCl (127). Though survival of the pathogen is significantly 16

Texas Tech University, Alex Brandt, May 2014

reduced in low aw foods, and its ideal aw for growth is 0.95 (33), its survival for up to 19 weeks at aw values as low as 0.16 at 5°C has been documented (84). Public Health and Economic Burden With the majority of cases (68%) linked to foodborne transmission, E. coli O157:H7’s annual public health burden in the U.S. involves over 63,000 cases of foodborne illness, greater than 2100 hospitalizations, and 20 deaths (318), resulting in an economic burden of $607 million (319). Likewise, its burden on the food supply is demonstrated by the 117 products that were recalled via 19 Class I recalls of FDA regulated products from 2003 to 2011 (85) and the 93 recalls of USDA-FSIS regulated products during the span from 2005 to 2011 (361). As with S. enterica, the case-fatality rate of O157:H7 is quite low at 0.5%, but due to the severe nature of HUS, especially in children under 10 where HUS develops in ~15% of O157:H7 infections (27), the hospitalization rate is estimated to be 46.2% (318). Using data from past foodborne outbreaks, the infectious dose of the organism is generally deemed to be fewer than 100 CFU (352), owing in part to its acid tolerance mechanisms that allow the small number of cells to survive the acidity of the stomach and reach the intestines to manifest disease. Infection and Hemolytic Uremic Syndrome As mentioned above, the more serious complication associated with E. coli O157:H7 infection is HUS, but the spectrum of disease also includes hemorrhagic colitis and less complicated presentations like abdominal cramps and nonbloody diarrhea (180). Following ingestion of the bacterium, it colonizes the colon over a three to four day incubation period. The first wave of symptoms begins soon afterward with abdominal cramps that can last for one to two days (341). Bloody diarrhea lasting from four to ten days ensues, but symptoms commonly resolve within seven days (341). However, among all ages about 6% of cases do not resolve, and will transition into HUS (131). Because of the toxico-infective nature of the illness, the use of antibiotics is discouraged, as it elevates systemic toxin levels and exacerbates the degree of illness in the infected individual (331). Thus, dialysis and blood transfusions are currently the sole remedies for the effects of HUS until infection resolves (341). Beyond the immediate health complications of an O157:H7 infection, elevated risk for long term sequelae such as hypertension, renal impairment and cardiovascular disease is also a major concern (65). 17

Texas Tech University, Alex Brandt, May 2014

Pathogenesis and Virulence Factors The signature virulence factors of E. coli O157:H7 (and all EHEC) are its ability to attach to and efface the intestinal epithelium, and its production of the cytolethal Shiga toxins Stx1 and Stx2 (259). Upon reaching the colon of the human host, surviving cells secrete small amounts of Stx2, which in turn causes localization of nucleolin on the host cell-surface (301). Intimin, the E. coli adhesin of primary importance in intestinal colonization, is expressed on the E. coli surface and can bind to the bundled host nucleolin (329). Intimin is encoded by eae, which is located in a 42 kb pathogenecity island known as the locus of enterocyte effacement (LEE) (238). The LEE also includes genes that encode for a translocated intimin receptor (Tir), machinery for a T3SS, and effectors delivered through the T3SS. The E. coli T3SS is a needle-like structure inserted into the colonocyte after initial attachment (167). It is assembled from the LEE proteins EscF and EspA (193, 418). The proteins EspB, EspD, SepL, and SepD interact to form a translocation pore in the host cell membrane (185, 194) through which Tir, LEE effectors, and non-LEE effectors are subsequently delivered into the epithelial cell. Tir localizes on the colonocyte surface and serves as a ligand for intimin, resulting in intimate attachment of the E. coli cell to the colonocyte (144). Signals originating from the intracellular domains of Tir are triggered by this binding and cooperate with the action of the non-LEE effector EspFu to subvert the host cell cytoskeleton and actin polymerization through interaction with the host cell proteins IRSp53 and N-WASP (50, 411). The result of this action is the formation of an actin pedestal on top of which the E. coli positions itself; the histopathological lesion that is characteristic of this pedestal formation is known as the attaching and effacing, or “A/E” lesion (251). Shiga toxins, so named for their similarity to the Shigella dysenteriae Type 1 cytotoxin (Shiga) (269), are also key virulence factors responsible for the hemorrhagic colitis and HUS that can develop in O157:H7 infection. Though most E. coli O157:H7 strains possess both Stx1 and Stx2, cells with only one toxin or neither are occasionally isolated (51, 321). The toxins are encoded on lambdoid prophages lysogenized in the chromosome (336), and are comprised of A and B subunits (48) that together form an

AB5 superstructure (188). The homopentamer formed from B subunits binds to the host cell surface glycolipid globotriaoslyceramide (Gb3) (214), which facilitates its uptake 18

Texas Tech University, Alex Brandt, May 2014 into the cell (188). Once internalized, the A portion acts as an N-glycosidase, causing removal of an adenosine of the eukaryotic 28S rRNA and subsequent inhibition of protein synthesis (101). This leads to irreparable cell damage and death. In the bloodstream the toxin is carried, possibly by polymorphonuclear lymphocytes (41), to tissues that are rich in Gb3 surface protein; renal glomerular endothelial cells are high in

Gb3 and are the key target for manifestation of HUS (391). This can lead to apoptosis of kidney cells and subsequent renal failure if dialysis is not performed. Other genetic elements such as the pO157 plasmid (212) and O islands (179) are also believed to contribute to pathogenesis. However, their roles remain to be fully elucidated. Foodborne Outbreaks and Current Regulation The E. coli O157:H7 outbreaks of 1982, which garnered its recognition as a foodborne pathogen, were separate outbreaks in Oregon and Michigan that resulted in 26 cases and 19 hospitalizations, and 21 cases and 14 hospitalizations, respectively (298). Both outbreaks were associated with undercooked hamburgers from two different locations of the same restaurant chain. A larger multistate outbreak occurred in 1993 in the western U.S. with 731 cases identified. Hospitalizations totaled 178, with 56 cases of HUS, and 4 deaths. The outbreak resulted from consumption of ground beef served by a restaurant chain that also was not sufficiently cooked to kill E. coli O157:H7 (352, 365). As a direct consequence of the 1993 outbreak, FSIS began to review measures being taken by the beef industry to control the pathogen. These efforts led to the 1994 declaration that raw ground beef contaminated with any detectable amount of the pathogen was considered adulterated; this was clarified in 1999 to include all non-intact beef products (354). These measures also culminated in FSIS’s enactment of the 1996 Pathogen Reduction; Hazard Analysis and Critical Control Point (HACCP) Systems Final Rule (353), which required meat and poultry establishments to implement HACCP to control reasonably likely to occur biological hazards (such as E. coli O157:H7) and to include testing for generic E. coli as a process control indicator. More recently, FSIS has expanded its position regarding adulteration of raw non-intact beef to include six other Shiga-toxin producing serotypes (O26, O45, O103, O111, O121, and O145) that are also of considerable public health importance in the U.S. (358).

Texas Tech University, Alex Brandt, May 2014

Foods other than non-intact beef that have also been implicated in notable outbreaks of E. coli O157:H7 illnesses include produce, apple cider and juice, waterborne infections, and fermented and cooked meats. Produce outbreaks have mainly involved leafy greens and sprouts. A 2006 outbreak, which spanned multiple U.S. states and Canada, involved 206 cases and 3 deaths that were associated with pre-packaged fresh spinach (374). An extremely large outbreak in Japan in 1996 was associated with consumption of white raddish sprouts and led to 7966 cases and 3 deaths (246). Apple cider was implicated in a 1991 Massachusetts outbreak that sickened 23 (32), and several years later in 1996, apple juice was implicated as the vehicle for 71 illnesses and 1 death in three western U.S. states and British Columbia (367). Though several outbreaks of illness have been associated with contaminated water, one of the largest recorded to date was a 1999 outbreak at a county fair in New York which produced over 900 infections, 65 hospitalizations, and two fatalities (370). A Scottish outbreak in 1995 involved cooked meat from a local butcher that sickened 501 people and killed 21 total (74). Other meat products that have been implicated in O157:H7 outbreaks include dry-cured salami on two occasions (226, 366), and jerky made from deer meat (184).

Bacterial Subtyping Overview and Considerations Much of the knowledge of the epidemiology and transmission of foodborne pathogens, especially in outbreak situations, is garnered from bacterial subtyping methods. Barrett and Gerner-Smidt describe bacterial subtyping as “the characterization of bacteria below the species (or subspecies) level” which “allow[s] the isolates to be placed into groups in which members are more or less similar to one another based on one or more (usually many) characteristics” (21). Because bacteria reproduce asexually, the progeny are essentially clones of the parent cell, except in rare cases where mutations or horizontal gene transfer alter the genetic composition of the organism (413). The rare nature of these alterations permits subtle differentiation of strains, but also allows progeny originating from a certain ancestor to be systematically grouped together. A plethora of subtyping methods are available to the investigator. This necessitates the consideration of certain criteria (such as cost, discriminatory ability, 20

Texas Tech University, Alex Brandt, May 2014 standardization and reproducibility, automation and ease of use, speed, and applicability to a given organism) in choosing an optimal method (413). Current subtyping methods can largely be divided into two categories: 1. phenotype-based methods that target metabolites and extracellular surface molecules and 2. DNA-based (molecular) methods that target nucleic acids (413). Typical phenotypic methods include serotyping, phage typing, biotyping, and antimicrobial susceptibility testing (21). Though DNA-based methods have largely become the benchmark for subtyping, phenotypic methods are still commonly combined with molecular methods for identification. A classic example is the continued use of S. enterica serotyping in conjunction with DNA-based methods (141). Within the DNA-based subtyping category exists the two sub-categories of band-based methods and sequence-based methods (272). Band-based methods largely involve the production of DNA fragments that are resolved to create a banding pattern. This pattern essentially servers as a barcode for the subtype. Sequence-based methods determine single nucleotide polymorphism (SNP) profiles that are unique to certain subtypes. Band-Based Molecular Subtyping Methods Band-based molecular subtyping methods are further divided into: 1. non-DNA based methods, 2. methods involving enzymatic restriction of DNA, 3. polymerase chain reaction (PCR) methods, and 4. methods which combine PCR and enzymatic restriction of DNA (21). Because the costs of DNA sequencing were comparably higher during the advent of molecular subtyping methods, band-based methods were some of the first to be widely adopted. Though they are gradually being replaced with advanced methods (311), they still retain utility in a variety of applications. Pulsed-field gel electrophoresis (PFGE) is a band-based subtyping method which has evolved to become the “gold standard” of molecular subtyping since its inception in the 1980s (322). The method entails the digestion of whole genomic DNA embedded in an agarose plug using a “rare-cutting” restriction enzyme, such as XbaI. This is followed by multi-directional agarose gel electrophoresis using one of a number of detailed algorithims for switching directions of the electrical field (164). Patterns produced by the DNA fragments correspond to bacterial subtypes, and are often referred to as “pulsotypes” (413). Though disadvantaged by cost and the amount of time required, good inter-laboratory reproducibility via highly standardized protocols is a key advantage for 21

Texas Tech University, Alex Brandt, May 2014 the use of PFGE (272). Consequently, since its establishment in the 1990s, the CDC PulseNet program has employed PFGE as its primary subtyping tool for foodborne pathogens across its contributing state public health laboratories (338). Other band-band based methods that are available for subtyping include multilocus enzyme electrophoresis (MLEE), restriction fragment length polymorphism (RFLP) analysis (333), repetitive extragenic palindromic PCR (REP-PCR) (405), random amplified polymorphic DNA (RAPD) (416), and amplified fragment length polymorphism (AFLP) (408). Though they are highly discriminatory and are low in cost per test, widespread adoption of the PCR-based REP-PCR and RAPD methods have largely been hindered by the concerns of low reproducibility, especially across laboratories (272). AFLP, which is a method that combines enzymatic restriction of whole genomic DNA, adaptor ligation, and PCR is hindered by cost, but has good reproducibility (272). AFLP has been used widely in bacterial, animal, and plant subtyping (316). MLEE is a non-DNA band-based method that distinguishes subtypes based on electrophoresis of intracellular metabolic enzymes; enzymes with different amino acid substitutions have different electrophoretic patterns when placed under an electrical field (209). RFLP involves the use of a “frequent-cutting” restriction enzyme that produces hundreds of small DNA-fragments, which are normally difficult to visualize. However, methodological variations such as EcoRI ribotyping (337), which involves transfer of restricted DNA to nylon or nitrocellulose membranes followed by hybridization with a labeled probe for rRNA genes, have gained traction. EcoRI ribotyping itself become has become highly automated with the advent of instruments such as the DuPont Riboprinter (43), and has been shown to have utility in subtyping foodborne organisms such as S. enterica (151) and L. monocytogenes (81). Sequence-Based Molecular Subtyping Methods Included in the sequence-based subtyping category are methods such as multilocus sequence typing (MLST), multi-locus variable number tandem repeat analysis (MLVA), and whole genome sequencing (WGS) (21). MLST usually involves the direct sequencing of 450- to 500-bp regions of seven highly conserved bacterial housekeeping genes (55). The obtained sequences are then compared with others in a database such as MLST.net (http://www.mlst.net) to assign a subtype to the isolate based on the 22

Texas Tech University, Alex Brandt, May 2014 sequencing profile. A modified MLST scheme has been recently employed which shifted the focus from housekeeping to virulence genes, to create a method known as multi- virulence locus sequence typing or MVLST (423). MLVA is a sequencing technique that focuses on the varied size, location, and complexity of genetic elements known as variable tandem number repeats (VNTR) (395). In this scheme, multiple VNTR loci can be sequenced and compared to create a subtyping profile much like that done in MLST. Allelic typing (AT) schemes that involve sequence typing and phylogenetic analysis based on a single gene have also proven useful for subtyping pathogens such as L. monocytogenes and other species within the Listeria genus (264). With costs constantly decreasing and throughput constantly increasing due to the implementation of next generation sequencing methods, WGS holds great promise as an optimal subtyping technique of the future (38). WGS provides the ultimate level of discrimination by targeting SNPs, mobile genetic elements, and other unique features in the entire genome. It has already proven its utility in the real-time investigation of several illness outbreaks (202), including foodborne outbreaks associated with L. monocytogenes (123) and Escherichia coli O104:H4 (132), and will likely continue to display its usefulness in future investigations as the speed of analysis is improved (311). Utility of Molecular Subtyping As implied above, gathering subtyping information along with epidemiological data in large databases, like CDC PulseNet, facilitates surveillance of foodborne pathogens, and aids in identification of foodborne outbreaks (338). If routine analysis reveals temporal and geographic clustering among certain pathogen subtypes, these relationships may be indicative of the occurrence of a foodborne outbreak (21). By matching subtypes of isolates from clinical cases and food products, the source of the outbreak can be rapidly identified and mitigated (120). A real-life example of the contribution of PulseNet to an outbreak investigation was seen in the 1998 listeriosis outbreak associated with frankfurters. Through routine analysis it was discovered that 113 of the 247 clinical L. monocytogenes in the U.S. had matching PFGE patterns (134). By conducting a case-control study where outbreak strain-infected patients were used as cases, and patients infected with other strains were used as controls, frankfurters produced by a certain facility were able to be implicated as the vehicle of the outbreak. 23

Texas Tech University, Alex Brandt, May 2014

Further expansion of these databases to global proportions can also be useful in identifying worldwide trends in disease, and improving ecological knowledge of foodborne pathogens (339). More recently, open source databases which contain strain subtyping information of foodborne pathogens, such as Food Microbe Tracker (www.foodmicrobetracker.com), have also become available, and are helping to build knowledge of the ecology of the foodborne pathogens throughout the food chain (400). Beyond outbreak investigations, subtyping has utility in tracking the emergence of new pathogen subtypes. A case-in-point is the study of the emergence of the multidrug resistant S. Typhimurium DT104 subtype (66). Molecular subtyping can also be combined with longitudinal sampling of processing facilities, herds, farms, flocks, and other sources of foodborne pathogens along the farm-to-fork continuum to determine distribution and transmission patterns of particular subtypes. This particular utility of molecular subtyping will be the major topic of discussion going forward in this review.

Molecular Ecology and Persistence of Foodborne Pathogens Molecular Ecology Advancements in molecular subtyping have driven rapid growth in the field of biology known as molecular ecology. Traditional ecology focuses on understanding population structures, interactions between species, and interactions between organisms and their environment; molecular ecology has the same focus but uses molecular subtyping tools (like those described above) to elucidate these relationships. These same principles have been applied to the study of foodborne pathogens in specific ecological niches. Knowledge gained through these studies can provide better revelations of how foodborne pathogens are distributed throughout the farm-to-fork continuum, how the pathogens interact with and adapt to their local environments, and how the pathogens interface with humans and other potential hosts. Numerous studies have been performed in the last several decades that probe the molecular ecology of L. monocytogenes, S. enterica, and E. coli O157:H7 in different ecological niches. L. monocytogenes ecology has been studied in livestock operations (e.g.(139, 265)), agricultural environments (e.g.(335)), urban and natural environments (e.g.(312, 314)), dairy production and processing environments (e.g.(153, 176, 200)), 24

Texas Tech University, Alex Brandt, May 2014 seafood processing environments (e.g.(198, 268, 305)), meat and poultry processing environments (e.g.(56, 201, 283, 325, 417)), chilled and other RTE food processing environments (e.g.(157, 186)), and at retail RTE food establishments (e.g.(154, 313)). Similarly, S. enterica ecology has been studied in livestock populations (e.g.(88, 118, 210, 211)), wildlife populations (e.g.(128, 260)), environmental sources (e.g.(163, 282)), foods and animal feed (e.g.(234, 262, 291)), and agricultural environments (e.g.(80, 308, 393)). Likewise, E. coli O157:H7 ecology has been investigated in livestock herds (e.g.(64, 116, 293, 307)) and on carcasses presented for slaughter (e.g.(20)). With a growing body of knowledge, the transmission dynamics and population structures of these pathogens in certain environments (with particular interest in those linked to food production) will help mitigate their spread throughout the food chain and minimize their ultimate transfer to humans. Furthermore, by combining longitudinal sampling and molecular subtyping in molecular ecological studies it has been revealed that certain foodborne pathogen strains seem to be more capable of long-term survival within particular niches. This phenomenon of recurrently isolating a pathogen from a certain niche, as revealed by longitudinal sampling and molecular subtyping (referred to as “persistence”), and its observation in foodborne pathogens, will be discussed below. Multiple Uses of the Term “Persistence” As discussed by Ferreira et al. (106), the term “persistence” can take on several meanings with regard to foodborne pathogens. One use of the term is in describing long- term human and animal infections, often where the host is in an asymptomatic carrier state. This definition encompasses persistence of S. enterica within lymph node macrophages (249), persistence and long-term shedding of S. enterica from the human gastrointestinal tract (45), and persistence of E. coli O157:H7 in the gastrointestinal tract of cattle (51). The long-term survival of pathogens in the host is especially problematic with regard to contamination of foods by infected food handlers (350) and introduction of pathogens into the rearing environment for food animals (307, 308). Yet another instance is the use of “persistence” to describe the long-term survival (usually in a non-proliferative, lower metabolic state) of a pathogen in a given food product (288). The long-term survival of L. monocytogenes in a variety of food products, 25

Texas Tech University, Alex Brandt, May 2014 especially at refrigeration temperatures, is well recognized (126, 252), and is commonly referred to as persistence. The same is true for the long-term survival of S. enterica in low-moisture foods and feeds (152, 175). This form of persistence is also problematic since the organism is usually able to survive during the entire shelf life of the food product, where it is later consumed by a host and initiates disease. One last meaning for “persistence” given by Ferreira (the definition of interest in this review) is the long-term survival of a specific foodborne pathogen strain in a complex natural or human-made environment. A food processing facility is an example of a complex human-made environment (106). What is interesting in this regard is that only certain strains seem to predominate and establish long-term survival as residents of the food processing facility microbial population, while others are only transient residents. While it is postulated that the success of a strain in such a complex niche requires specific adaptation to the diverse environmental conditions that it encounters, much still remains unclear as to why this phenomenon is observed (52). Garnering a better understanding of this form of persistence is of the essence, as persistent strains may recurrently contaminate food products over time if their presence in the food processing environment is not effectively controlled. Furthermore, recent multi-state outbreaks of foodborne illness linked to contamination with presumed persistent strains (274, 275, 369, 372, 375) are demonstrative of the public health risk posed by this phenomenon. Evidence for Persistence of L. monocytogenes RTE foods (especially deli meats, frankfurters, and seafood) that are exposed to the processing environment after receiving a lethality heat treatment, and before packaging, are considered high risk foods with regard to listeriosis. Cross-contamination with L. monocytogenes cells that originate from contaminated surfaces in the post- lethality food processing environment is considered the main route by which the pathogen enters the production chain for these products. Thus, effective control programs designed to mitigate the pathogen’s presence in the processing environments for these high risk foods are considered essential to reducing foodborne listeriosis (351). However, there is a large body of evidence which suggests that certain strains of L. monocytogenes are capable of persisting in the food processing environment for extended periods of time, despite control measures that are designed to eliminate them. 26

Texas Tech University, Alex Brandt, May 2014

This is rather concerning, because even after regular cleaning and sanitation, the persistent strains can be transferred to (or remain on) product-contact surfaces (known as transfer points); this can then lead to recurrent product contamination. Table 1.1 summarizes the pertinent information for a number of studies where ecology of L. monocytogenes in food production facilities was investigated, and evidence for persistence of certain pathogen strains was able to be inferred. Based on the diversity of these studies it is evident that L. monocytogenes persistence is a ubiquitous challenge for the food industry. Results like those found by Orsi et al. are especially concerning because they demonstrate how a single pathogen strain can persist in the food processing environment for very long periods of time (275). The study suggested persistence of one EcoRI ribotype DUP-1053A L. monocytogenes strain for up to 12 years, and what’s more, this strain was tied to both a sporadic case and multi-state outbreak of listeriosis (364, 372). Several other foodborne illness outbreaks (and large recalls with no diseases reported) have been associated with persistence of L. monocytogenes in the processing environments of the implicated foods (136, 267, 286, 355, 371), further reinforcing the food safety threat posed by fully virulent strains that also persist in processing facilities. Investigators have attempted to elucidate the characteristics unique to persistent strains of L. monocytogenes from within the set of traits that would logically contribute to persistence (adherence to surfaces, disinfectant tolerance, tolerance to stresses, etc.), but have been unable to establish any clear relationships (52); the findings of these studies will be discussed in greater detail below. On the contrary, it is well recognized that the pathogen often takes refuge in a number of environmental sites such as drains and floors, hollow rollers on conveyors, cracks in tubular structures, rubber sealant strips on doors, cracks in floors and walls, tight fitting metal-to-metal and metal-to-plastic spaces, and even processing equipment compartments (351). In fact, almost any cool and damp site that is not easily cleaned and sanitized, and where food residues can accumulate, can provide refuge for the pathogen. Sites that have these characteristics are generally referred to as “niches,” but to avoid confusion with the ecological niches that are also discussed in this review, another appropriate term “harborage sites,” will be used instead. Harborage sites are generally found through routine environmental sampling conducted in the processing facility, emphasizing the need for effective sampling and

Texas Tech University, Alex Brandt, May 2014 testing components of Listeria Control Programs (351). Once identified, corrective actions should be taken to both eliminate L. monocytogenes present in the harborage site and eliminate the physical problem creating the harborage site (i.e. welding a tubular opening shut, sealing a floor crack, etc.) (46). Following these actions, regular sampling of the harborage site area should be conducted to ensure that the hazard is eliminated. In cases where physical remediation cannot be accomplished (i.e. a recurrently positive drain cannot be eliminated because of the need to have drainage from a certain area of the facility) cleaning and sanitation actions should be modified in order to reduce L. monocytogenes in the site and eliminate its transfer to other surfaces (46). Yet, even after the implementation of stringent interventions, persistent subtypes may still remain present. Autio et al. conducted a facility-wide eradication program using hot steam, hot air, and hot water, and had success with eliminating persistent strains, as verified by five months worth of negative samples following the implementation of the program (16). However, even though reductions in L. monocytogenes prevalence were observed after implementation of plant-specific interventions in studies performed by Almeida et al. (10), Lappi et al. (198), and Malley et al. (230), persistent subtypes were still isolated for the full duration of each study. These observations stress the gravity of the challenge that L. monocytogenes persistence has for food processors as well as the dire needs to understand it so that more effective control strategies can be designed. Though significant attention has been given to L. monocytogenes persistence within the confines of food processing facilities, increasing consideration has been given to how other environments (e.g. pristine natural environments, agricultural environments, animal herds, and urban environments) may serve as the primary reservoirs of the persistent strains that eventually colonize these establishments. Concerns have been raised that if a strain persists in one of these reservoirs, it may also have characteristics that allow it to persist in a processing facility. Thus, if one of these reservoirs is closely associated with a processing facility (e.g. it is in the same geographic region, is a source of raw materials for the processing facility, or is location that employees commonly come into contact with) it may also require attention in the way of control measure development and implementation. Even if strains that originate from these reservoirs are not able to establish persistence within the processing facility, their constant re-

Texas Tech University, Alex Brandt, May 2014 introduction from the reservoir is still an aspect for the food processor to contend with. Contaminating strains may be re-introduced from the reservoir via raw materials, packaging items, mobile equipment, and employee articles such as shoes and clothes. Research on the persistence of L. monocytogenes in other complex environments is limited, and only a handful of studies have documented its long-term survival in settings other than food processing facilities. Minimally-disturbed natural environments are one of the non-processing facility sources from which the pathogen has been isolated in the past (104). In fact, one of the first studies on the ecology of L. monocytogenes showed that it could reside in a saprophytic state within soil and decaying vegetation of these natural environments (410). However, the pathogen has only been observed in relatively low levels in these natural settings, and evidence points to comparably higher prevalence of the organism among agricultural and urban environments. Nightingale et al. found that ruminant livestock constitute important reservoirs through which L. monocytogenes can be disseminated (265). Other studies have also found that the pathogen is also able to survive for extended periods of time in matrices commonly found in primary agricultural environments, such as silage, animal feed, liquid manure (315), and sewage sludge (409). In fact, the 1981 Nova Scotia listeriosis outbreak was tied to cabbage that had been repeatedly fertilized with non-composted sheep manure (320). Strawn et al. also found L. monocytogenes strains of identical sigB allelic types in surface water sites of produce farms over the course of one year of sampling (335), suggesting that certain strains may also persist in these types of agricultural operations. Sauders et al. found a 22.2% prevalence of all Listeria species among 898 urban samples collected in 2001 and 2002 (314). This suggests that certain locations within urban environments may also serve as reservoirs for the pathogen. However, much remains to be ascertained about the ecology of the pathogen in these non-processing facility environments, and with more research in the area, clearer paradigms for transfer of the organism from these reservoirs to the food production chain will likely be revealed. Evidence for Persistence of S. enterica In contrast to the abundance of material published on L. monocytogenes, studies that probe the ecology of S. enterica in food processing environments are considerably limited in number. However, persistence of S. enterica in other complex ecological 29

Texas Tech University, Alex Brandt, May 2014 niches has been demonstrated. This observation is still concerning because it suggests that persistent strains may have keen adaptability which allows them to respond to diverse conditions in these niches in order to survive. This may ultimately translate into their ability to adapt to and survive under the environmental conditions of food processing facilities, thereby becoming persistent residents of the population. Several studies have shown that livestock herds and their associated production environments are niches that can harbor persistent strains of S. enterica. Callaway et al. used PFGE to investigate the diversity of the S. enterica genotypes in swine wallows over a ten month period (49). The authors determined that some strains were able to persist in wallows for more than five months, and that strains of identical subtypes were distributed across multiple wallows. Cobbold et al. conducted a longitudinal study where fecal and environmental isolates were collected from two dairies in Australia (67). Results showed that one cow shed S. Newport for up to 90 days, and that the prevalence of the organism was high in bedding material, feed refusals, lagoon slurries, and milk filters. Davies et al. tracked persistence of S. Enteritidis Phage Type 4 in a free-range chicken farm following removal of birds from the farm (80), and demonstrated that the strain was able to persist in the dried feces, litter, and feed for the entire 26 month study period. The organism was also found in soil samples after an eight month period as well as in the feces of wild animals that were in close proximity to the operation. Also, Sandvang et al. used PFGE to determine the environmental survival and spread of S. Typhimurium in three Danish swine farms (308). Their results suggested persistence of one clone in Farm 3 which comprised 100% of the isolates collected from environmental sources and feces. Outside of primary livestock production environments, the persistence of S. enterica has also been observed in crop-production and marine environments. Uesugi et al. sampled for S. Enteritidis in 17 almond orchards on three farms that were linked to an outbreak associated with consumption of raw almonds (393). Longitudinal drag swab sampling revealed that S. Enteritidis Phage Type 30 strains associated with the outbreak (which belonged to two very similar PFGE types) persisted in the orchard over a five year period. Martinez-Urtaza et al. demonstrated the persistence of three S. Seftenberg PFGE pulsotypes in coastal environments for a period of more than four years (235). This observation merits considerable concern because these same pulsotypes were later found

Texas Tech University, Alex Brandt, May 2014 in local mussel processing plants, demonstrating the potential for transfer of a persistent subtype from the surrounding environment into nearby food processing facilities. S. enterica persistence in more confined environments has also been observed. One example is given by Nesse et al. who demonstrated that distinct S. enterica pulsotyes persisted in Norwegian fish feed processing facilities for at least three years (262). A phenotypic study that included these persistent strains (along with several others isolated from fish meal and feed factories from 1991 to 2006) demonstrated that presumed persistent strains had better biofilm-forming capabilities than presumed non-persistent strains (406). Consequently, the authors concluded that biofilm-forming capability might be a key risk factor for persistence of the pathogen in processing facilities. Dunowska et al. used PFGE and antimicrobial drug susceptibility to compare a set of 56 S. enterica isolates that included strains isolated from patients and the environment of a veterinary hospital as well as epidemiologically unrelated strains (97). Their observations revealed that clones originating from a past outbreak within the hospital may have survived in the environment despite rigorous hygiene and sanitation, and likely caused subsequent nosocomial infections. Recurrent contamination of a carcass splitter with S. Rissen in a pork slaughterhouse sampled by van Hoek et al. provides one of the best examples for the possible persistence of S. enterica in a food-related environment (26). The authors concluded that this recurrent contamination was likely the result of endemic “house flora” being present in the facility. Another study providing evidence of S. enterica persistence in a food-related environment was conducted by Martinez-Urtaza et al. who probed the genetic relatedness of S. enterica isolates from eight mussel processing facilities with a subset of human, animal, feed, and environmental isolates (235). PFGE subtyping of these isolates suggested persistence of a specific pulsotype, X19. This strain was also the dominant pulsotype in one processing facility over the course of the five-year study. One of the most compelling pieces of evidence for environmental persistence of S. enterica originates from two temporally distinct outbreaks where salmonellosis was associated with consumption of dried cereals produced in the same processing facility (369, 375). The two outbreaks, which occurred ten years apart, involved an identical pulsotype of S. Agona. The outbreak strain was isolated from environmental samples taken in the facility, and it was concluded that a construction project in the facility may

Texas Tech University, Alex Brandt, May 2014 have allowed reintroduction of the strain into the production area. As a result, the authors emphasized the importance of good cleaning and sanitation, even in low-moisture food processing operations, as S. enterica may still be able to persist under these conditions. Evidence for Persistence of E. coli O157:H7 Much of the current body of literature that involves the use of molecular methods to probe E. coli O157:H7 “persistence” does so in the context of its longitudinal existence among animals in dairy and beef cattle herds, sheep flocks, and in the surrounding farm environment (e.g.(51, 112, 116, 169, 187, 293)). For instance, Carlson et al. recently demonstrated that certain closely related subtypes of E. coli O157:H7 were able to persist in the environment of a cattle feedlot facility for a period of more than two years, and that these strains subsequently persisted in the gastrointestinal tract of animals housed on the operation for 110 d (51). Similarly, a molecular subtyping study by Lahti et al. revealed that a Finnish cattle operation was colonized with a certain E. coli O157:H7 PFGE pulsotype that accounted for 79.6% of the isolates (197). The authors concluded that the presence of the pathogen on water cups, floors, and feed passages were indicative of its ability to persist on barn surfaces for extended periods of time. Likewise, LeJeune et al. demonstrated that E. coli O157:H7 belonging to a select few restriction endonuclease digestion pattern (REDP) subtypes persisted in a farm environment over an entire study period despite turnover in the population of cattle on the farm (208). This observation suggested that persistence in the farm environment, not cattle introduction, was the main means by which the pathogen was transferred to incoming animals. Also, the results of a study by Shere et al. suggested that an E. coli O157:H7 strain of a certain REDP type persisted on a farm for more than two years, and that a contaminated drinking water source likely served as a focal point for dissemination throughout the herd (328). In a situation similar to that of S. enterica, studies which probe the molecular ecology of E. coli O157:H7 in food processing facility environments are considerably limited in number, dictating the need for such research. However, glimpses of environmental transmission of E. coli O157:H7 to food products have been given by a handful of studies. A study by Barkocy-Gallagher et al., which sought to identify transmission patterns of E. coli O157:H7 on carcasses from pre- to post-harvest using PFGE (20), demonstrated that within lots nearly 31.8% of isolates recovered from the 32

Texas Tech University, Alex Brandt, May 2014 post-harvest carcass did not match those recovered from the pre-harvest live animal. This observation seems to suggest that, along with contamination from the hide, contamination from the processing environment may have contributed to the composition of the final E. coli O157:H7 population on the carcass. Gun et al. investigated E. coli O157:H7 contamination of carcasses and the processing environment in a Turkish cattle abattoir, and were able to recover the pathogen from 3.6% carcasses, as well as in two knife samples, two employee hand samples, one apron sample, and one floor sample (142). This also suggests that environmental sites may serves as transfer points for the pathogen during slaughter. Likewise, Aslam et al. conducted sampling and molecular subtyping of E. coli O157:H7 isolates from a Canadian beef-packing plant using RAPD (13). The isolation of the same subtypes of E. coli O157:H7 in ground beef as those found in carcass breaking equipment and conveyors further demonstrates the role of environmental sources in the contamination of beef products. The presence of E. coli O157:H7 has also been noted in food processing environments other than those that process beef. For instance, Cagri-Mehmetoglu et al. isolated the pathogen in environmental sites of a kasar cheese processing facility (47). The pathogen was isolated from a raw milk sample, but was also isolated from the hands and gloves of workers, as well as the drains and the floors of the facility. Though it did not involve specific isolation of O157:H7, a large three-year assessment of the microbial ecology of four different U.K. RTE chilled food processing facilities may provide the best example of E. coli’s ability to persist for extended periods of time in a food processing environment (157). Over the course of the study, ten different ribogroups of generic E. coli showed evidence for persistence, as defined by the repeated isolation of the same strain, from the same processing site, over an extended period of time. Though much of this persistence was product oriented, this observation still demonstrates the potential for E. coli (including O157:H7) to be disseminated by environmental point sources where it causes recurrent contamination of products. Collectively, these studies reinforce that proper sanitation and employee hygiene are critical to mitigate the transfer of the pathogen once it enters a processing environment. Challenges in Defining Persistence Despite the extensive documentation of pathogen persistence in various 33

Texas Tech University, Alex Brandt, May 2014 environments, current knowledge of the phenomenon is still fragmented. It is agreed that one of the biggest obstacles preventing a clear understanding of persistence (as well as means to mitigate it) is an inconsistent definition of persistent strains (230). Table 1.1 provides an excellent demonstration of this conundrum. Studies on persistence are often conducted in facilities producing diverse food products, employing a wide array of subtyping methods, using different sampling frequencies, and can span various periods of time. This alone provides substantial inter-study inconsistencies as different subtyping methods provide different levels of discriminatory power, different sampling frequencies may yield different presumed time spans for persistence, and different food processing and sanitation procedures have different inputs and flows. However, the discrepancies in the criteria set by the authors to define a persistent strain oftentimes contribute even more confusion. For instance, Dauphin et al. strictly defined a persistent L. monocytogenes strain as one that was “isolated on six different sampling dates in a 2-month interval (79),” while Lundén et al. used a more lenient definition of “found repeatedly in the plant during a period of several months or years (221).” What’s more, the majority of studies do not state the criteria that they use to delineate a persistent strain. Uncertainty regarding whether persistent strains actually reside in the processing environment or are continually reintroduced (by raw product, employees, or other contaminated materials entering the facility) further confounds definitions of persistence. Also, the isolation of clones which seem to predominate in the facility (and are presumed to be persistent) may solely be influenced by the overall historical prevalence of those subtypes in the region (261, 305). This increased prevalence may elevate the likelihood that subtypes would be isolated by chance alone, and not necessarily due to their persistent status. Furthermore, sampling procedures and enrichment selectivity may also influence the strains that are isolated (44). Taken together, these minor individual issues can interact to create a major distortion of the perceived persistence of a subtype and warrant consideration in defining persistence. Elucidating Risk Factors for Persistence Despite repeated measures to genotypically and phenotypically differentiate persistent and sporadically isolated subtypes, no risk factors unique to persistent subtypes have been consistently identified. One should be keenly aware that this inconsistency in 34

Texas Tech University, Alex Brandt, May 2014 risk factor identification may be a direct consequence of inconsistent persistence definitions and classifications, which cause results to be incoherent; however the current body of knowledge is still worthy of discussion. To date, most phenotypic and genotypic characterization studies have focused on the ability of persistent strains to 1) attach to abiotic surfaces and form biofilms, 2) tolerate the action of sanitizers and antimicrobials, and 3) tolerate stresses such as heat, cold, acid, and low aw. Current knowledge regarding the relationships of these characteristics to strain persistence will be discussed below. Biofilms are bacterial communities embedded within complex three-dimensional extracellular matrices (usually comprised of exopolysaccharide), which are adherent to each other and to surfaces and interfaces (73). These communities can be comprised of diverse members, and exhibit alternative phenotypes in terms of growth rates and gene transcription with respect to planktonic cells (91). Because biofilms impart resistance to mechanical removal from abiotic surfaces and tolerance to chemical stresses (e.g.(122, 195)), the ability of a strain to attach to surfaces and induce biofilm formation has long been considered a potential contributor to its persistence in processing facilities. Attachment of L. monocytogenes, S. enterica, and E. coli O157:H7 to a variety of surface materials common to food processing environments, such as stainless steel, polypropylene, high density polyethylene, polystyrene, glass, and rubber has been documented (25, 109, 148, 228). Initial attachment is thought to be mediated by surface charge interactions between the chemical groups of abiotic materials and extracellular molecules such as curli fimbriae, type I fimbriae, and flagella (60, 72, 334, 401). Components of food soils may aid in this attachment (8), and may further protect the organism’s removal through cleaning (148). Stages of biofilm development that occur after attachment include proliferation of cells to form layers, and formation of channels through which water and nutrients flow (73). Once fully mature, a biofilm can generate large amounts of planktonic cells which can be released to contaminate food and food handling equipment. It should be noted however, that there is an ongoing debate about whether foodborne pathogens (L. monocytogenes in particular) actually exhibit all the complex features of biofilm formation, and whether the perceived formation of biofilms in experimental settings is actually just strong cellular attachment to abiotic surfaces (106). Indeed, a study by Kalmokoff et al. demonstrated through scanning electron

Texas Tech University, Alex Brandt, May 2014 microscopy that only 1 of 36 strains of L. monocytogenes actually formed a classic biofilm on stainless steel (177). Thus, the term “adherence” will be used in place of “biofilm formation” going forward in order to comply with this observation. Studies that have examined the concordance of abiotic surface adherence and persistence status for foodborne pathogens have provided mixed conclusions. A study by Vestby et al. compared adherence ability of persistent and non-persistent strains of S. enterica using a polystyrene microtiter plate assay and found that persistent strains adhered better (406). A study which used L. monocytogenes isolates collected from dairy milking facilities found that predominant and persistent strains had significantly more adherence to a polyvinyl chloride microtiter plate than strains that were either only persistent, predominant, or sporadic (199). Similarly, persistent L. monocytogenes strains have been shown to achieve higher cell density in a given area when grown in TSBYE than non-persistent strains (37), and display up to two to 11 times greater adherence than non-persistent strains within a 1 to 2 hour time frame (225). However, the latter study (225) also demonstrated that adherence of persistent and non-persistent strains reached the same levels after 72 h. The results from a study by Harvey et al. also showed no difference in adherence (via microtiter plate and petri dish adherence assays) between persistent and non-persistent strains of L. monocytogenes from food, animal, environmental and clinical sources (146). Cruz et al. also showed no significant difference in adherence among persistent and non-persistent L. monocytogenes isolates from mussel processing facilities (75). Likewise, a study by Djordjevic et al. demonstrated no significant difference in adhered cell density between persistent and non-persistent strains when grown in modified Welshimer’s broth (87), providing further evidence that strain adherence and persistence status may not be correlated. Overall, with the currently available evidence it seems that a clear relationship between persistence status and adherence ability cannot be drawn. While it is possible that no relationship exists between the two variables, it is also argued that the adherence conditions in these in vitro tests are overly simplified. Bacterial adherence requires complex interactions with a surface, and the absence of food residues, background microflora, and complexities of abiotic surface textures and chemistries normally encountered in a food plant might preclude the observation of the true potential of strain

Texas Tech University, Alex Brandt, May 2014 adherence in in vitro studies. This is a valid argument, especially with regard to the interaction of background flora. Strains of Flavobacterium (39), Kocuria varians (53), Staphylococcus capitis (53), Stenotrophomonas maltophilia (53), Comamonas testosteroni (53), and Pseudomonas fragi (310) have all been shown to have a positive effect on L. monocytogenes adherence to abiotic surfaces. Thus, studies which more accurately mimic the complexities of a food processing environment may provide a clearer perspective of the relationship between persistence and adherence, and may offer insight into effective strategies to mitigate adherence and subsequent strain persistence. Along with surface adherence, tolerance of foodbore pathogens to chemical disinfectants commonly used in food processing facilities has also been considered a key contributor to strain persistence. Some of the most common chemical disinfectants used in the food industry for sanitation purposes include chlorine- and iodine-based sanitizers, quaternary ammonium compounds, and peroxyacetic acids, all of which have varying cellular targets and modes of action. As long as their application is preceded by thorough cleaning and removal of heavy soils, proper use of sanitizers at approved bactericidal concentrations will ensure bacteria are eliminated, and no evidence of populations that are truly “resistant” to full-strength sanitizer concentrations has been documented to date (35). However, “tolerance” via adaptation to sanitizer exposure has been noted (54, 203), and is generally observed when bacteria are exposed to sub-lethal concentrations much lower than the recommended bactericidal concentration. This situation can occur when the disinfectant is diluted out upon surfaces during application or when a pathogen is residing within a harborage site that cannot be penetrated by the full-strength sanitizer solution. Though investigators have probed the correlation between strain persistence status and sanitizer tolerance, a definitive relationship has yet to be established. A weak indication that persistence might be tied to disinfectant tolerance was given by Aase et al. who found that 20 isolates of L. monocytogenes originating from food-related environments (including one persistent strain isolated from a fish processing plant) had minimum inhibitory concentrations (MICs) of benzalkonium chloride that were twice that of another subset of 174 strains (1). Fox et al. also found that tolerance of two persistent strains of L. monocytogenes to quaternary ammonium compounds was higher than that for two non-persistent strains, and that this tolerance involved a complex

Texas Tech University, Alex Brandt, May 2014 transcriptional response centered on enhanced peptidoglycan biosynthesis (111). However, studies showing no apparent relationship have also been published. Lundén et al. noted higher initial MICs of two persistent L. monocytogenes strains to quaternary ammonium compounds, one tertiary alkylamine, sodium hypochlorite, and potassium persulfate, when compared to non-persistent strains (221). However, all strains reached approximately similar MICs after a short term exposure to the compounds, and this tolerance lasted for up to 28 days for three compounds. Both Earnshaw and Lawrence (98) and Holah et al. (158) investigated the ability of persistent and non-persistent L. monocytogenes strains to tolerate the application of quaternary ammonium and alkali metal hydroxide sanitizers, but found no correlations. Similarly, Heir et al. found no links between tolerance to benzalkonium chloride and the persistence status of L. monocytogenes isolates as determined by PFGE (147). Furthermore, it seems that the attachment status of strains to abiotic surfaces during these experiments may further complicate the interpretation of in vitro results obtained for sanitizer tolerance. Kastbjerg and Gram allowed both persistent and non- persistent strains of L. monocytogenes to attach to stainless steel coupons or remain in a planktonic state before exposure to two disinfectants, and found no differences based on attachment status or persistence (181). However, Skandamis et al. observed that strains attached to agar surfaces had much better tolerance to acid treatment than cells in a liquid medium (330). Therefore, it is apparent that current studies do not provide a clear picture of the relationship between persistence and sanitizer tolerance, and that experimental conditions may further complicate the interpretation of the validity of observed results. Tolerance of foodborne pathogens to stresses commonly encountered in the food processing chain, such as nutrient starvation, heat, cold, acid, and low aw have also been postulated to have a role in persistence. In L. monocytogenes, stress responses to stimuli such as acid and oxidative stress, osmotolerance, and carbon starvation are largely controlled by the alternative sigma factor σB (105, 273, 414), whose expression is activated by an extracellular stimulus and a pathway that is mediated by RsbV and RsbT (58). Following activation of σB, the transcription of an array of stress response proteins is upregulated. Because the σB regulon includes a diverse set of stress response proteins, stimulation that is triggered by one stress may also provide cross-protection against other

Texas Tech University, Alex Brandt, May 2014 stresses that are subsequently encountered (28, 124). In addition to σB, other stress response mechanisms such as the SOS response (396, 397), and transcriptional regulators such as σH (296), σL (294), and σC (422) have also been found to play a role in tolerance to high and low temperatures, acid stress, high osmolarity, alkaline conditions, and bile stress in L. monocytogenes. The counterpart to σB in S. enterica and E. coli is the alternative sigma factor RpoS or σS (22, 89). Genes under the regulation of RpoS include those that are also involved in response to stresses encountered in food processing environments such as heat, osmotic stress, acid stress, oxidative stress, and nutrient starvation (63, 96, 149, 207, 300). Enhanced activity of these stress response regulators in persistent strains would be logical given their long-term colonization of processing facilities where these stresses are regularly encountered, and consequently has been considered worthy of investigation. However, as with surface adherence and disinfectant tolerance, studies on the connection between persistence and enhanced stress responses in foodborne pathogens have provided conflicting results. Porsby, et al. evaluated the fates of a persistent fish smokehouse isolate of L. monocytogenes and an isolate of the EGD clinical reference strain when exposed to the hurdles of salting, drying, and cold smoking (289). No differences were observed between the persistent strain and clinical strain, leading the authors to conclude that stress response mechanisms were not particularly contributing to the fish smokehouse strain’s persistence. Lundén et al. did find a significantly higher (P = 0.02) acid tolerance for persistent strains when compared to non-persistent strains, but did not observe the same trend with regard to heat tolerance (222). Likewise, Ringus et al. used quantitative reverse-transcription PCR to compare transcript levels of σB and CtsR (an L. monocytogenes stress response repressor) in persistent and non-persistent strains from fish and meat processing environments following exposure to salt stress (299). The authors found no significant differences between the strains, and concluded that their results were consistent with an emerging theory that establishment of L. monocytogenes strain persistence is largely stochastic in nature (52). This theory mentioned by Ringus et al. is one recently proposed by Carpentier and Cerf (52). In their review, the authors include an array of studies that probe the linkage between persistence and the factors traditionally thought to contribute to it (i.e. biofilm

Texas Tech University, Alex Brandt, May 2014 formation, disinfectant tolerance, contribution of background flora, etc.). Based on the heterogeneity in the study outcomes, the authors ultimately concluded that no strains of L. monocytogenes have unique properties that allow them to persist in food processing environments. Their alternate explanation for persistence is that there are only unique harborage sites that provide favorable conditions where strains can establish persistence. Furthermore, they note that because persistence seems to be largely based on favorable growth rather than survival of stresses or surface attachment, any strain could theoretically establish persistence as long as it encounters a suitable harborage site in a food processing facility that favors its growth needs. This theory therefore re-emphasizes the need to base environmental monitoring programs for Listeria on aggressive “seek- and-destroy” approaches (46) that proactively address contamination at these sites through routine environmental sampling and testing. Admittedly, this theory proposed by Carpentier and Cerf may be entirely true, and persistence may be explained simply by specific environmental risk factors with no unique traits observable in the bacterium. Few studies have actually probed facility- specific risk factors rather than bacterium-specific risk factors, so this knowledge is most definitely lacking. However, the Carpentier and Cerf theory still seems to be overly simplistic. As is concluded by Ferreira et al. (106), evidence still points to an increased likelihood for certain strains to establish persistence. Thus, Ferreira et al. suggest that persistence is more likely the culmination of a complex but harmonious interplay between unique pathogen-specific traits and a multitude of facility-specific risk factors. Consequently, accounting for both the facility and bacterial ends of the persistence spectrum in future research may provide a clearer picture of the phenomenon and subsequently may assist in the development of effective methods to counteract it. Another novel theory with regard to the establishment of L. monocytogenes persistence is one that focuses on rapid evolution via recombination in the prophage that is often embedded within the comK gene of the pathogen (404). Verghese et al. sequenced different regions of the comK prophage in several EC strains and found that the sequences were specific to individual processing plants, suggesting that the region may serve as a “rapid adaptation island” where extensive recombination quickly provides specificity to the conditions of the facility as the prophage region evolves. Furthermore,

Texas Tech University, Alex Brandt, May 2014 strains that harbored a prophage had significantly higher cell densities when attached to food conditioning films than strains that lacked a prophage, possibly helping explain persistence of these strains. What is interesting is that this theory is also in alignment with the results of Orsi et al. who demonstrated that, over a 12 year span, the same L. monocytogenes strain, which caused two separate outbreaks, had very little change in the backbone genome, but had extensive differences in the comK prophage sequence (275). Much still remains to be understood about this theory, but it holds promise for helping to explain the persistence phenomenon. Establishing Consistent Definitions of Persistence The availability of an improved definition of strain persistence for future research studies will likely aid in the elucidation of relationships between bacterial characteristics and facility risk factors that are believed to be associated with strain persistence. To address the shortcomings and inconsistencies in defining persistent strains, Malley, et al. recently proposed two statistical approaches to defining subtype persistence (230). The first is a statistical test that compares the frequency of a subtype in a particular processing facility (or one site within the processing facility) and the frequency of the subtype in a reference distribution, such as the collection of isolates indexed in the Food Microbe Tracker Database (400). While this method is easily biased by the origins and diversity of the isolates included in the reference distribution, it helps eliminate the influence of historical strain prevalence in arriving at a determination of persistence. In other words, for a subtype to be considered persistent, the likelihood of isolating the subtype from the collection of isolates gathered from a facility (or site) must be greater than the likelihood of isolating that particular subtype from the general population of isolates that the reference distribution (or a suitable subset of the reference distribution) is taken from. Thus, if a greater probability of isolating a strain from a facility as compared to the reference distribution is found, it should indicate that the strain is uniquely associated with the facility, and is not being isolated more frequently than others by chance alone as a result of its high prevalence in the surrounding environment. Likewise, Malley, et al. also proposed a binomial test for persistence that evaluates the likelihood of recovering certain subtypes in terms of a sequence of events. To carry out the test, a two-by-two table is constructed and statistically analyzed (via 41

Texas Tech University, Alex Brandt, May 2014

Chi-Squared or Fisher’s Exact tests) where the A cell contains the number of sample sites that were both currently positive and positive at the immediately previous sampling time (++), the B cell contains the number of sample sites that were currently positive but previously negative (+-), the C cell contains the number of sample sites that were currently negative but previously positive (-+), and the D cell contains the number of sample sites that were both currently and previously negative (--). If the analysis reveals that a previous positive result with a given subtype is a significant risk factor for a current positive result with the same subtype, the subtype should therefore be interpreted as being persistent. Because the test is not based on a reference distribution, it is more robust with regard to bias, but is less appropriate for use at an individual site level where the small number of observations precludes proper analysis. Though they cannot completely define persistence, these tests certainly improve the definition and will allow better inter-study comparison. The application of these tests to subtyping data for L. monocytogenes, non-pathogenic Listeria species, S. enterica, and E. coli O157:H7 collected via longitudinal sampling from meat processing facilities will be a core element of the following studies that are included in this dissertation.

References 1. Aase B, Sundheim G, Langsrud S, Rørvik LM. 2000. Occurrence of and a possible mechanism for resistance to a quaternary ammonium compound in Listeria monocytogenes. Int. J. Food Microbiol. 62:57-63.

2. Abbott SL, O'Connor J, Robin T, Zimmer BL, Janda JM. 2003. Biochemical properties of a newly described Escherichia species, Escherichia albertii. J. Clin. Microbiol. 41:4852-4854.

3. Aberle ED, Forrest JC, Gerrard DE, Mills EW, (ed.). 2001. Principles of meat science, 4th ed. Kendall Hunt, Dubuque, IA.

4. Abram M, Doric M. 1997. Primary Listeria monocytogenes infection in gestating mice. Fol. Microbiol. 42:65-71.

5. Adélia M, Ferreira SS, Lund BM. 1987. The influence of pH and temperature on initiation of growth of Salmonella spp. Lett. Appl. Microbiol. 5:67-70.

Texas Tech University, Alex Brandt, May 2014

6. Agbaje M, Begum RH, Oyekunle MA, Ojo OE, Adenubi OT. 2011. Evolution of Salmonella nomenclature: a critical note. Fol. Microbiol. 56:497-503.

7. Aguado V, Vitas AI, García-Jalón I. 2004. Characterization of Listeria monocytogenes and Listeria innocua from a vegetable processing plant by RAPD and REA. Int. J. Food Microbiol. 90:341-347.

8. al-Makhlafi H, McGuire J, Daeschel M. 1994. Influence of preadsorbed milk proteins on adhesion of Listeria monocytogenes to hydrophobic and hydrophilic silica surfaces. Appl. Environ. Microbiol. 60:3560-3565.

9. Alcaine SD, Warnick LD, Wiedmann M. 2007. Antimicrobial resistance in nontyphoidal Salmonella. J. Food Prot. 70:780-790.

10. Almeida G, Magalhaes R, Carneiro L, Santos I, Silva J, Ferreira V, Hogg T, Teixeira P. 2013. Foci of contamination of Listeria monocytogenes in different cheese processing plants. Int. J. Food Microbiol. 167:303-309.

11. Alvarez-Domínguez C, Vázquez-Boland JA, Carrasco-Marín E, López-Mato P, Leyva-Cobián F. 1997. Host cell heparan sulfate proteoglycans mediate attachment and entry of Listeria monocytogenes, and the listerial surface protein ActA is involved in heparan sulfate receptor recognition. Infect. Immun. 65:78- 88.

12. Angelotti R, Foter MJ, Lewis KH. 1961. Time-temperature effects on salmonellae and staphylococci in foods. I. behavior in refrigerated foods. II. behavior at warm holding temperatures. Am. J. Public Health Nat. Health 51:76- 88.

13. Aslam M, Greer GG, Nattress FM, Gill CO, McMullen LM. 2004. Genotypic analysis of Escherichia coli recovered from product and equipment at a beef- packing plant. J. Appl. Microbiol. 97:78-86.

14. Asplund K, Nurmi E. 1991. The growth of salmonellae in tomatoes. Int. J. Food Microbiol. 13:177-181.

15. Aureli P, Fiorucci GC, Caroli D, Marchiaro G, Novara O, Leone L, Salmaso S. 2000. An outbreak of febrile gastroenteritis associated with corn contaminated by Listeria monocytogenes. N. Eng. J. Med. 342:1236-1241.

16. Autio T, Hielm S, Miettinen M, Sjoberg AM, Aarnisalo K, Bjorkroth J, Mattila-Sandholm T, Korkeala H. 1999. Sources of Listeria monocytogenes

Texas Tech University, Alex Brandt, May 2014

contamination in a cold-smoked rainbow trout processing plant detected by pulsed-field gel electrophoresis typing. Appl. Environ. Microbiol. 65:150-155.

17. Bakowski MA, Braun V, Brumell JH. 2008. Salmonella-containing vacuoles: directing traffic and nesting to grow. Traffic 9:2022-2031.

18. Banerji A, Noya FJD. 1999. Brain abscess associated with neonatal listeriosis. Pediatr. Infect. Dis. J. 18:305-307.

19. Barancelli GV, Camargo TM, Gagliardi NG, Porto E, Souza RA, Campioni F, Falcão JP, Hofer E, Cruz AG, Oliveira CAF. 2014. Pulsed-field gel electrophoresis characterization of Listeria monocytogenes isolates from cheese manufacturing plants in São Paulo, Brazil. Int. J. Food Microbiol. 173:21-29.

20. Barkocy-Gallagher GA, Arthur TM, Siragusa GR, Keen JE, Elder RO, Laegreid WW, Koohmaraie M. 2001. Genotypic analyses of Escherichia coli O157:H7 and O157 nonmotile isolates recovered from beef cattle and carcasses at processing plants in the midwestern states of the United States. Appl. Environ. Microbiol. 67:3810-3818.

21. Barrett TJ, Gerner-Smidt P. 2007. Molecular source tracking and molecular subtyping, p. 987-1004. In Doyle MP, Beuchat LR (ed.), Food microbiology: fundamentals and frontiers, 3rd ed. ASM Press, Washington, DC.

22. Battesti A, Majdalani N, Gottesman S. 2011. The RpoS-mediated general stress response in Escherichia coli. Ann. Rev. Microbiol. 65:189-213.

23. Baumler AJ, Tsolis RM, Bowe FA, Kusters JG, Hoffmann S, Heffron F. 1996. The pef fimbrial operon of Salmonella typhimurium mediates adhesion to murine small intestine and is necessary for fluid accumulation in the infant mouse. Infect. Immun. 64:61-68.

24. Baumler AJ, Tsolis RM, Heffron F. 1996. The lpf fimbrial operon mediates adhesion of Salmonella typhimurium to murine Peyer's patches. Proc. Nat. Acad. Sci. USA 93:279-283.

25. Beauchamp CS, Dourou D, Geornaras I, Yoon Y, Scanga JA, Belk KE, Smith GC, Nychas G-JE, Sofos JN. 2012. Transfer, attachment, and formation of biofilms by Escherichia coli O157:H7 on meat-contact surface materials. J. Food Sci.:M1-M5.

26. Behravesh CB, Ferraro A, Deasy M, Dato V, Moll M, Sandt C, Rea NK, Rickert R, Marriott C, Warren K, Urdaneta V, Salehi E, Villamil E, Ayers T,

Texas Tech University, Alex Brandt, May 2014

Hoekstra RM, Austin JL, Ostroff S, Williams IT. 2010. Human Salmonella infections linked to contaminated dry dog and cat food, 2006-2008. Pediatrics 126:477-483.

27. Bell BP, Griffin PM, Lozano P, Christie DL, Kobayashi JM, Tarr PI. 1997. Predictors of hemolytic uremic syndrome in children during a large outbreak of Escherichia coli O157:H7 infections. Pediatrics 100:e12.

28. Bergholz TM, Bowen B, Wiedmann M, Boor KJ. 2012. Listeria monocytogenes shows temperature-dependent and -independent responses to salt stress, including responses that induce cross-protection against other stresses. Appl. Environ. Microbiol. 78:2602-2612.

29. Berrang ME, Meinersmann RJ, Frank JF, Ladely SR. 2010. Colonization of a newly constructed commercial chicken further processing plant with Listeria monocytogenes. J. Food Prot. 73:286-291.

30. Berrang ME, Meinersmann RJ, Frank JF, Smith DP, Genzlinger LL. 2005. Distribution of Listeria monocytogenes subtypes within a poultry further processing plant. J. Food Prot. 68:980-985.

31. Bērziᶇš A, Hellström S, Siliᶇš I, Korkeala H. 2010. Contamination patterns of Listeria monocytogenes in cold-smoked pork processing. J. Food Prot. 73:2103- 2109.

32. Besser RE, Lett SM, Weber JT, Doyle MP, Barrett TJ, Wells JG, Griffin PM. 1993. An outbreak of diarrhea and hemolytic uremic syndrome from Escherichia coli O157:H7 in fresh-pressed apple cider. J. Am. Med. Assoc. 269:2217-2220.

33. Beuchat LR, Komitopoulou E, Beckers H, Betts RP, Bourdichon F, Fanning S, Joosten HM, Ter Kuile BH. 2013. Low-water activity foods: increased concern as vehicles of foodborne pathogens. J. Food Prot. 76:150-172.

34. Blatter S, Giezendanner N, Stephan R, Zweifel C. 2010. Phenotypic and molecular typing of Listeria monocytogenes isolated from the processing environment and products of a sandwich-producing plant. Food Contr. 21:1519- 1523.

35. Bodnaruk P. 2011. Is it necessary to rotate sanitizer use in food processing operations. Qual. Assur. 2:66.

Texas Tech University, Alex Brandt, May 2014

36. Bonifield HR, Hughes KT. 2003. Flagellar phase variation in Salmonella enterica is mediated by a posttranscriptional control mechanism. J. Bacteriol. 185:3567-3574.

37. Borucki MK, Peppin JD, White D, Loge F, Call DR. 2003. Variation in biofilm formation among strains of Listeria monocytogenes. Appl. Environ. Microbiol. 69:7336-7342.

38. Boxrud D. 2010. Advances in subtyping methods of foodborne disease pathogens. Curr. Opin. Biotech. 21:137-141.

39. Bremer PJ, Monk I, Osborne CM. 2001. Survival of Listeria monocytogenes attached to stainless steel surfaces in the presence or absence of Flavobacterium spp. J. Food Prot. 64:1369-1376.

40. Brenner FW, Villar RG, Angulo FJ, Tauxe R, Swaminathan B. 2000. Salmonella nomenclature. J. Clin. Microbiol. 38:2465-2467.

41. Brigotti M, Tazzari PL, Ravanelli E, Carnicelli D, Barbieri S, Rocchi L, Arfilli V, Scavia G, Ricci F, Bontadini A, Alfieri RR, Petronini PG, Pecoraro C, Tozzi AE, Caprioli A. 2010. Endothelial damage induced by Shiga toxins delivered by neutrophils during transmigration. J. Leukoc. Biol. 88:201-210.

42. Brooks JT, Sowers EG, Wells JG, Greene KD, Griffin PM, Hoekstra RM, Strockbine NA. 2005. Non-O157 Shiga toxin-producing Escherichia coli infections in the United States, 1983-2002. J. Infect. Dis. 192:1422-1429.

43. Bruce J. 1996. Automated system rapidly identifies and characterizes microorganisms in food. Food Technol. 50:77-81.

44. Bruhn JB, Vogel BF, Gram L. 2005. Bias in the Listeria monocytogenes enrichment procedure: lineage 2 strains outcompete lineage 1 strains in University of Vermont selective enrichments. Appl. Environ. Microbiol. 71:961-967.

45. Buchwald DS, Blaser MJ. 1984. A review of human salmonellosis: II. duration of excretion following infection with nontyphi Salmonella. Rev. Infect. Dis. 6:345-356.

46. Butts J. 2003. Seek & destroy: identifying and controlling Listeria monocytogenes growth niches. Available At: http://www.foodsafetymagazine .com/magazine-archive1/aprilmay-2003/seek-destroy-identifying-and-controlling- listeria-monocytogenes-growth-niches. Accessed On: 2/13/14

Texas Tech University, Alex Brandt, May 2014

47. Cagri-Mehmetoglu A, Yaldirak G, Bodur T, Simsek M, Bozkir H, Eren NM. 2011. Incidence of Listeria monocytogenes and Escherichia coli O157:H7 in two kasar cheese processing environments. Food Contr. 22:762-766.

48. Calderwood SB, Auclair F, Donohuerolfe A, Keusch GT, Mekalanos JJ. 1987. Nucleotide sequence of the Shiga-like toxin genes of Escherichia coli. Proc. Nat. Acad. Sci. USA 84:4364-4368.

49. Callaway TR, Morrow JL, Johnson AK, Dailey JW, Wallace FM, Wagstrom EA, McGlone JJ, Lewis AR, Dowd SE, Poole TL, Edrington TS, Anderson RC, Genovese KJ, Byrd JA, Harvey RB, Nisbet DJ. 2005. Environmental prevalence and persistence of Salmonella spp. in outdoor swine wallows. Foodborne Pathog. Dis. 2:263-273.

50. Campellone KG, Robbins D, Leong JM. 2004. EspFU is a translocated EHEC effector that interacts with Tir and N-WASP and promotes Nck-independent actin assembly. Dev. Cell 7:217-228.

51. Carlson BA, Nightingale KK, Mason GL, Ruby JR, Choat WT, Loneragan GH, Smith GC, Sofos JN, Belk KE. 2009. Escherichia coli O157:H7 strains that persist in feedlot cattle are genetically related and demonstrate an enhanced ability to adhere to intestinal epithelial cells. Appl. Environ. Microbiol. 75:5927- 5937.

52. Carpentier B, Cerf O. 2011. Review - persistence of Listeria monocytogenes in food industry equipment and premises. Int. J. Food Microbiol. 145:1-8.

53. Carpentier B, Chassaing D. 2004. Interactions in biofilms between Listeria monocytogenes and resident microorganisms from food industry premises. Int. J. Food Microbiol. 97:111-122.

54. Cerf O, Carpentier B, Sanders P. 2010. Tests for determining in-use concentrations of antibiotics and disinfectants are based on entirely different concepts: “resistance” has different meanings. Int. J. Food Microbiol. 136:247- 254.

55. Chan MS, Maiden MC, Spratt BG. 2001. Database-driven multi locus sequence typing (MLST) of bacterial pathogens. Bioinform. 17:1077-1083.

56. Chasseignaux E, Gérault P, Toquin M-T, Salvat G, Colin P, Ermel G. 2002. Ecology of Listeria monocytogenes in the environment of raw poultry meat and raw pork meat processing plants. FEMS Microbiol. Lett. 210:271-275.

Texas Tech University, Alex Brandt, May 2014

57. Chasseignaux E, Toquin MT, Ragimbeau C, Salvat G, Colin P, Ermel G. 2001. Molecular epidemiology of Listeria monocytogenes isolates collected from the environment, raw meat and raw products in two poultry- and pork-processing plants. J. Appl. Microbiol. 91:888-899.

58. Chaturongakul S, Boor KJ. 2004. RsbT and RsbV contribute to σB-dependent survival under environmental, energy, and intracellular stress conditions in Listeria monocytogenes. Appl. Environ. Microbiol. 70:5349-5356.

59. Chaudhuri S, Bruno JC, Alonzo F, III, Xayarath B, Cianciotto NP, Freitag NE. 2010. Contribution of chitinases to Listeria monocytogenes pathogenesis. Appl. Environ. Microbiol. 76:7302-7305.

60. Chavant P, Martinie B, Meylheuc T, Bellon-Fontaine MN, Hebraud M. 2002. Listeria monocytogenes LO28: surface physicochemical properties and ability to form biofilms at different temperatures and growth phases. Appl. Environ. Microbiol. 68:728-737.

61. Chen BY, Pyla R, Kim TJ, Silva JL, Jung YS. 2010. Incidence and persistence of Listeria monocytogenes in the catfish processing environment and fresh fillets. J. Food Prot. 73:1641-1650.

62. Chen Y, Zhang W, Knabel SJ. 2007. Multi-virulence-locus sequence typing identifies single nucleotide polymorphisms which differentiate epidemic clones and outbreak strains of Listeria monocytogenes. J. Clin. Microbiol. 45:835-846.

63. Cheville AM, Arnold KW, Buchrieser C, Cheng CM, Kaspar CW. 1996. rpoS regulation of acid, heat, and salt tolerance in Escherichia coli O157:H7. Appl. Environ. Microbiol. 62:1822-1824.

64. Childs KD, Simpson CA, Warren-Serna W, Bellenger G, Centrella B, Bowling RA, Ruby J, Stefanek J, Vote DJ, Choat T, Scanga JA, Sofos JN, Smith GC, Belk KE. 2006. Molecular characterization of Escherichia coli O157:H7 hide contamination routes: feedlot to harvest. J. Food Prot. 69:1240- 1247.

65. Clark WF, Sontrop JM, Macnab JJ, Salvadori M, Moist L, Suri R, Garg AX. 2010. Long term risk for hypertension, renal impairment, and cardiovascular disease after gastroenteritis from drinking water contaminated with Escherichia coli O157:H7: a prospective cohort study. Brit. Med. J. 341:c6020.

Texas Tech University, Alex Brandt, May 2014

66. Cloeckaert A, Schwarz S. 2001. Molecular characterization, spread and evolution of multidrug resistance in Salmonella enterica Typhimurium DT104. Vet. Res. 32:301-310.

67. Cobbold RN, Rice DH, Davis MA, Besser TE, Hancock DD. 2006. Long-term persistence of multi-drug-resistant Salmonella enterica serovar Newport in two dairy herds. J. Am. Vet. Med. Assoc. 228:585-591.

68. Cokes C, France AM, Reddy V, Hanson H, Lee L, Kornstein L, Stavinsky F, Balter S. 2011. Serving high-risk foods in a high-risk setting: survey of hospital food service practices after an outbreak of listeriosis in a hospital. Infect. Contr. Hosp. Epidemiol. 32:380-386.

69. Collazo CM, Galan JE. 1997. The invasion-associated type III system of Salmonella typhimurium directs the translocation of Sip proteins into the host cell. Mol. Microbiol. 24:747-756.

70. Collazo CM, Zierler MK, Galán JE. 1995. Functional analysis of the Salmonella typhimurium invasion genes invI and invJ and identification of a target of the protein secretion apparatus encoded in the inv locus. Mol. Microbiol. 15:25-38.

71. Conner DE, Kotrola JS. 1995. Growth and survival of Escherichia coli O157:H7 under acidic conditions. Appl. Environ. Microbiol. 61:382-385.

72. Cookson AL, Cooley WA, Woodward MJ. 2002. The role of type 1 and curli fimbriae of Shiga toxin-producing Escherichia coli in adherence to abiotic surfaces. Int. J. Med. Microbiol. 292:195-205.

73. Costerton JW, Lewandowski Z, Caldwell DE, Korber DR, Lappin-Scott HM. 1995. Microbial biofilms. Ann. Rev. Microbiol. 49:711-745.

74. Cowden JM. 1997. Scottish outbreak of Escherichia coli O157, November- December 1996. Eurosurveillance 2:1-2.

75. Cruz CD, Fletcher GC. 2011. Prevalence and biofilm-forming ability of Listeria monocytogenes in New Zealand mussel (Perna canaliculus) processing plants. Food Microbiol. 28:1387-1393.

76. D'Aoust J-Y. 1991. Pathogenicity of foodborne Salmonella. Int. J. Food Microbiol. 12:17-40.

Texas Tech University, Alex Brandt, May 2014

77. D'Aoust J-Y. 1991. Psychrotrophy and foodborne Salmonella. Int. J. Food Microbiol. 13:207-215.

78. Dalton CB, Austin CC, Sobel J, Hayes PS, Bibb WF, Graves LM, Swaminathan B, Proctor ME, Griffin PM. 1997. An outbreak of gastroenteritis and fever due to Listeria monocytogenes in milk. N. Engl. J. Med. 336:100-105.

79. Dauphin G, Ragimbeau C, Malle P. 2001. Use of PFGE typing for tracing contamination with Listeria monocytogenes in three cold-smoked salmon processing plants. Int. J. Food Microbiol. 64:51-61.

80. Davies RH, Breslin M. 2003. Persistence of Salmonella Enteritidis Phage Type 4 in the environment and arthropod vectors on an empty free-range chicken farm. Environ. Microbiol. 5:79-84.

81. De Cesare A, Bruce JL, Dambaugh TR, Guerzoni ME, Wiedmann M. 2001. Automated ribotyping using different enzymes to improve discrimination of Listeria monocytogenes isolates, with a particular focus on serotype 4b strains. J. Clin. Microbiol. 39:3002-3005.

82. den Bakker HC, Moreno Switt AI, Govoni G, Cummings CA, Ranieri ML, Degoricija L, Hoelzer K, Rodriguez-Rivera LD, Brown S, Bolchacova E, Furtado MR, Wiedmann M. 2011. Genome sequencing reveals diversification of virulence factor content and possible host adaptation in distinct subpopulations of Salmonella enterica. BMC Genom. 12:425.

83. den Bakker HC, Warchocki S, Wright EM, Allred AF, Ahlstrom C, Manuel CS, Stasiewicz MJ, Burrell A, Roof S, Strawn L, Fortes ED, Nightingale KK, Kephart D, Wiedmann M. 2014. Five new species of Listeria (L. floridensis sp. nov, L. aquatica sp. nov., L. cornellensis sp. nov. L. riparia sp. nov., and L. grandensis sp. nov.) from agricultural and natural environments in the United States. Int. J. Syst. Evol. Microbiol.

84. Deng Y, Ryu JH, Beuchat LR. 1998. Influence of temperature and pH on survival of Escherichia coli O157:H7 in dry foods and growth in reconstituted infant rice cereal. Int. J. Food Microbiol. 45:173-184.

85. Dey M, Mayo JA, Saville D, Wolyniak C, Klontz KC. 2013. Recalls of foods due to microbiological contamination classified by the U.S. Food and Drug Administration, fiscal years 2003 through 2011. J. Food Prot. 76:932-938.

86. Di Ciccio P, Meloni D, Festino AR, Conter M, Zanardi E, Ghidini S, Vergara A, Mazzette R, Ianieri A. 2012. Longitudinal study on the sources of Listeria

Texas Tech University, Alex Brandt, May 2014

monocytogenes contamination in cold-smoked salmon and its processing environment in Italy. Int. J. Food Microbiol. 158:79-84.

87. Djordjevic D, Wiedmann M, McLandsborough LA. 2002. Microtiter plate assay for assessment of Listeria monocytogenes biofilm formation. Appl. Environ. Microbiol. 68:2950-2958.

88. Dodd CC, Renter DG, Shi X, Alam MJ, Nagaraja TG, Sanderson MW. 2011. Prevalence and persistence of Salmonella in cohorts of feedlot cattle. Foodborne Pathog. Dis. 8:781-789.

89. Dodd CE, Aldsworth TG. 2002. The importance of RpoS in the survival of bacteria through food processing. Int. J. Food Microbiol. 74:189-194.

90. Domann E, Wehland J, Rohde M, Pistor S, Hartl M, Goebel W, Leimeisterwachter M, Wuenscher M, Chakraborty T. 1992. A novel bacterial virulence gene in Listeria monocytogenes required for host-cell microfilament interaction with homology to the proline-rich region of vinculin. EMBO J. 11:1981-1990.

91. Donlan RM, Costerton JW. 2002. Biofilms: survival mechanisms of clinically relevant microorganisms. Clin. Microbiol. Rev. 15:167-193.

92. Doumith M, Buchrieser C, Glaser P, Jacquet C, Martin P. 2004. Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. J. Clin. Microbiol. 42:3819-3822.

93. Doyle MP, Schoeni JL. 1984. Survival and growth characteristics of Escherichia coli associated with hemorrhagic colitis. Appl. Environ. Microbiol. 48:855-856.

94. Dramsi S, Biswas I, Maguin E, Braun L, Mastroeni P, Cossart P. 1995. Entry of Listeria monocytogenes into hepatocytes requires expression of InlB, a surface protein of the internalin multigene family. Mol. Microbiol. 16:251-261.

95. Dramsi S, Bourdichon F, Cabanes D, Lecuit M, Fsihi H, Cossart P. 2004. FbpA, a novel multifunctional Listeria monocytogenes virulence factor. Mol. Microbiol. 53:639-649.

96. Dukan S, Touati D. 1996. Hypochlorous acid stress in Escherichia coli: resistance, DNA damage, and comparison with hydrogen peroxide stress. J. Bacteriol. 178:6145-6150.

Texas Tech University, Alex Brandt, May 2014

97. Dunowska M, Morley PS, Traub-Dargatz JL, Davis MA, Patterson G, Frye JG, Hyatt DR, Dargatz DA. 2007. Comparison of Salmonella enterica serotype Infantis isolates from a veterinary teaching hospital. J. Appl. Microbiol. 102:1527-1536.

98. Earnshaw AM, Lawrence LM. 1998. Sensitivity to commercial disinfectants, and the occurrence of plasmids within various Listeria monocytogenes genotypes isolated from poultry products and the poultry processing environment. J. Appl. Microbiol. 84:642-648.

99. Eggertson L. 2009. Federal listeriosis investigation has limited mandate. Can. Med. Assoc. J. 180:503.

100. Eifert JD, Curtis PA, Bazaco MC, Meinersmann RJ, Berrang ME, Kernodle S, Stam C, Jaykus LA, Kathariou S. 2005. Molecular characterization of Listeria monocytogenes of the serotype 4b complex (4b, 4d, 4e) from two turkey processing plants. Foodborne Pathog. Dis. 2:192-200.

101. Endo Y, Tsurugi K, Yutsudo T, Takeda Y, Ogasawara T, Igarashi K. 1988. Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes: RNA N-glycosidase activity of the toxins. Eur. J. Biochem. 171:45-50.

102. Farber JM, Daley E, Coates F, Beausoleil N, Fournier J. 1991. Feeding trials of Listeria monocytogenes with a nonhuman primate model. J. Clin. Microbiol. 29:2606-2608.

103. Farber JM, Daley EM, Mackie MT, Limerick B. 2000. A small outbreak of listeriosis potentially linked to the consumption of imitation crab meat. Lett. Appl. Microbiol. 31:100-104.

104. Fenlon DR, Wilson J, Donachie W. 1996. The incidence and level of Listeria monocytogenes contamination of food sources at primary production and initial processing. J. Appl. Bacteriol. 81:641-650.

105. Ferreira A, O'Byrne CP, Boor KJ. 2001. Role of σB in heat, ethanol, acid, and oxidative stress resistance and during carbon starvation in Listeria monocytogenes. Appl. Environ. Microbiol. 67:4454-4457.

106. Ferreira V, Wiedmann M, Teixeira P, Stasiewicz MJ. 2014. Listeria monocytogenes persistence in food-associated environments: epidemiology, strain characteristics, and implications for public health. J. Food Prot. 77:150-170.

Texas Tech University, Alex Brandt, May 2014

107. Fleming DW, Cochi SL, MacDonald KL, Brondum J, Hayes PS, Plikaytis BD, Holmes MB, Audurier A, Broome CV, Reingold AL. 1985. Pasteurized milk as a vehicle of infection in an outbreak of listeriosis. N. Eng. J. Med. 312:404-407.

108. Fookes M, Schroeder GN, Langridge GC, Blondel CJ, Mammina C, Connor TR, Seth-Smith H, Vernikos GS, Robinson KS, Sanders M, Petty NK, Kingsley RA, Baumler AJ, Nuccio SP, Contreras I, Santiviago CA, Maskell D, Barrow P, Humphrey T, Nastasi A, Roberts M, Frankel G, Parkhill J, Dougan G, Thomson NR. 2011. Salmonella bongori provides insights into the evolution of the salmonellae. PLoS Pathog. 7:e1002191.

109. Fouladkhah A, Geornaras I, Sofos JN. 2013. Biofilm formation of O157 and non-O157 Shiga toxin-producing Escherichia coli and multidrug-resistant and susceptible Salmonella Typhimurium and Newport and their inactivation by sanitizers. J. Food Sci. 78:M880-M886.

110. Fox E, Hunt K, O'Brien M, Jordan K. 2011. Listeria monocytogenes in Irish farmhouse cheese processing environments. Int. J. Food Microbiol. 145S:S39- S45.

111. Fox EM, Leonard N, Jordan K. 2011. Physiological and transcriptional characterization of persistent and nonpersistent Listeria monocytogenes isolates. Appl. Environ. Microbiol. 77:6559-6569.

112. Fremaux B, Prigent-Combaret C, Vernozy-Rozand C. 2008. Long-term survival of Shiga toxin-producing Escherichia coli in cattle effluents and environment: an updated review. Vet. Microbiol. 132:1-18.

113. Gahring LC, Heffron F, Finlay BB, Falkow S. 1990. Invasion and replication of Salmonella typhimurium in animal cells. Infect. Immun. 58:443-448.

114. Gaillard JL, Berche P, Frehel C, Gouin E, Cossart P. 1991. Entry of Listeria monocytogenes into cells is mediated by internalin, a repeat protein reminiscent of surface antigens from Gram-positive cocci. Cell 65:1127-1141.

115. Gaillard JL, Berche P, Mounier J, Richard S, Sansonetti P. 1987. In vitro model of penetration and intracellular growth of Listeria monocytogenes in the human enterocyte-like cell line Caco-2. Infect. Immun. 55:2822-2829.

116. Galland JC, Hyatt DR, Crupper SS, Acheson DW. 2001. Prevalence, antibiotic susceptibility, and diversity of Escherichia coli O157:H7 isolates from a

Texas Tech University, Alex Brandt, May 2014

longitudinal study of beef cattle feedlots. Appl. Environ. Microbiol. 67:1619- 1627.

117. Gaul LK, Farag NH, Shim T, Kingsley MA, Silk BJ, Hyytia-Trees E. 2013. Hospital-acquired listeriosis outbreak caused by contaminated diced celery-- Texas, 2010. Clin. Infect. Dis. 56:20-26.

118. Gebreyes WA, Altier C, Thakur S. 2006. Molecular epidemiology and diversity of Salmonella serovar Typhimurium in pigs using phenotypic and genotypic approaches. Epidemiol. Infect. 134:187-198.

119. Gerner-Smidt P, Ethelberg S, Schiellerup P, Christensen JJ, Engberg J, Fussing V, Jensen A, Jensen C, Petersen AM, Bruun BG. 2005. Invasive listeriosis in Denmark 1994-2003: a review of 299 cases with special emphasis on risk factors for mortality. Clin. Microbiol. Infect. 11:618-624.

120. Gerner-Smidt P, Hise K, Kincaid J, Hunter S, Rolando S, Hyytia-Trees E, Ribot EM, Swaminathan B. 2006. PulseNet USA: a five-year update. Foodborne Pathog. Dis. 3:9-19.

121. Giammanco GM, Pignato S, Mammina C, Grimont F, Grimont PA, Nastasi A, Giammanco G. 2002. Persistent endemicity of Salmonella bongori 48:z(35):-- in southern Italy: molecular characterization of human, animal, and environmental isolates. J. Clin. Microbiol. 40:3502-3505.

122. Giao MS, Keevil CW. 2013. Hydrodynamic shear stress to remove Listeria monocytogenes biofilms from stainless steel and polytetrafluoroethylene surfaces. J. Appl. Microbiol. 114:256-265.

123. Gilmour MW, Graham M, Van Domselaar G, Tyler S, Kent H, Trout-Yakel KM, Larios O, Allen V, Lee B, Nadon C. 2010. High-throughput genome sequencing of two Listeria monocytogenes clinical isolates during a large foodborne outbreak. BMC Genom. 11:120.

124. Giotis ES, Julotok M, Wilkinson BJ, Blair IS, McDowell DA. 2008. Role of sigma B factor in the alkaline tolerance response of Listeria monocytogenes 10403S and cross-protection against subsequent ethanol and osmotic stress. J. Food Prot. 71:1481-1485.

125. Giovannacci I, Ragimbeau C, Queguiner S, Salvat G, Vendeuvre JL, Carlier V, Ermel G. 1999. Listeria monocytogenes in pork slaughtering and cutting plants: use of RAPD, PFGE and PCR-REA for tracing and molecular epidemiology. Int. J. Food Microbiol. 53:127-140.

Texas Tech University, Alex Brandt, May 2014

126. Glass KA, Doyle MP. 1989. Fate of Listeria monocytogenes in processed meat products during refrigerated storage. Appl. Environ. Microbiol. 55:1565-1569.

127. Glass KA, Loeffelholz JM, Ford JP, Doyle MP. 1992. Fate of Escherichia coli O157:H7 as affected by pH or sodium chloride and in fermented, dry sausage. Appl. Environ. Microbiol. 58:2513-2516.

128. Gorski L, Jay-Russell MT, Liang AS, Walker S, Bengson Y, Govoni J, Mandrell RE. 2013. Diversity of pulsed-field gel electrophoresis pulsotypes, serovars, and antibiotic resistance among Salmonella isolates from wild amphibians and reptiles in the California central coast. Foodborne Pathog. Dis. 10:540-548.

129. Gottlieb SL, Newbern EC, Griffin PM, Graves LM, Hoekstra RM, Baker NL, Hunter SB, Holt KG, Ramsey F, Head M, Levine P, Johnson G, Schoonmaker-Bopp D, Reddy V, Kornstein L, Gerwel M, Nsubuga J, Edwards L, Stonecipher S, Hurd S, Austin D, Jefferson MA, Young SD, Hise K, Chernak ED, Sobel J. 2006. Multistate outbreak of listeriosis linked to turkey deli meat and subsequent changes in US regulatory policy. Clin. Infect. Dis. 42:29-36.

130. Gouin E, Adib-Conquy M, Balestrino D, Nahori MA, Villiers V, Colland F, Dramsi S, Dussurget O, Cossart P. 2010. The Listeria monocytogenes InlC protein interferes with innate immune responses by targeting the IκB kinase subunit IKKα. Proc. Nat. Acad. Sci. USA 107:17333-17338.

131. Gould LH, Demma L, Jones TF, Hurd S, Vugia DJ, Smith K, Shiferaw B, Segler S, Palmer A, Zansky S, Griffin PM. 2009. Hemolytic uremic syndrome and death in persons with Escherichia coli O157:H7 infection, Foodborne Diseases Active Surveillance Network sites, 2000–2006. Clin. Infect. Dis. 49:1480-1485.

132. Grad YH, Lipsitch M, Feldgarden M, Arachchi HM, Cerqueira GC, Fitzgerald M, Godfrey P, Haas BJ, Murphy CI, Russ C, Sykes S, Walker BJ, Wortman JR, Young S, Zeng Q, Abouelleil A, Bochicchio J, Chauvin S, Desmet T, Gujja S, McCowan C, Montmayeur A, Steelman S, Frimodt- Moller J, Petersen AM, Struve C, Krogfelt KA, Bingen E, Weill FX, Lander ES, Nusbaum C, Birren BW, Hung DT, Hanage WP. 2012. Genomic epidemiology of the Escherichia coli O104:H4 outbreaks in Europe, 2011. Proc. Nat. Acad. Sci. USA 109:3065-3070.

133. Gragg SE, Loneragan GH, Brashears MM, Arthur TM, Bosilevac JM, Kalchayanand N, Wang R, Schmidt JW, Brooks JC, Shackelford SD, Wheeler TL, Brown TR, Edrington TS, Brichta-Harhay DM. 2013. Cross- 55

Texas Tech University, Alex Brandt, May 2014

sectional study examining Salmonella enterica carriage in subiliac lymph nodes of cull and feedlot cattle at harvest. Foodborne Pathog. Dis. 10:368-374.

134. Graves LM, Hunter SB, Ong AR, Schoonmaker-Bopp D, Hise K, Kornstein L, DeWitt WE, Hayes PS, Dunne E, Mead P, Swaminathan B. 2005. Microbiological aspects of the investigation that traced the 1998 outbreak of listeriosis in the United States to contaminated hot dogs and establishment of molecular subtyping-based surveillance for Listeria monocytogenes in the PulseNet network. J. Clin. Microbiol. 43:2350-2355.

135. Groisman EA. 2001. The pleiotropic two-component regulatory system PhoP- PhoQ. J. Bacteriol. 183:1835-1842.

136. Gale Group. 2006. Bar-S Foods company: international directory of company histories. Available At: http://www.answers.com/topic/bar-s-foods-co. Accessed On: 2/12/14

137. Gründling A, Burrack LS, Bouwer HGA, Higgins DE, Mekalanos JJ. 2004. Listeria monocytogenes regulates flagellar motility gene expression through MogR, a transcriptional repressor required for virulence. Proc. Nat. Acad. Sci. USA 101:12318-12323.

138. Gründling A, Gonzalez MD, Higgins DE. 2003. Requirement of the Listeria monocytogenes broad-range phospholipase PC-PLC during infection of human epithelial cells. J. Bacteriol. 185:6295-6307.

139. Gudmundsdottir KB, Aalbaek B, Sigurdarson S, Gunnarsson E. 2004. The diversity of Listeria monocytogenes strains from 10 Icelandic sheep farms. J. Appl. Microbiol. 96:913-921.

140. Gudmundsdóttir S, Gudbjörnsdóttir B, Einarsson H, Kristinsson KG, Kristjánsson M. 2006. Contamination of cooked peeled shrimp (Pandalus borealis) by Listeria monocytogenes during processing at two processing plants. J. Food Prot. 69:1304-1311.

141. Guibourdenche M, Roggentin P, Mikoleit M, Fields PI, Bockemuhl J, Grimont PA, Weill FX. 2010. Supplement 2003-2007 (no. 47) to the White- Kauffmann-Le Minor scheme. Res. Microbiol. 161:26-29.

142. Gun H, Yilmaz A, Turker S, Tanlasi A, Yilmaz H. 2003. Contamination of bovine carcasses and abattoir environment by Escherichia coli O157:H7 in Istanbul. Int. J. Food Microbiol. 84:339-344.

Texas Tech University, Alex Brandt, May 2014

143. Hansen-Wester I, Hensel M. 2001. Salmonella pathogenicity islands encoding type III secretion systems. Microb. Infect. 3:549-559.

144. Hartland EL, Batchelor M, Delahay RM, Hale C, Matthews S, Dougan G, Knutton S, Connerton I, Frankel G. 1999. Binding of intimin from enteropathogenic Escherichia coli to Tir and to host cells. Mol. Microbiol. 32:151-158.

145. Harvey J, Gilmour A. 1994. Application of multilocus enzyme electrophoresis and restriction fragment length polymorphism analysis to the typing of Listeria monocytogenes strains isolated from raw milk, nondairy foods, and clinical and veterinary sources. Appl. Environ. Microbiol. 60:1547-1553.

146. Harvey J, Keenan KP, Gilmour A. 2007. Assessing biofilm formation by Listeria monocytogenes strains. Food Microbiol. 24:380-392.

147. Heir E, Lindstedt B-A, Røtterud O-J, Vardund T, Kapperud G, Nesbakken T. 2004. Molecular epidemiology and disinfectant susceptibility of Listeria monocytogenes from meat processing plants and human infections. Int. J. Food Microbiol. 96:85-96.

148. Helke DM, Wong ACL. 1994. Survival and growth characteristics of Listeria monocytogenes and Salmonella typhimurium on stainless steel and buna-n-rubber. J. Food Prot. 57:963-968.

149. Hengge-Aronis R, Klein W, Lange R, Rimmele M, Boos W. 1991. Trehalose synthesis genes are controlled by the putative sigma factor encoded by rpoS and are involved in stationary-phase thermotolerance in Escherichia coli. J. Bacteriol. 173:7918-7924.

150. Hensel M. 2000. Salmonella pathogenicity island 2. Mol. Microbiol. 36:1015- 1023.

151. Hilton AC, Penn CW. 1998. Comparison of ribotyping and arbitrarily primed PCR for molecular typing of Salmonella enterica and relationships between strains on the basis of these molecular markers. J. Appl. Microbiol. 85:933-940.

152. Hiramatsu R, Matsumoto M, Sakae K, Miyazaki Y. 2005. Ability of Shiga toxin-producing Escherichia coli and Salmonella spp. to survive in a desiccation model system and in dry foods. Appl. Environ. Microbiol. 71:6657-6663.

Texas Tech University, Alex Brandt, May 2014

153. Ho AJ, Lappi VR, Wiedmann M. 2007. Longitudinal monitoring of Listeria monocytogenes contamination patterns in a farmstead dairy processing facility. J. Dairy Sci. 90:2517-2524.

154. Hoelzer K, Sauders BD, Sanchez MD, Olsen PT, Pickett MM, Mangione KJ, Rice DH, Corby J, Stich S, Fortes ED, Roof SE, Grohn YT, Wiedmann M, Oliver HF. 2011. Prevalence, distribution, and diversity of Listeria monocytogenes in retail environments, focusing on small establishments and establishments with a history of failed inspections. J. Food Prot. 74:1083-1095.

155. Hoffman AD, Gall KL, Norton DM, Wiedmann M. 2003. Listeria monocytogenes contamination patterns for the smoked fish processing environment and for raw fish. J. Food Prot. 66:52-60.

156. Hohmann EL. 2001. Nontyphoidal salmonellosis. Clin. Infect. Dis. 32:263-269.

157. Holah JT, Bird J, Hall KE. 2004. The microbial ecology of high-risk, chilled food factories; evidence for persistent Listeria spp. and Escherichia coli strains. J. Appl. Microbiol. 97:68-77.

158. Holah JT, Taylor JH, Dawson DJ, Hall KE. 2002. Biocide use in the food industry and the disinfectant resistance of persistent strains of Listeria monocytogenes and Escherichia coli. J. Appl. Microbiol. 92S:111S-120S.

159. Holley RA, Proulx M. 1986. Use of egg washwater pH to prevent survival of Salmonella at moderate temperatures. Poult. Sci. 65:922-928.

160. Holt JG, Krieg NR, Sneath PHA, Staley JT, Williams ST. 2000. Bergey's manual of determinative bacteriology, 9th ed. Lippincott Williams & Wilkins, Philadelphia, PA.

161. Horwitz MA, Pollard RA, Merson MH, Martin SM. 1977. A large outbreak of foodborne salmonellosis on the Navajo Nation Indian Reservation, epidemiology and secondary transmission. Am. J. Pub. Health 67:1071-1076.

162. House HR, Ehlers JP. 2008. Travel-related infections. Emerg. Med. Clin. N. Am. 26:499-516.

163. Huang KH, Hsu BM, Chou MY, Tsai HL, Kao PM, Wang HJ, Hsiao HY, Su MJ, Huang YL. 2014. Application of molecular biological techniques to analyze Salmonella seasonal distribution in stream water. FEMS Microbiol. Lett. DOI:10.1111/1574-6968.12381.

Texas Tech University, Alex Brandt, May 2014

164. Hunter SB, Vauterin P, Lambert-Fair MA, Van Duyne MS, Kubota K, Graves L, Wrigley D, Barrett T, Ribot E. 2005. Establishment of a universal size standard strain for use with the PulseNet standardized pulsed-field gel electrophoresis protocols: converting the national databases to the new size standard. J. Clin. Microbiol. 43:1045-1050.

165. Jacquet C, Catimel B, Brosch R, Buchrieser C, Dehaumont P, Goulet V, Lepoutre A, Veit P, Rocourt J. 1995. Investigations related to the epidemic strain involved in the French listeriosis outbreak in 1992. Appl. Environ. Microbiol. 61:2242-2246.

166. Janning B, in't Veld PH, Notermans S, Kramer J. 1994. Resistance of bacterial strains to dry conditions: use of anhydrous silica gel in a desiccation model system. J. Appl. Bacteriol. 77:319-324.

167. Jarvis KG, Kaper JB. 1996. Secretion of extracellular proteins by enterohemorrhagic Escherichia coli via a putative type III secretion system. Infect. Immun. 64:4826-4829.

168. Jay JM, Loessner MJ, Golden DA. 2005. Foodborne gastroenteritis caused by Salmonella and Shigella, p. 619-636. In Jay JM, Loessner MJ, Golden DA (ed.), Modern Food Microbiology, 7 ed. Springer Science+Business Media, Inc., New York, NY.

169. Jeong KC, Hiki O, Kang MY, Park D, Kaspar CW. 2013. Prevalent and persistent Escherichia coli O157:H7 strains on farms are selected by bovine passage. Vet. Microbiol. 162:912-920.

170. Johansson T, Rantala L, Palmu L, Honkanen-Buzalski T. 1999. Occurrence and typing of Listeria monocytogenes strains in retail vacuum-packed fish products and in a production plant. Int. J. Food Microbiol. 47:111-119.

171. Jones BD. 1997. Host responses to pathogenic Salmonella infection. Genes Dev. 11:679-687.

172. Jones BD, Paterson HF, Hall A, Falkow S. 1993. Salmonella typhimurium induces membrane ruffling by a growth factor-receptor-independent mechanism. Proc. Nat. Acad. Sci. USA 90:10390-10394.

173. Jonquieres R, Bierne H, Mengaud J, Cossart P. 1998. The inlA gene of Listeria monocytogenes LO28 harbors a nonsense mutation resulting in release of internalin. Infect. Immun. 66:3420-3422.

Texas Tech University, Alex Brandt, May 2014

174. Junttila JR, Niemela SI, Hirn J. 1988. Minimum growth temperatures of Listeria monocytogenes and non-haemolytic Listeria. J. Appl. Bacteriol. 65:321- 327.

175. Juven BJ, Cox NA, Bailey JS, Thomson JE, Charles OW, Shutze JV. 1984. Survival of Salmonella in dry food and feed. J. Food Prot. 47:445–448.

176. Kabuki DY, Kuaye AY, Wiedmann M, Boor KJ. 2004. Molecular subtyping and tracking of Listeria monocytogenes in latin-style fresh-cheese processing plants. J. Dairy Sci. 87:2803-2812.

177. Kalmokoff ML, Austin JW, Wan XD, Sanders G, Banerjee S, Farber JM. 2001. Adsorption, attachment and biofilm formation among isolates of Listeria monocytogenes using model conditions. J. Appl. Bacteriol. 91:725-734.

178. Kapperud G, Gustavsen S, Hellesnes I, Hansen AH, Lassen J, Hirn J, Jahkola M, Montenegro MA, Helmuth R. 1990. Outbreak of Salmonella typhimurium infection traced to contaminated chocolate and caused by a strain lacking the 60-megadalton virulence plasmid. J. Clin. Microbiol. 28:2597-2601.

179. Karmali MA, Mascarenhas M, Shen S, Ziebell K, Johnson S, Reid-Smith R, Isaac-Renton J, Clark C, Rahn K, Kaper JB. 2003. Association of genomic O island 122 of Escherichia coli EDL 933 with verocytotoxin-producing Escherichia coli seropathotypes that are linked to epidemic and/or serious disease. J. Clin. Microbiol. 41:4930-4940.

180. Karmali MA, Steele BT, Petric M, Lim C. 1983. Sporadic cases of haemolytic- uraemic syndrome associated with faecal cytotoxin and cytotoxin-producing Escherichia coli in stools. Lancet 1:619-620.

181. Kastbjerg VG, Gram L. 2009. Model systems allowing quantification of sensitivity to disinfectants and comparison of disinfectant susceptibility of persistent and presumed nonpersistent Listeria monocytogenes. J. Appl. Microbiol. 106:1667-1681.

182. Kathariou S. 2002. Listeria monocytogenes virulence and pathogenicity, a food safety perspective. J. Food Prot. 65:1811-1829.

183. Kaur J, Jain SK. 2012. Role of antigens and virulence factors of Salmonella enterica serovar Typhi in its pathogenesis. Microbiol. Res. 167:199-210.

Texas Tech University, Alex Brandt, May 2014

184. Keene WE, Sazie E, Kok J, Rice DH, Hancock DD, Balan VK, Zhao T, Doyle MP. 1997. An outbreak of Escherichia coli O157:H7 infections traced to jerky made from deer meat. J. Am. Med. Assoc. 277:1229-1231.

185. Kenny B, DeVinney R, Stein M, Reinscheid DJ, Frey EA, Finlay BB. 1997. Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells. Cell 91:511-520.

186. Keto-Timonen R, Tolvanen R, Lunden J, Korkeala H. 2007. An 8-year surveillance of the diversity and persistence of Listeria monocytogenes in a chilled food processing plant analyzed by amplified fragment length polymorphism. J. Food Prot. 70:1866-1873.

187. Kilonzo C, Atwill ER, Mandrell R, Garrick M, Villanueva V, Hoar BR. 2011. Prevalence and molecular characterization of Escherichia coli O157:H7 by multiple-locus variable-number tandem repeat analysis and pulsed-field gel electrophoresis in three sheep farming operations in California. J. Food Prot. 74:1413-1421.

188. Kitov PI, Sadowska JM, Mulvey G, Armstrong GD, Ling H, Pannu NS, Read RJ, Bundle DR. 2000. Shiga-like toxins are neutralized by tailored multivalent carbohydrate ligands. Nature 403:669-672.

189. Klæboe H, Lunestad BT, Borlaug K, Paulauskas A, Rosef O. 2010. Persistence and diversity of Listeria monocytogenes isolates in Norwegian processing plants. Vet. Med. Zoot. 50:42-47.

190. Klæboe H, Rosef O, Fortes E, Wiedmann M. 2006. Ribotype diversity of Listeria monocytogenes isolates from two salmon processing plants in Norway. Int. J. Environ. Health Res. 16:375-383.

191. Knabel SJ, Reimer A, Verghese B, Lok M, Ziegler J, Farber J, Pagotto F, Graham M, Nadon CA, Gilmour MW. 2012. Sequence typing confirms that a predominant Listeria monocytogenes clone caused human listeriosis cases and outbreaks in Canada from 1988 to 2010. J. Clin. Microbiol. 50:1748-1751.

192. Knodler LA, Steele-Mortimer O. 2003. Taking possession: biogenesis of the Salmonella-containing vacuole. Traffic 4:587-599.

193. Knutton S, Rosenshine I, Pallen MJ, Nisan I, Neves BC, Bain C, Wolff C, Dougan G, Frankel G. 1998. A novel EspA-associated surface organelle of enteropathogenic Escherichia coli involved in protein translocation into epithelial cells. EMBO J. 17:2166-2176.

Texas Tech University, Alex Brandt, May 2014

194. Kresse AU, Beltrametti F, Muller A, Ebel F, Guzman CA. 2000. Characterization of SepL of enterohemorrhagic Escherichia coli. J. Bacteriol. 182:6490-6498.

195. Krysinski EP, Brown LJ, Marchisello TJ. 1992. Effect of cleaners and sanitizers on Listeria monocytogenes attached to product contact surfaces. J. Food Prot. 55:246-251.

196. Kuhle V, Abrahams GL, Hensel M. 2006. Intracellular Salmonella enterica redirect exocytic transport processes in a Salmonella pathogenicity island 2- dependent manner. Traffic 7:716-730.

197. Lahti E, Ruoho O, Rantala L, Hanninen ML, Honkanen-Buzalski T. 2003. Longitudinal study of Escherichia coli O157 in a cattle finishing unit. Appl. Environ. Microbiol. 69:554-561.

198. Lappi VR, Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Longitudinal studies on Listeria in smoked fish plants: impact of intervention strategies on contamination patterns. J. Food Prot. 67:2500-2514.

199. Latorre AA, Van Kessel JA, Karns JS, Zurakowski MJ, Pradhan AK, Boor KJ, Adolph E, Sukhnanand S, Schukken YH. 2011. Increased in vitro adherence and on-farm persistence of predominant and persistent Listeria monocytogenes strains in the milking system. Appl. Environ. Microbiol. 77:3676- 3684.

200. Latorre AA, Van Kessel JAS, Karns JS, Zurakowski MJ, Pradhan AK, Zadoks RN, Boor KJ, Schukken YH. 2009. Molecular ecology of Listeria monocytogenes: evidence for a reservoir in milking equipment on a dairy farm. Appl. Environ. Microbiol. 75:1315-1323.

201. Lawrence LM, Gilmour A. 1995. Characterization of Listeria monocytogenes isolated from poultry products and from the poultry-processing environment by random amplification of polymorphic DNA and multilocus enzyme electrophoresis. Appl. Environ. Microbiol. 61:2139-2144.

202. Le VT, Diep BA. 2013. Selected insights from application of whole-genome sequencing for outbreak investigations. Curr. Opin. Crit. Care 19:432-439.

203. Lear JC, Maillard JY, Dettmar PW, Goddard PA, Russell AD. 2002. Chloroxylenol- and triclosan-tolerant bacteria from industrial sources. J. Ind. Microbiol. Biotechnol. 29:238-242.

Texas Tech University, Alex Brandt, May 2014

204. Lecos C. 1986. Of microbes and milk: probing America's worst Salmonella outbreak. Dairy Food Sanit. 6:136-140.

205. Lecuit M, Hurme R, Pizarro-Cerda J, Ohayon H, Geiger B, Cossart P. 2000. A role for alpha-and beta-catenins in bacterial uptake. Proc. Nat. Acad. Sci. USA 97:10008-10013.

206. Lecuit M, Ohayon H, Braun L, Mengaud J, Cossart P. 1997. Internalin of Listeria monocytogenes with an intact leucine-rich repeat region is sufficient to promote internalization. Infect. Immun. 65:5309-5319.

207. Lee IS, Lin J, Hall HK, Bearson B, Foster JW. 1995. The stationary-phase sigma factor sigma S (RpoS) is required for a sustained acid tolerance response in virulent Salmonella typhimurium. Mol. Microbiol. 17:155-167.

208. LeJeune JT, Besser TE, Rice DH, Berg JL, Stilborn RP, Hancock DD. 2004. Longitudinal study of fecal shedding of Escherichia coli O157:H7 in feedlot cattle: predominance and persistence of specific clonal types despite massive cattle population turnover. Appl. Environ. Microbiol. 70:377-384.

209. Lewontin RC, Hubby JL. 1966. A molecular approach to the study of genic heterozygosity in natural populations II: amount of variation and degree of heterozygosity in natural populations of Drosophila pseudoobscura. Genetics 54:595-609.

210. Li X, Payne JB, Santos FB, Levine JF, Anderson KE, Sheldon BW. 2007. Salmonella populations and prevalence in layer feces from commercial high-rise houses and characterization of the Salmonella isolates by serotyping, antibiotic resistance analysis, and pulsed field gel electrophoresis. Poult. Sci. 86:591-597.

211. Liebana E, Garcia-Migura L, Breslin MF, Davies RH, Woodward MJ. 2001. Diversity of strains of Salmonella enterica serotype Enteritidis from English poultry farms assessed by multiple genetic fingerprinting. J. Clin. Microbiol. 39:154-161.

212. Lim JY, Yoon J, Hovde CJ. 2010. A brief overview of Escherichia coli O157:H7 and its plasmid O157. J. Microbiol. Biotechnol. 20:5-14.

213. Lin J, Smith MP, Chapin KC, Baik HS, Bennett GN, Foster JW. 1996. Mechanisms of acid resistance in enterohemorrhagic Escherichia coli. Appl. Environ. Microbiol. 62:3094-3100.

Texas Tech University, Alex Brandt, May 2014

214. Ling H, Boodhoo A, Hazes B, Cummings MD, Armstrong GD, Brunton JL, Read RJ. 1998. Structure of the Shiga-like toxin I B-pentamer complexed with an analogue of its receptor Gb3. Biochem. 37:1777-1788.

215. Lingnau A, Domann E, Hudel M, Bock M, Nichterlein T, Wehland J, Chakraborty T. 1995. Expression of the Listeria monocytogenes EGD inlA and inlB genes, whose products mediate bacterial entry into tissue culture cell lines, by PrfA-dependent and -independent mechanisms. Infect. Immun. 63:3896-3903.

216. Linnan MJ, Mascola L, Lou XD, Goulet V, May S, Salminen C, Hird DW, Yonekura ML, Hayes P, Weaver R, Audurier A, Plikaytis BF, Fannin SL, Kleks A, Broome CV. 1988. Epidemic listeriosis associated with Mexican-style cheese. N. Engl. J. Med. 319:823-828.

217. Liu DY, Lawrence ML, Gorski L, Mandrell RE, Ainsworth AJ, Austin FW. 2006. Listeria monocytogenes serotype 4b strains belonging to lineages I and III possess distinct molecular features. J. Clin. Microbiol. 44:214-217.

218. Lomonaco S, Verghese B, Gerner-Smidt P, Tarr C, Gladney L, Joseph L, Katz L, Turnsek M, Frace M, Chen Y, Brown E, Meinersmann R, Berrang M, Knabel S. 2013. Novel epidemic clones of Listeria monocytogenes, United States, 2011. Emerg. Infect. Dis. 19:147-150.

219. López V, Ortiz S, Corujo A, López P, Poza D, Navas J, Moreno R, Martínez- Suárez JV. 2008. Different contamination patterns of lineage I and II strains of Listeria monocytogenes in a Spanish broiler abattoir. Poult. Sci. 87:1874-1882.

220. Lukinmaa S, Miettinen M, Nakari UM, Korkeala H, Siitonen A. 2003. Listeria monocytogenes isolates from invasive infections: variation of sero- and genotypes during an 11-year period in Finland. J. Clin. Microbiol. 41:1694-1700.

221. Lundén J, Autio T, Markkula A, Hellström S, Korkeala H. 2003. Adaptive and cross-adaptive responses of persistent and non-persistent Listeria monocytogenes strains to disinfectants. Int. J. Food Microbiol. 82:265-272.

222. Lundén J, Tolvanen R, Korkeala H. 2008. Acid and heat tolerance of persistent and nonpersistent Listeria monocytogenes food plant strains. Lett. Appl. Microbiol. 46:276-280.

223. Lundén JM, Autio TJ, Korkeala HJ. 2002. Transfer of persistent Listeria monocytogenes contamination between food-processing plants associated with a dicing machine. J. Food Prot. 65:1129-1133.

Texas Tech University, Alex Brandt, May 2014

224. Lundén JM, Autio TJ, Sjöberg AM, Korkeala HJ. 2003. Persistent and nonpersistent Listeria monocytogenes contamination in meat and poultry processing plants. J. Food Prot. 66:2062-2069.

225. Lundén JM, Miettinen MK, Autio TJ, Korkeala HJ. 2000. Persistent Listeria monocytogenes strains show enhanced adherence to food contact surface after short contact times. J. Food Prot. 63:1204-1207.

226. MacDonald DM, Fyfe M, Paccagnella A, Trinidad A, Louie K, Patrick D. 1999. Escherichia coli O157:H7 outbreak linked to salami, British Columbia, Canada, 1999. Epidemiol. Infect. 132:283-289.

227. MacDonald PD, Whitwam RE, Boggs JD, MacCormack JN, Anderson KL, Reardon JW, Saah JR, Graves LM, Hunter SB, Sobel J. 2005. Outbreak of listeriosis among Mexican immigrants as a result of consumption of illicitly produced Mexican-style cheese. Clin. Infect. Dis. 40:677-682.

228. Mafu AA, Roy D, Goulet J, Magny P. 1990. Attachment of Listeria monocytogenes to stainless steel, glass, polypropylene, and rubber surfaces after short contact times. J. Food Prot. 53:742-746.

229. Malinverni R, Bille J, Perret C, Regli F, Tanner F, Glauser MP. 1985. Epidemic listeriosis: report of 25 cases in 15 months at the Vaud University Hospital Center. Schweiz. Med. Wochenschr. 115:2-10.

230. Malley TJ, Stasiewicz MJ, Grohn YT, Roof S, Warchocki S, Nightingale K, Wiedmann M. 2013. Implementation of statistical tools to support identification and management of persistent Listeria monocytogenes contamination in smoked fish processing plants. J. Food Prot. 76:796-811.

231. Mallo GV, Espina M, Smith AC, Terebiznik MR, Aleman A, Finlay BB, Rameh LE, Grinstein S, Brumell JH. 2008. SopB promotes phosphatidylinositol 3-phosphate formation on Salmonella vacuoles by recruiting Rab5 and Vps34. J. Cell Biol. 182:741-752.

232. Marco AJ, Prats N, Ramos JA, Briones V, Blanco M, Dominguez L, Domingo M. 1992. A microbiological, histopathological and immunohistological study of the intragastric inoculation of Listeria monocytogenes in mice. J. Compar. Path. 107:1-9.

233. Marler B. 2013. Publisher’s platform: why isn’t Salmonella a legal adulterant? Available At: http://www.foodsafetynews.com/2013/02/publishers-platform-why- isnt-salmonella-a-legal-adulterant/#.UumpfbRdyYc. Accessed On: 1/29/14

Texas Tech University, Alex Brandt, May 2014

234. Martin B, Garriga M, Aymerich T. 2011. Prevalence of Salmonella spp. and Listeria monocytogenes at small-scale Spanish factories producing traditional fermented sausages. J. Food Prot. 74:812-815.

235. Martinez-Urtaza J, Liebana E. 2005. Investigation of clonal distribution and persistence of Salmonella Senftenberg in the marine environment and identification of potential sources of contamination. FEMS Microbiol. Ecol. 52:255-263.

236. Mathew AG, Cissell R, Liamthong S. 2007. Antibiotic resistance in bacteria associated with food animals: a United States perspective of livestock production. Foodborne Pathog. Dis. 4:115-133.

237. McCollum JT, Cronquist AB, Silk BJ, Jackson KA, O'Connor KA, Cosgrove S, Gossack JP, Parachini SS, Jain NS, Ettestad P, Ibraheem M, Cantu V, Joshi M, DuVernoy T, Fogg NW, Gorny JR, Mogen KM, Spires C, Teitell P, Joseph LA, Tarr CL, Imanishi M, Neil KP, Tauxe RV, Mahon BE. 2013. Multistate outbreak of listeriosis associated with cantaloupe. N. Eng. J. Med. 369:944-953.

238. McDaniel TK, Jarvis KG, Donnenberg MS, Kaper JB. 1995. A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens. Proc. Nat. Acad. Sci. USA 92:1664-1668.

239. McGhie EJ, Hayward RD, Koronakis V. 2001. Cooperation between actin- binding proteins of invasive Salmonella: SipA potentiates SipC nucleation and bundling of actin. EMBO J. 20:2131-2139.

240. McLauchlin J. 1990. Distribution of serovars of Listeria monocytogenes isolated from different categories of patients with listeriosis. Eur. J. Clin. Microbiol. Infect. Dis. 9:210-213.

241. Mead PS, Dunne EF, Graves L, Wiedmann M, Patrick M, Hunter S, Salehi E, Mostashari F, Craig A, Mshar P, Bannerman T, Sauders BD, Hayes P, Dewitt W, Sparling P, Griffin P, Morse D, Slutsker L, Swaminathan B. 2006. Nationwide outbreak of listeriosis due to contaminated meat. Epidemiol. Infect. 134:744-751.

242. Medrala D. 2003. Persistence of Listeria monocytogenes strains isolated from products in a Polish fish-processing plant over a 1-year period. Food Microbiol. 20:715-724.

Texas Tech University, Alex Brandt, May 2014

243. Meloni D, Piras F, Mureddu A, Fois F, Consolati SG, Lamon S, Mazzette R. 2013. Listeria monocytogenes in five Sardinian swine slaughterhouses: prevalence, serotype, and genotype characterization. J. Food Prot. 76:1863-1867.

244. Mengaud J, Braun-Breton C, Cossart P. 1991. Identification of phosphatidylinositol-specific phospholipase C activity in Listeria monocytogenes: a novel type of virulence factor. Mol. Microbiol. 5:367-372.

245. Mengaud J, Ohayon H, Gounon P, Mege RM, Cossart P. 1996. E-cadherin is the receptor for internalin, a surface protein required for entry of L. monocytogenes into epithelial cells. Cell 84:923-932.

246. Michino H, Araki K, Minami S, Takaya S, Sakai N, Miyazaki M, Ono A, Yanagawa H. 1999. Massive outbreak of Escherichia coli O157:H7 infection in schoolchildren in Sakai City, Japan, associated with consumption of white radish sprouts. Am. J. Epidemiol. 150:787-796.

247. Miettinen M. 1999. Characterization of Listeria monocytogenes from an ice cream plant by serotyping and pulsed-field gel electrophoresis. Int. J. Food Microbiol. 46:187-192.

248. Milohanic E, Jonquières R, Cossart P, Berche P, Gaillard J-L. 2001. The autolysin Ami contributes to the adhesion of Listeria monocytogenes to eukaryotic cells via its cell wall anchor. Mol. Microbiol. 39:1212-1224.

249. Monack DM, Bouley DM, Falkow S. 2004. Salmonella typhimurium persists within macrophages in the mesenteric lymph nodes of chronically infected Nramp1+/+ mice and can be reactivated by IFNγneutralization. J. Exp. Med. 199:231-241.

250. Monk IR, Casey PG, Hill C, Gahan CG. 2010. Directed evolution and targeted mutagenesis to murinize Listeria monocytogenes internalin A for enhanced infectivity in the murine oral infection model. BMC Microbiol. 10:318.

251. Moon HW, Whipp SC, Argenzio RA, Levine MM, Giannella RA. 1983. Attaching and effacing activities of rabbit and human enteropathogenic Escherichia coli in pig and rabbit intestines. Infect. Immun. 41:1340-1351.

252. Morgan F, Bonnin V, Mallereau MP, Perrin G. 2001. Survival of Listeria monocytogenes during manufacture, ripening and storage of soft lactic cheese made from raw goat milk. Int. J. Food Microbiol. 64:217-221.

Texas Tech University, Alex Brandt, May 2014

253. Morris IJ, Ribeiro CD. 1989. Listeria monocytogenes and pâté. Lancet 334:1285-1286.

254. Mounier J, Ryter A, Coquis-Rondon M, Sansonetti PJ. 1990. Intracellular and cell-to-cell spread of Listeria monocytogenes involves interaction with F-actin in the enterocytelike cell line Caco-2. Infect. Immun. 58:1048-1058.

255. Murray EG. 1955. A characterization of listeriosis in man and other animals. Can. Med. Assoc. J. 72:99-103.

256. Murray EGD, Webb RA, Swann MBR. 1926. A disease of rabbits characterised by a large mononuclear leucocytosis, caused by a hitherto undescribed bacillus Bacterium monocytogenes (n.sp.). J. Pathol. Bacteriol. 29:407-439.

257. Mylonakis E, Hohmann EL, Calderwood SB. 1998. Central nervous system infection with Listeria monocytogenes: 33 years' experience at a general hospital and review of 776 episodes from the literature. Medicine 77:313-336.

258. Mylonakis E, Paliou M, Hohmann EL, Calderwood SB, Wing EJ. 2002. Listeriosis during pregnancy: a case series and review of 222 cases. Medicine 81:260-269.

259. Nataro JP, Kaper JB. 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11:142-201.

260. Navarro-Gonzalez N, Mentaberre G, Porrero CM, Serrano E, Mateos A, Lopez-Martin JM, Lavin S, Dominguez L. 2012. Effect of cattle on Salmonella carriage, diversity and antimicrobial resistance in free-ranging wild boar (Sus scrofa) in northeastern Spain. PLoS One 7:e51614.

261. Nesbakken T, Kapperud G, Caugant DA. 1996. Pathways of Listeria monocytogenes contamination in the meat processing industry. Int. J. Food Microbiol. 31:161-171.

262. Nesse LL, Nordby K, Heir E, Bergsjoe B, Vardund T, Nygaard H, Holstad G. 2003. Molecular analyses of Salmonella enterica isolates from fish feed factories and fish feed ingredients. Appl. Environ. Microbiol. 69:1075-1081.

263. Nieman RE, Lorber B. 1980. Listeriosis in adults, a changing pattern: report of eight cases and review of the literature, 1968-1978. Rev. Infect. Dis. 2:207-227.

264. Nightingale K, Bovell L, Grajczyk A, Wiedmann M. 2007. Combined sigB allelic typing and multiplex PCR provide improved discriminatory power and 68

Texas Tech University, Alex Brandt, May 2014

reliability for Listeria monocytogenes molecular serotyping. J. Microbiol. Meth. 68:52-59.

265. Nightingale KK, Schukken YH, Nightingale CR, Fortes ED, Ho AJ, Her Z, Grohn YT, McDonough PL, Wiedmann M. 2004. Ecology and transmission of Listeria monocytogenes infecting ruminants and in the farm environment. Appl. Environ. Microbiol. 70:4458-4467.

266. Nightingale KK, Windham K, Martin KE, Yeung M, Wiedmann M. 2005. Select Listeria monocytogenes subtypes commonly found in foods carry distinct nonsense mutations in inlA, leading to expression of truncated and secreted internalin A, and are associated with a reduced invasion phenotype for human intestinal epithelial cells. Appl. Environ. Microbiol. 71:8764-8772.

267. Nocera D, Bannerman E, Rocourt J, Jaton-Ogay K, Bille J. 1990. Characterization by DNA restriction endonuclease analysis of Listeria monocytogenes strains related to the Swiss epidemic of listeriosis. J. Clin. Microbiol. 28:2259-2263.

268. Norton DM, McCamey MA, Gall KL, Scarlett JM, Boor KJ, Wiedmann M. 2001. Molecular studies on the ecology of Listeria monocytogenes in the smoked fish processing industry. Appl. Environ. Microbiol. 67:198-205.

269. O'Brien AD, LaVeck GD, Thompson MR, Formal SB. 1982. Production of Shigella dysenteriae type 1-like cytotoxin by Escherichia coli. J. Infect. Dis. 146:763-769.

270. Ojeniyi B, Christensen J, Bisgaard M. 2000. Comparative investigations of Listeria monocytogenes isolated from a turkey processing plant, turkey products, and from human cases of listeriosis in Denmark. Epidemiol. Infect. 125:303-308.

271. Olier M, Pierre F, Lemaitre JP, Divies C, Rousset A, Guzzo J. 2002. Assessment of the pathogenic potential of two Listeria monocytogenes human faecal carriage isolates. Microbiol. 148:1855-1862.

272. Olive DM, Bean P. 1999. Principles and applications of methods for DNA-based typing of microbial organisms. J. Clin. Microbiol. 37:1661-1669.

273. Oliver HF, Orsi RH, Wiedmann M, Boor KJ. 2010. Listeria monocytogenes σB has a small core regulon and a conserved role in virulence but makes differential contributions to stress tolerance across a diverse collection of strains. Appl. Environ. Microbiol. 76:4216-4232.

Texas Tech University, Alex Brandt, May 2014

274. Olsen SJ, Patrick M, Hunter SB, Reddy V, Kornstein L, MacKenzie WR, Lane K, Bidol S, Stoltman GA, Frye DM, Lee I, Hurd S, Jones TF, LaPorte TN, Dewitt W, Graves L, Wiedmann M, Schoonmaker-Bopp DJ, Huang AJ, Vincent C, Bugenhagen A, Corby J, Carloni ER, Holcomb ME, Woron RF, Zansky SM, Dowdle G, Smith F, Ahrabi-Fard S, Ong AR, Tucker N, Hynes NA, Mead P. 2005. Multistate outbreak of Listeria monocytogenes infection linked to delicatessen turkey meat. Clin. Infect. Dis. 40:962-967.

275. Orsi RH, Borowsky ML, Lauer P, Young SK, Nusbaum C, Galagan JE, Birren BW, Ivy RA, Sun Q, Graves LM, Swaminathan B, Wiedmann M. 2008. Short-term genome evolution of Listeria monocytogenes in a non-controlled environment. BMC Genom. 9:539.

276. Orsi RH, denBakker HC, Wiedmann M. 2010. Listeria monocytogenes lineages: genomics, evolution, ecology, and phenotypic characteristics. Int. J. Med. Microbiol. DOI:10.1016/j.ijmm.2010.1005.1002.

277. Orskov F, Orskov I. 1992. Escherichia coli serotyping and disease in man and animals. Can. J. Microbiol. 38:699-704.

278. Ortiz S, Lopez V, Villatoro D, Lopez P, Davila JC, Martinez-Suarez JV. 2010. A 3-year surveillance of the genetic diversity and persistence of Listeria monocytogenes in an Iberian pig slaughterhouse and processing plant. Foodborne Pathog. Dis. 7:1177-1184.

279. Pagotto F, Ng LK, Clark C, Farber J. 2006. Canadian listeriosis reference service. Foodborne Pathog. Dis. 3:132-137.

280. Pandiripally VK, Westbrook DG, Sunki GR, Bhunia AK. 1999. Surface protein p104 is involved in adhesion of Listeria monocytogenes to human intestinal cell line, Caco-2. J. Med. Microbiol. 48:117-124.

281. Parish ME, Higgins DP. 1989. Survival of Listeria monocytogenes in low pH model broth systems. J. Food Prot. 52:144-147.

282. Patchanee P, Molla B, White N, Line DE, Gebreyes WA. 2010. Tracking Salmonella contamination in various watersheds and phenotypic and genotypic diversity. Foodborne Pathog. Dis. 7:1113-1120.

283. Peccio A, Autio T, Korkeala H, Rosmini R, Trevisani M. 2003. Listeria monocytogenes occurrence and characterization in meat-producing plants. Lett. Appl. Microbiol. 37:234-238.

Texas Tech University, Alex Brandt, May 2014

284. Peel M, Donachie W, Shaw A. 1988. Temperature-dependent expression of flagella of Listeria monocytogenes studied by electron microscopy, SDS-PAGE and western blotting. J. Gen. Microbiol. 134:2171-2178.

285. Perrin M, Bemer M, Delamare C. 2003. Fatal case of Listeria innocua bacteremia. J. Clin. Microbiol. 41:5308-5309.

286. Public Health Agency of Canada. 2008. Lessons learned: public health agency of Canada's response to the 2008 listeriosis outbreak. Available At: http://www .phac-aspc.gc.ca/fs-sa/listeria/2008-lessons-lecons-eng.php. Accessed On: 2/12/14

287. Pinner RW, Schuchat A, Swaminathan BB, Hayes PS, Deaver KA, Weaver RE, Plikaytis BD, Reeves M, Broome CV, Wenger JD. 1992. Role of foods in sporadic listeriosis: II: microbilogic and epidemiologic investigation. J. Am. Med. Assoc. 267:2046-2050.

288. Podolak R, Enache E, Stone W, Black DG, Elliott PH. 2010. Sources and risk factors for contamination, survival, persistence, and heat resistance of Salmonella in low-moisture foods. J. Food Prot. 73:1919-1936.

289. Porsby CH, Vogel BF, Mohr M, Gram L. 2008. Influence of processing steps in cold-smoked salmon production on survival and growth of persistent and presumed non-persistent Listeria monocytogenes. Int. J. Food Microbiol. 122:287-295.

290. Poyart C, Abachin E, Razafimanantsoa I, Berche P. 1993. The zinc metalloprotease of Listeria monocytogenes is required for maturation of phosphatidylcholine phospholipase C: direct evidence obtained by gene complementation. Infect. Immun. 61:1576-1580.

291. Prendergast DM, Duggan SJ, Gonzales-Barron U, Fanning S, Butler F, Cormican M, Duffy G. 2009. Prevalence, numbers and characteristics of Salmonella spp. on Irish retail pork. Int. J. Food Microbiol. 131:233-239.

292. Rabinovich L, Sigal N, Borovok I, Nir-Paz R, Herskovits Anat A. 2012. Prophage excision activates Listeria competence genes that promote phagosomal escape and virulence. Cell 150:792-802.

293. Rahn K, Renwick SA, Johnson RP, Wilson JB, Clarke RC, Alves D, McEwen S, Lior H, Spika J. 1997. Persistence of Escherichia coli O157:H7 in dairy cattle and the dairy farm environment. Epidemiol. Infect. 119:251-259.

Texas Tech University, Alex Brandt, May 2014

294. Raimann E, Schmid B, Stephan R, Tasara T. 2009. The alternative sigma factor σL of L. monocytogenes promotes growth under diverse environmental stresses. Foodborne Pathog. Dis. 6:583-591.

295. Rapose A, Lick SD, Ismail N. 2008. Listeria grayi bacteremia in a heart transplant recipient. Transpl. Infect. Dis. 10:434-436.

296. Rea RB, Gahan CG, Hill C. 2004. Disruption of putative regulatory loci in Listeria monocytogenes demonstrates a significant role for Fur and PerR in virulence. Infect. Immun. 72:717-727.

297. Reitler R, Yarom D, Seligmann R. 1960. The enhancing effect of staphylococcal enterotoxin on Salmonella infection. Med. Off. 104:181.

298. Riley LW, Remis RS, Helgerson SD, McGee HB, Wells JG, Davis BR, Hebert RJ, Olcott ES, Johnson LM, Hargrett NT, Blake PA, Cohen ML. 1983. Hemorrhagic colitis associated with a rare Escherichia coli serotype. N. Engl. J. Med. 308:681-685.

299. Ringus DL, Ivy RA, Wiedmann M, Boor KJ. 2012. Salt stress-induced transcription of σB- and CtsR-regulated genes in persistent and non-persistent Listeria monocytogenes strains from food processing plants. Foodborne Pathog. Dis. 9:198-206.

300. Ritter AC, Bacciu D, Santi L, Silva WO, Vainstein MH, Rubino S, Uzzau S, Tondo EC. 2012. Investigation of rpoS and dps genes in sodium hypochlorite resistance of Salmonella Enteritidis SE86 isolated from foodborne illness outbreaks in southern Brazil. J. Food Prot. 75:437-442.

301. Robinson CM, Sinclair JF, Smith MJ, O'Brien AD. 2006. Shiga toxin of enterohemorrhagic Escherichia coli type O157:H7 promotes intestinal colonization. Proc. Nat. Acad. Sci. USA 103:9667-9672.

302. Rocourt J, Hof H, Schrettenbrunner A, Malinverni R, Bille J. 1986. Acute purulent Listeria seelingeri meningitis in an immunocompetent adult. Schweiz. Med. Wochenschr. 116:248-251.

303. Rørvik LM, Aase B, Alvestad T, Caugant DA. 2000. Molecular epidemiological survey of Listeria monocytogenes in seafoods and seafood- processing plants. Appl. Environ. Microbiol. 66:4779-4784.

Texas Tech University, Alex Brandt, May 2014

304. Rørvik LM, Aase B, Alvestad T, Caugant DA. 2003. Molecular epidemiological survey of Listeria monocytogenes in broilers and poultry products. J. Appl. Microbiol. 94:633-640.

305. Rørvik LM, Caugant DA, Yndestad M. 1995. Contamination pattern of Listeria monocytogenes and other Listeria spp. in a salmon slaughterhouse and smoked salmon processing plant. Int. J. Food Microbiol. 25:19-27.

306. Ruschkowski S, Rosenshine I, Finlay BB. 1992. Salmonella typhimurium induces an inositol phosphate flux in infected epithelial cells. FEMS Microbiol. Lett. 74:121-126.

307. Sanderson MW, Sargeant JM, Shi X, Nagaraja TG, Zurek L, Alam MJ. 2006. Longitudinal emergence and distribution of Escherichia coli O157 genotypes in a beef feedlot. Appl. Environ. Microbiol. 72:7614-7619.

308. Sandvang D, Jensen LB, Baggesen DL, Baloda SB. 2000. Persistence of a Salmonella enterica serotype Typhimurium clone in Danish pig production units and farmhouse environment studied by pulsed field gel electrophoresis (PFGE). FEMS Microbiol. Lett. 187:21-25.

309. Santos RL, Tsolis RM, Baumler AJ, Adams LG. 2003. Pathogenesis of Salmonella-induced enteritis. Braz. J. Med. Biol. Res. 36:3-12.

310. Sasahara KC, Zottola EA. 1993. Biofilm formation by Listeria monocytogenes utilizes a primary colonizing microorganism in flowing systems. J. Food Prot. 56:1022-1028.

311. Sauders BD. 2013. CDC/FDA partnership targets whole genome sequencing of Listeria monocytogenes. Available At: http://www.foodsafetynews.com/2013/11/ cdc-begins-real-time-whole-genome-sequencing-of-listeria-monocytogenes/#. Uv0LfbRfSv8. Accessed On: 2/13/14

312. Sauders BD, Durak MZ, Fortes E, Windham K, Schukken Y, Lembo AJ, Akey B, Nightingale KK, Wiedmann M. 2006. Molecular characterization of Listeria monocytogenes from natural and urban environments. J. Food Prot. 69:93-105.

313. Sauders BD, Mangione K, Vincent C, Schermerhorn J, Farchione CM, Dumas NB, Bopp D, Kornstein L, Fortes ED, Windham K, Wiedmann M. 2004. Distribution of Listeria monocytogenes molecular subtypes among human and food isolates from New York state shows persistence of human disease-

Texas Tech University, Alex Brandt, May 2014

associated Listeria monocytogenes strains in retail environments. J. Food Prot. 67:1417-1428.

314. Sauders BD, Overdevest J, Fortes E, Windham K, Schukken Y, Lembo A, Wiedmann M. 2012. Diversity of Listeria species in urban and natural environments. Appl. Environ. Microbiol. 78:4420-4433.

315. Sauders BD, Wiedmann M. 2007. Ecology of Listeria species and L. monocytogenes in the natural environment, p. 21-53. In Ryser ET, Marth EH (ed.), Listeria, listeriosis, and food safety, 3rd ed. CRC Press, Boca Raton, FL.

316. Savelkoul PH, Aarts HJ, de Haas J, Dijkshoorn L, Duim B, Otsen M, Rademaker JL, Schouls L, Lenstra JA. 1999. Amplified-fragment length polymorphism analysis: the state of an art. J. Clin. Microbiol. 37:3083-3091.

317. Scallan E, Griffin PM, Angulo FJ, Tauxe RV, Hoekstra RM. 2011. Foodborne illness acquired in the United States-unspecified agents. Emerg. Infect. Dis. 17:16-22.

318. Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson MA, Roy SL, Jones JL, Griffin PM. 2011. Foodborne illness acquired in the United States- major pathogens. Emerg. Infect. Dis. 17:7-15.

319. Scharff RL. 2012. Economic burden from health losses due to foodborne illness in the United States. J. Food Prot. 75:123-131.

320. Schlech WF, III, Lavigne PM, Bortolussi RA, Allen AC, Haldane EV, Wort AJ, Hightower AW, Johnson SE, King SH, Nicholls ES, Broome CV. 1983. Epidemic listeriosis - evidence for transmission by food. N. Eng. J. Med. 308:203-206.

321. Schmidt H, Scheef J, Huppertz HI, Frosch M, Karch H. 1999. Escherichia coli O157:H7 and O157:H(-) strains that do not produce Shiga toxin: phenotypic and genetic characterization of isolates associated with diarrhea and hemolytic- uremic syndrome. J. Clin. Microbiol. 37:3491-3496.

322. Schwartz DC, Saffran W, Welsh J, Haas R, Goldenberg M, Cantor CR. 1983. New techniques for purifying large DNAs and studying their properties and packaging. Cold Spring Harb. Symp. Quant. Biol. 47:189-195.

323. Seeliger HPR, Höhne K. 1979. Chapter 11: serotyping of Listeria monocytogenes and related species. Meth. Microbiol. 13:31-49.

Texas Tech University, Alex Brandt, May 2014

324. Selander RK, Beltran P, Smith NH, Helmuth R, Rubin FA, Kopecko DJ, Ferris K, Tall BD, Cravioto A, Musser JM. 1990. Evolutionary genetic relationships of clones of Salmonella serovars that cause human typhoid and other enteric fevers. Infect. Immun. 58:2262-2275.

325. Senczek D, Stephan R, Untermann F. 2000. Pulsed-field gel electrophoresis (PFGE) typing of Listeria strains isolated from a meat processing plant over a 2- year period. Int. J. Food Microbiol. 62:155-159.

326. Shahamat M, Seaman A, Woodbine M. 1980. Survival of Listeria monocytogenes in high salt concentrations. Zent. Bakter. 246:506-511.

327. Shank FR, Elliot EL, Wachsmuth IK, Losikoff ME. 1996. US position on Listeria monocytogenes in foods. Food Contr. 7:229-234.

328. Shere JA, Bartlett KJ, Kaspar CW. 1998. Longitudinal study of Escherichia coli O157:H7 dissemination on four dairy farms in Wisconsin. Appl. Environ. Microbiol. 64:1390-1399.

329. Sinclair JF, O'Brien AD. 2002. Cell surface-localized nucleolin is a eukaryotic receptor for the adhesin intimin-gamma of enterohemorrhagic Escherichia coli O157:H7. J. Biol. Chem. 277:2876-2885.

330. Skandamis PN, Gounadaki AS, Geornaras I, Sofos JN. 2012. Adaptive acid tolerance response of Listeria monocytogenes strains under planktonic and immobilized growth conditions. Int. J. Food Microbiol. 159:160-166.

331. Slutsker L, Ries AA, Maloney K, Wells JG, Greene KD, Griffin PM. 1998. A nationwide case-control study of Escherichia coli O157:H7 infection in the United States. J. Infect. Dis. 177:962-966.

332. Smith MA, Takeuchi K, Brackett RE, McClure HM, Raybourne RB, Williams KM, Babu US, Ware GO, Broderson JR, Doyle MP. 2003. Nonhuman primate model for Listeria monocytogenes-induced stillbirths. Infect. Immun. 71:1574-1579.

333. Southern EM. 1979. Analysis of restriction-fragment patterns from complex deoxyribonucleic acid species. Biochem. Soc. Symp. 44:37-41.

334. Steenackers H, Hermans K, Vanderleyden J, De Keersmaecker SCJ. 2012. Salmonella biofilms: an overview on occurrence, structure, regulation and eradication. Food Res. Int. 45:502-531.

Texas Tech University, Alex Brandt, May 2014

335. Strawn LK, Fortes ED, Bihn EA, Nightingale KK, Grohn YT, Worobo RW, Wiedmann M, Bergholz PW. 2013. Landscape and meteorological factors affecting prevalence of three food-borne pathogens in fruit and vegetable farms. Appl. Environ. Microbiol. 79:588-600.

336. Strockbine NA, Marques LR, Newland JW, Smith HW, Holmes RK, O'Brien AD. 1986. Two toxin-converting phages from Escherichia coli O157:H7 strain 933 encode antigenically distinct toxins with similar biologic activities. Infect. Immun. 53:135-140.

337. Stull TL, LiPuma JJ, Edlind TD. 1988. A broad-spectrum probe for molecular epidemiology of bacteria: ribosomal RNA. J. Infect. Dis. 157:280-286.

338. Swaminathan B, Barrett TJ, Hunter SB, Tauxe RV. 2001. PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg. Infect. Dis. 7:382-389.

339. Swaminathan B, Gerner-Smidt P, Ng LK, Lukinmaa S, Kam KM, Rolando S, Gutierrez EP, Binsztein N. 2006. Building PulseNet International: an interconnected system of laboratory networks to facilitate timely public health recognition and response to foodborne disease outbreaks and emerging foodborne diseases. Foodborne Pathog. Dis. 3:36-50.

340. Swanson JA, Baer SC. 1995. Phagocytosis by zippers and triggers. Trends Cell Biol. 5:89-93.

341. Tarr PI, Gordon CA, Chandler WL. 2005. Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet 365:1073-1086.

342. Tenaillon O, Skurnik D, Picard B, Denamur E. 2010. The population genetics of commensal Escherichia coli. Nat. Rev. Microbiol. 8:207-217.

343. Teunis PFM, Kasuga F, Fazil A, Ogden ID, Rotariu O, Strachan NJC. 2010. Dose–response modeling of Salmonella using outbreak data. Int. J. Food Microbiol. 144:243-249.

344. Thévenot D, Delignette-Muller ML, Christieans S, Leroy S, Kodjo A, Vernozy-Rozand C. 2006. Serological and molecular ecology of Listeria monocytogenes isolates collected from 13 French pork meat salting–curing plants and their products. Int. J. Food Microbiol. 112:153-161.

Texas Tech University, Alex Brandt, May 2014

345. Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Tracking of Listeria monocytogenes in smoked fish processing plants. J. Food Prot. 67:328- 341.

346. Thimothe J, Walker J, Suvanich V, Gall KL, Moody MW, Wiedmann M. 2002. Detection of Listeria in crawfish processing plants and in raw, whole crawfish and processed crawfish (Procambarus spp.). J. Food Prot. 65:1735- 1739.

347. Thomas LV, Wimpenny JW, Peters AC. 1992. Testing multiple variables on the growth of a mixed inoculum of Salmonella strains using gradient plates. Int. J. Food Microbiol. 15:165-175.

348. Threlfall EJ, Ward LR, Rowe B. 1978. Spread of multiresistant strains of Salmonella typhimurium phage types 204 and 193 in Britain. Brit. Med. J. 2:997.

349. Tilney LG, Portnoy DA. 1989. Actin filaments and the growth, movement, and spread of the intracellular bacterial parasite, Listeria monocytogenes. J. Cell Biol. 109:1597-1608.

350. Todd EC, Greig JD, Bartleson CA, Michaels BS. 2008. Outbreaks where food workers have been implicated in the spread of foodborne disease, part 4: infective doses and pathogen carriage. J. Food Prot. 71:2339-2373.

351. Tompkin RB. 2002. Control of Listeria monocytogenes in the food-processing environment. J. Food Prot. 65:709-725.

352. Tuttle J, Gomez T, Doyle MP, Wells JG, Zhao T, Tauxe RV, Griffin PM. 1999. Lessons from a large outbreak of Escherichia coli O157:H7 infections: insights into the infectious dose and method of widespread contamination of hamburger patties. Epidemiol. Infect. 122:185-192.

353. U.S. Department of Agriculture, Food Safety and Inspection Service. 1996. Pathogen reduction; hazard analysis and critical control point (HACCP) systems; final rule. Fed. Reg. 61:38805-38989.

354. U.S. Department of Agriculture, Food Safety and Inspection Service. 1999. Beef products contaminated with Escherichia coli O157:H7. Fed. Reg. 64:2803- 2805.

355. U.S. Department of Agriculture, Food Safety and Inspection Service. 2001. Oklahoma firm recalls meat and poultry products for possible Listeria

Texas Tech University, Alex Brandt, May 2014

contamination. Available At: http://www.fsis.usda.gov/Oa/recalls/prelease/pr021- 2001.htm?redirecthttp=true. Accessed On: 2/12/14

356. U.S. Department of Agriculture, Food Safety and Inspection Service. 2003. Control of Listeria monocytogenes in ready-to-eat meat and poultry products; final rule. Fed. Reg. 68:34207-34254.

357. U.S. Department of Agriculture, Food Safety and Inspection Service. 2010. Rhode Island firm recalls Italian sausage products due to possible Salmonella contamination Available At: http://www.fsis.usda.gov/wps/wcm/connect/fsis- archives-content/internet/main/topics/recalls-and-public-health-alerts/recall-case- archive/archives/ct_index13a. Accessed On: 1/29/14

358. U.S. Department of Agriculture, Food Safety and Inspection Service. 2012. FSIS verification testing for non-O157 Shiga toxin-producing Escherichia coli (non-O157 STEC) under MT60, MT52, and MT53 sampling programs. Available At: http://foodpoisoningbulletin.com/wp-content/uploads/FSISNONSTECNOTIC E.pdf. Accessed On: 2/2/14

359. U.S. Department of Agriculture, Food Safety and Inspection Service. 2014. Food Safety and Inspection Service’s annual sampling program plan fiscal year 2014. Available At: http://www.fsis.usda.gov/wps/wcm/connect/62fad225-9052- 47fa-9ffe-397f436dc96a/Sampling-Program-Plan-FY2014.pdf?MOD=AJPERES. Accessed On: 1/30/14

360. U.S. Department of Agriculture, Food Safety and Inspection Service. 2014. FSIS compliance guideline: controlling Listeria monocytogenes in post-lethality exposed ready-to-eat meat and poultry products. Available At: http://www.fsis. usda.gov/shared/PDF/Controlling_LM_RTE_guideline_0912.pdf. Accessed On: 2/12/14

361. U.S. Department of Agriculture, Food Safety and Inspection Service. 2014. Recalls and public health alerts. Available At: http://www.fsis.usda.gov /wps/portal/fsis/topics/recalls-and-public-health-alerts/recall-summaries/!ut/p/a1/0 4_Sj9CPykssy0xPLMnMz0vMAfGjzOINAg3MDC2dDbz83RzdDDz9jN3CLPzc DQ38zYAKIlEUWBqCFIQF-ns7OxtY-BkTqR8HcDQgpN-LCAuMinydfdP1ow oSSzJ0M_PS8vUjilKTE3NydItLc3MTizJTi_XD9aNQTTIwBEKgScEmHl5-xgb- JugKsHgVogC3XwpyQyOqfNKCPdMVFQE1ZBvN/?1dmy¤t=true&urile =wcm%3apath%3a%2FFSIS-Content%2Finternet%2Fmain%2Ftopics%2Frecalls -and-public-health-alerts. Accessed On: 4/19/14

362. U.S. Department of Health and Human Services. 2013. Food safety - healthy people 2020. Available At: http://www.healthypeople.gov/2020/topicsobjectives 2020/overview.aspx?topicid=14. Accessed On: 1/12/14 78

Texas Tech University, Alex Brandt, May 2014

363. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1985. Listeriosis outbreak associated with Mexican-style cheese. Morb. Mortal. Wkly. Rep. 34:357-359.

364. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1989. Listeriosis associated with consumption of turkey franks. Morb. Mortal. Wkly. Rep. 38:267-268.

365. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1993. Update: multistate outbreak of Escherichia coli O157:H7 infections from hamburgers--western United States, 1992-1993. J. Am. Med. Assoc. 269:2194-2196.

366. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1995. Escherichia coli O157:H7 outbreak linked to commercially distributed dry-cured salami - Washington and California, 1994. Morb. Mortal. Wkly. Rep. 44:157.

367. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1996. Outbreak of Escherichia coli O157:H7 infections associated with drinking unpasteurized commercial apple juice--British Columbia, California, Colorado, and Washington, October 1996. Morb. Mortal. Wkly. Rep. 45:975.

368. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1998. Multistate outbreak of listeriosis - United States, 1998. Morb. Mortal. Wkly. Rep. 47:1085-1086.

369. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1998. Multistate outbreak of Salmonella serotype Agona infections linked to toasted oats cereal--United States, April-May, 1998. J. Am. Med. Assoc. 280:411.

370. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1999. Outbreak of Escherichia coli O157:H7 and Campylobacter among attendees of the Washington County Fair-New York, 1999. Morb. Mortal. Wkly. Rep. 48:803-805.

371. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1999. Update: multistate outbreak of listeriosis - United States, 1998-1999. Morb. Mortal. Wkly. Rep. 47:1117-1118.

Texas Tech University, Alex Brandt, May 2014

372. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2000. Multistate outbreak of listeriosis - United States, 2000. Morb. Mortal. Wkly. Rep. 49:1129-1130.

373. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2002. Public health dispatch: outbreak of listeriosis - northeastern United States, 2002. Morb. Mortal. Wkly. Rep. 51:950-951.

374. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2006. Multi-state outbreak of E. coli O157:H7 infections from spinach. Available At: http://www.cdc.gov/ecoli/2006/september/updates. Accessed On: 2/2/14

375. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2008. Multistate outbreak of Salmonella Agona infections linked to rice & wheat puff cereal (final update). Available At: http://www.cdc .gov/salmonella/agona. Accessed On: 1/15/14

376. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2008. Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce items--United States, 2008. Morb. Mortal. Wkly. Rep. 57:929-934.

377. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2009. Multistate outbreak of human Salmonella Saintpaul infections linked to raw alfalfa sprouts (final update). Available At: http://www. cdc.gov/salmonella/saintpaul-alfalfa/index.html. Accessed On: 1/29/14

378. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2009. Multistate outbreak of Salmonella infections linked to pistachio nuts (final update). Available At: http://www.cdc.gov/salmonella/ pistachios/index.html. Accessed On: 1/29/14

379. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2009. Multistate outbreak of Salmonella Typhimurium infections linked to peanut butter, 2008–2009. Available At: http://www.cdc.gov/ salmonella/typhimurium. Accessed On: 1/29/14

380. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2010. Multistate outbreak of human Salmonella Enteritidis infections associated with shell eggs (final update). Available At: http://www.cdc. gov/salmonella/enteritidis. Accessed On: 1/29/14

Texas Tech University, Alex Brandt, May 2014

381. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2010. Multistate outbreak of human Salmonella Montevideo infections (final update). Available At: http://www.cdc.gov/salmonella/monte video/index.html. Accessed On: 1/29/14

382. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2011. Multistate outbreak of human Salmonella Heidelberg infections linked to ground turkey. Available At: http://www.cdc.gov/salmonella /heidelberg/index.html. Accessed On: 1/29/14

383. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2011. Multistate outbreak of human Salmonella I 4,[5],12:i:- infections linked to alfalfa sprouts (final update). Available At: http://www.cdc. gov/salmonella/i4512i-/index.html. Accessed On: 1/29/14

384. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2012. Multistate outbreak of Salmonella Enteritidis infections linked to ground beef (final update). Available At: http://www.cdc.gov /salmonella/enteritidis-07-12/index.html. Accessed On: 1/29/14

385. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2012. Multistate outbreak of Salmonella Typhimurium and Salmonella Newport infections linked to cantaloupe (final update). Available At: http://www.cdc.gov/salmonella/typhimurium-cantaloupe-08-12/index.html. Accessed On: 1/29/14

386. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2013. Foodborne diseases active surveillance network (FoodNet). Available At: http://www.cdc.gov/foodnet/about.html. Accessed On: 1/20/14

387. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2013. Outbreak of Escherichia coli O104:H4 infections associated with sprout consumption - Europe and North America, May-July 2011. Morb. Mortal. Wkly. Rep. 62:1029-1031.

388. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2014. Multistate outbreak of multidrug-resistant Salmonella Heidelberg infections linked to Foster Farms brand chicken. Available At: http://www.cdc.gov/salmonella/heidelberg-10-13/index.html. Accessed On: 1/29/14

389. U.S. Department of Health and Human Services, Food and Drug Administration. 2013. Risk & safety assessment. Available At: http://www.fda. 81

Texas Tech University, Alex Brandt, May 2014

gov/Food/FoodScienceResearch/RiskSafetyAssessment/default.htm. Accessed On: 1/29/14

390. U.S. Department of Health and Human Services, Food and Drug Administration, U.S. Department of Agriculture, Food Safety Inspection Service. 2003. Quantitative assessment of relative risk to public health from foodborne Listeria monocytogenes among selected categories of ready-to-eat foods. Available At: http://www.fda.gov/Food/FoodScienceResearch/RiskSafety Assessment/ucm183966.htm. Accessed On: 2/12/14

391. Uchida H, Kiyokawa N, Horie H, Fujimoto J, Takeda T. 1999. The detection of Shiga toxins in the kidney of a patient with hemolytic uremic syndrome. Pediatr. Res. 45:133-137.

392. Uchiya K, Barbieri MA, Funato K, Shah AH, Stahl PD, Groisman EA. 1999. A Salmonella virulence protein that inhibits cellular trafficking. EMBO J. 18:3924-3933.

393. Uesugi AR, Danyluk MD, Mandrell RE, Harris LJ. 2007. Isolation of Salmonella Enteritidis phage type 30 from a single almond orchard over a 5-year period. J. Food Prot. 70:1784-1789.

394. Unnerstad H, Bannerman E, Bille J, Danielsson-Tham ML, Waak E, Tham W. 1996. Prolonged contamination of a dairy with Listeria monocytogenes. Neth. Milk Dairy J. 50:493-500.

395. van Belkum A. 1999. Short sequence repeats in microbial pathogenesis and evolution. Cell Mol. Life Sci. 56:729-734.

396. van der Veen S, Abee T. 2011. Contribution of Listeria monocytogenes RecA to acid and bile survival and invasion of human intestinal Caco-2 cells. Int. J. Med. Microbiol. 301:334-340.

397. van der Veen S, van Schalkwijk S, Molenaar D, de Vos WM, Abee T, Wells- Bennik MH. 2010. The SOS response of Listeria monocytogenes is involved in stress resistance and mutagenesis. Microbiol. 156:374-384.

398. Van Stelten A, Simpson JM, Chen Y, Scott VN, Whiting RC, Ross WH, Nightingale KK. 2011. Significant shift in median guinea pig infectious dose shown by an outbreak-associated Listeria monocytogenes epidemic clone strain and a strain carrying a premature stop codon mutation in inlA. Appl. Environ. Microbiol. 77:2479-2487.

Texas Tech University, Alex Brandt, May 2014

399. Van Stelten A, Simpson JM, Ward TJ, Nightingale KK. 2010. Revelation by single-nucleotide polymorphism genotyping that mutations leading to a premature stop codon in inlA are common among Listeria monocytogenes isolates from ready-to-eat foods but not human listeriosis cases. Appl. Environ. Microbiol. 76:2783-2790.

400. Vangay P, Fugett EB, Sun Q, Wiedmann M. 2013. Food Microbe Tracker: a web-based tool for storage and comparison of food-associated microbes. J. Food Prot. 76:283-294.

401. Vatanyoopaisarn S, Nazli A, Dodd CE, Rees CE, Waites WM. 2000. Effect of flagella on initial attachment of Listeria monocytogenes to stainless steel. Appl. Environ. Microbiol. 66:860-863.

402. Vázquez-Boland JA, Domínguez-Bernal G, González-Zorn B, Kreft J, Goebel W. 2001. Pathogenicity islands and virulence evolution in Listeria. Microb. Infect. 3:571-584.

403. Vazquez-Boland JA, Kuhn M, Berche P, Chakraborty T, Dominguez-Bernal G, Goebel W, Gonzalez-Zorn B, Wehland J, Kreft J. 2001. Listeria pathogenesis and molecular virulence determinants. Clin. Microbiol. Rev. 14:584- 640.

404. Verghese B, Lok M, Wen J, Alessandria V, Chen Y, Kathariou S, Knabel S. 2011. comK prophage junction fragments as markers for Listeria monocytogenes genotypes unique to individual meat and poultry processing plants and a model for rapid niche-specific adaptation, biofilm formation, and persistence. Appl. Environ. Microbiol. 77:3279-3292.

405. Versalovic J, Koeuth T, Lupski JR. 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucl. Acid. Res. 19:6823-6831.

406. Vestby LK, Møretrø T, Langsrud S, Heir E, Nesse LL. 2009. Biofilm forming abilities of Salmonella are correlated with persistence in fish meal- and feed factories. BMC Vet. Res. 5:20.

407. Vogel BF, Jørgensen LV, Ojeniyi B, Huss HH, Gram L. 2001. Diversity of Listeria monocytogenes isolates from cold-smoked salmon produced in different smokehouses as assessed by random amplified polymorphic DNA analyses. Int. J. Food Microbiol. 65:83-92.

Texas Tech University, Alex Brandt, May 2014

408. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M. 1995. AFLP: a new technique for DNA fingerprinting. Nucl. Acid. Res. 23:4407-4414.

409. Watkins J, Sleath KP. 1981. Isolation and enumeration of Listeria monocytogenes from sewage, sewage sludge and river water. J. Appl. Bacteriol. 50:1-9.

410. Weis J, Seeliger HPR. 1975. Incidence of Listeria monocytogenes in nature. Appl. Microbiol. 30:29-32.

411. Weiss SM, Ladwein M, Schmidt D, Ehinger J, Lommel S, Stading K, Beutling U, Disanza A, Frank R, Jansch L, Scita G, Gunzer F, Rottner K, Stradal TE. 2009. IRSp53 links the enterohemorrhagic E. coli effectors Tir and EspFU for actin pedestal formation. Cell Host Microb. 5:244-258.

412. Welshimer HJ, Donker-Voet J. 1971. Listeria monocytogenes in nature. Appl. Microbiol. 21:516-519.

413. Wiedmann M. 2002. Subtyping of bacterial foodborne pathogens. Nutr. Rev. 60:201-208.

414. Wiedmann M, Arvik TJ, Hurley RJ, Boor KJ. 1998. General stress transcription factor σB and its role in acid tolerance and virulence of Listeria monocytogenes. J. Bacteriol. 180:3650-3656.

415. Wiedmann M, Bruce JL, Keating C, Johnson AE, McDonough PL, Batt CA. 1997. Ribotypes and virulence gene polymorphisms suggest three distinct Listeria monocytogenes lineages with differences in pathogenic potential. Infect. Immun. 65:2707-2716.

416. Williams JG, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucl. Acid. Res. 18:6531-6535.

417. Williams SK, Roof S, Boyle EA, Burson D, Thippareddi H, Geornaras I, Sofos JN, Wiedmann M, Nightingale K. 2011. Molecular ecology of Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. J. Food Prot. 74:63-77.

418. Wilson RK, Shaw RK, Daniell S, Knutton S, Frankel G. 2001. Role of EscF, a putative needle complex protein, in the type III protein translocation system of enteropathogenic Escherichia coli. Cell Microbiol. 3:753-762.

Texas Tech University, Alex Brandt, May 2014

419. Wu KY, Liu GR, Liu WQ, Wang AQ, Zhan S, Sanderson KE, Johnston RN, Liu SL. 2005. The genome of Salmonella enterica serovar Gallinarum: distinct insertions/deletions and rare rearrangements. J. Bacteriol. 187:4720-4727.

420. Wulff G, Gram L, Ahrens P, Vogel BF. 2006. One group of genetically similar Listeria monocytogenes strains frequently dominates and persists in several fish slaughter- and smokehouses. Appl. Environ. Microbiol. 72:4313-4322.

421. Zeiner SA, Dwyer BE, Clegg S. 2012. FimA, FimF, and FimH are necessary for assembly of type 1 fimbriae on Salmonella enterica serovar Typhimurium. Infect. Immun. 80:3289-3296.

422. Zhang C, Nietfeldt J, Zhang M, Benson AK. 2005. Functional consequences of genome evolution in Listeria monocytogenes: the lmo0423 and lmo0422 genes encode σC and LstR, a lineage II-specific heat shock system. J. Bacteriol. 187:7243-7253.

423. Zhang W, Jayarao BM, Knabel SJ. 2004. Multi-virulence-locus sequence typing of Listeria monocytogenes. Appl. Environ. Microbiol. 70:913-920.

424. Zhang XL, Jeza VT, Pan Q. 2008. Salmonella typhi: from a human pathogen to a vaccine vector. Cell Mol. Immunol. 5:91-97.

425. Zhao T, Doyle MP, Besser RE. 1993. Fate of enterohemorrhagic Escherichia coli O157:H7 in apple cider with and without preservatives. Appl. Environ. Microbiol. 59:2526-2530.

Texas Tech University, Alex Brandt, May 2014

TABLE 1.1 Summary of selected studies where L. monocytogenes persistence was inferred Number of Persistence Number Facility Subtyping Presumed Duration Country of Reference Typea Method(s)c Persistent Able to be Facilitiesb Subtypesd Inferred MLEE, Milk; Dairy Northern 4; 3 RFLP, 4; 2 1-6 months (145) Farm Ireland Serotyping

Northern RAPD, Poultry 1 2 5-6 months (201) Ireland MLEE

Salmon MLEE, 4 weeks-8 Slaughter and Norway 1 2 (305) Serotyping months Smoking

Cold-Smoked United 3 Ribotyping 4 3.5-7 months (268) Fish States

RTE Chilled United 4 Ribotyping 7 8-21 months (157) Food Kingdom

Dairy Barn; United Cheese 1;1 Ribotyping 1;1 18 months (153) States Processing

Mussel New PFGE, 3 3 3-12 months (75) Processing Zealand Serotyping

RTE Smoked United 2 Ribotyping 4 5-12 months (230) Fish States

PFGE, Pork Virulence Sardinia 5 3 ≤ 3 years (243) Slaughter Gene PCR, Serotyping

Cheese PFGE, Brazil 3 4 3-9 months (19) Processing Serotyping

Texas Tech University, Alex Brandt, May 2014

TABLE 1.1, Cont. Summary of selected studies where L. monocytogenes persistence was inferred Number of Persistence Number Facility Subtyping Presumed Duration Country of Reference Typea Method(s)c Persistent Able to be Facilitiesb Subtypesd Inferred

MLEE, RTE Meat Norway 5 1 4 years (261) RFLP

PFGE, Cheese Scandinavia 1 1 7 years (394) Serotyping

RAPD, Pork PFGE, Slaughter and France 5 6 11-14 months (125) PCR-REA, Cutting Serotyping

Smoked and PFGE, Cold-Salted Finland 1 1 14 months (170) Serotyping Fish

PFGE, Ice Cream Finland 1 1 7 years (247) Serotyping

Turkey Denmark 1 PFGE 2 2 months (270)

Salmon Norway, MLEE, Slaughter; 2; 4 3; 1 3-45 months (303) Europe REA Smokehouse

Meat Switzerland 1 PFGE 2 2 years (325)

PFGE, 9 months; 3 Poultry; Pork France 1; 1 1 (57) Serotyping months

Cold-Smoked PFGE, France 3 1 2 months (79) Salmon Serotyping

Texas Tech University, Alex Brandt, May 2014

Cold-Smoked RAPD, Denmark 10 2 9 months (407) Salmon PFGE

PFGE, RTE Meat Finland 3 3 4-15 months (223) Serotyping

RTE United 2 Ribotyping 0 N/A (346) Crawfish States

Cold Smoked United 2 Ribotyping 10 3-16 weeks (155) Fish States

RTE Meat; PFGE, Finland 3; 1 14; 6 ≥ 3 months (224) Poultry Serotyping

Cold Smoked RAPD, Sliced Poland 1 3 8-10 months (242) RFLP-PFGE Salmon

Beef Slaughter; Italy 1; 1 PFGE 1 16 months (283) Pork Slaughter

MLEE, Poultry Norway 1 1 2 years (304) REA

RTE Spain 1 RAPD, REA 0 N/A (7) Vegetables

9 months-4 RTE Meat Norway 4 PFGE 6 (147) years

Texas Tech University, Alex Brandt, May 2014

RTE Smoked United 4 Ribotyping 10 3-22 months (198) Fish States

Ribotyping, Retail RTE United 10 hlyA PCR- 5 2-4 weeks (313) Foods States RFLP

RTE Smoked United 4 Ribotyping 9 1-10 months (345) Fish States

Chicken United actA 6 weeks-12 Further 1 8 (30) States Sequencing months Processing

PFGE, United Turkey 2 Epidemic 4 4-9 months (100) States Clone PCR

Cooked PFGE, 13 months- Peeled Iceland 2 3 (140) Serotyping 3.5 years Shrimp

Salmon Sampling Slaughter; Norway 1; 1 Ribotyping 3; 4 peformed for (190) Smokehouse 31 weeks

Saucissons PFGE, France 13 22 ≥ 2 weeks (344) Processing Serotyping

Fish Denmark; RAPD, Slaugther; Faroe 2; 2 AFLP, PCR- 1 1.5 years (420) Smokehouse Islands RFLP

Texas Tech University, Alex Brandt, May 2014

RTE Chilled Finland 1 AFLP 7 21-94 months (186) Food

PFGE, InlA Broiler Spain 1 Western Blot, 6 ≤ 1.5 years (219) Abbattoir Serotyping

United PFGE, RTE Turkey 1 1 12 years (275) States Ribotyping

Dairy United 4.5-35 Milking 1 PFGE 3 (200) States months Facility

Chicken Sampling United actA Further 1 5 peformed for (29) States Sequencing Processing 21 weeks

Cold-Smoked PFGE, Pork Latvia 1 3 2-62 months (31) Serotyping Processing

PFGE, REP- Sandwich Switzerland 1 PCR,Genetic 2 ≤ 1 year (34) Production Lineage PCR

PFGE, Catfish United 2 ERIC-PCR, 1 1 year (61) Processing States Serotyping

Salmon Norway 5 Ribotyping 8 ≤ 8 months (189) Processing

Pork PFGE, Slaughter and Spain 1 8 9-35 months (278) Serotyping RTE Pork

Texas Tech University, Alex Brandt, May 2014

Persistent Cheese PFGE, Ireland 16 Strains in 2 6-20 months (110) Processing Serotyping Facilities

United PFGE, RTE Meat 6 2 1-3 months (417) States Ribotyping

PFGE, Cold-Smoked Virulence Italy 1 2 ≤ 1 year (86) Salmon Gene PCR, Serotyping PFGE, Cheese Serotyping, Portugal 3 6 2-35 months (10) Processing Epidemic Clone PCR a In studies where multiple facility types were sampled, a semicolon (;) is used to separate the different facility classifications. b For studies with multiple facility types, the numbers of each type of facility are given and are separated with a semicolon in the order corresponding to their mention in the first column. c Abbreviations for subtyping methods include: PFGE (Pulsed Field Gel Electrophoresis), MLEE (Multi-Locus Enzyme Electrophoresis), RAPD (Random Amplified Polymorphic DNA), RFLP (Restriction Fragment Length Polymorphism), REA (Restriction Enzyme Analysis), AFLP (Amplified Fragment Length Polymorphism), REP-PCR (Repetitive Element Palindromic PCR), and ERIC-PCR (Enterobacterial Repetitive Intergenic Consensus PCR); Serotyping was performed either conventionally or with a molecular serogrouping method (92). d For studies with multiple facility types, the numbers of persistent strains in each facility type are given and are separated with a semicolon in the order corresponding to their mention in the first column.

Texas Tech University, Alex Brandt, May 2014

CHAPTER II INTEGRATING LONGITUDINAL SAMPLING, MOLECULAR SUBTYPING, AND STATISTICAL ANALYSIS TO INVESTIGATE PERSISTENCE OF LISTERIA MONOCYTOGENES, SALMONELLA ENTERICA, AND ESCHERICHIA COLI O157:H7 IN MEAT PROCESSING FACILITY ENVIRONMENTS

Introduction Several multi-state foodborne illness outbreaks within the past 20 years have been linked to meat and poultry products contaminated with either Listeria monocytogenes, Salmonella enterica, or Escherichia coli O157:H7 (15, 66, 68, 69, 71, 72, 74). Despite rigorous food safety control measures (35, 62, 64), the three pathogens remain prevalent in products routinely tested by the U.S. Food and Drug Administration and the U.S. Department of Agriculture (USDA) Food Safety and Inspection Service (17). This suggests that current controls may not account for all sources of product contamination. Likely included in these sources is contamination that originates from environmental sites within the food processing establishment. Although transmission of L. monocytogenes from harborage sites within the processing facility to the surfaces of post-lethality exposed ready-to-eat (RTE) meat and poultry products is well recognized for the significant public health hazard it poses (61, 64, 65), little is known about whether S. enterica and E. coli O157:H7 contaminate products in a similar manner. Much of the knowledge regarding transmission of L. monocytogenes within the processing environment is owed to numerous studies that have combined longitudinal sampling and molecular subtyping to probe its ecology in various food processing establishments (reviewed in (7, 15, 61)). One particular challenge associated with L. monocytogenes is its long-term colonization of processing facilities – a phenomenon termed “persistence.” Strain persistence for as long as seven (42, 75) and twelve (51) years has been observed, and several large product recalls and multistate foodborne illness outbreaks have been attributed to persistent L. monocytogenes strains (19, 49, 53, 63, 68, 69). However, recent evidence suggests that this phenomenon may not be exclusive to L. monocytogenes. For instance, a 2008 multistate salmonellosis outbreak linked to puffed cereal was attributed to an S. enterica serovar Agona strain that was also

Texas Tech University, Alex Brandt, May 2014 implicated as the etiologic agent of a foodborne illness outbreak that occurred ten years earlier (67, 70). In 2008 swabs taken from the facility environment were positive for the outbreak strain, and it was concluded that a construction project disturbed a location where the pathogen persisted, and reintroduced it into the facility environment (57). Studies aimed at defining risk factors for persistence have focused on criteria such as surface adherence, disinfectant tolerance, and stress tolerance, but overall these studies have produced inconsistent results (1, 4, 9, 12, 14, 16, 22, 23, 26, 36, 37, 54, 55). Some of this inter-study disagreement may be attributed to “loose” definitions of persistent strains (7). For example, Dauphin et al. (11) defined persistent strains as those that were “isolated on six different sampling dates in a 2-month interval,” while Lundén et al. (37) defined persistent strains as those “found repeatedly in the plant during a period of several months or years.” Inconsistent definitions among studies may lead to inaccurate categorization of strains, which may in turn cloud the overall perspective of persistence. Considering these critical knowledge gaps, the overall objective of this study was to integrate longitudinal sampling, molecular subtyping and statistical analysis to probe persistence of pathogens in the meat processing environment. Specific aims within this overall objective were to i) conduct longitudinal sampling for L. monocytogenes, S. enterica, and E. coli O157:H7 in environmental sites and food products of four small or very small meat processing facilities, ii) perform molecular subtyping to elucidate strain contamination patterns and identify harborage sites, and iii) employ recently-published statistical methods (39) to define strain persistence in a standardized manner. To our knowledge this is the first study to simultaneously investigate the molecular ecology of L. monocytogenes, S. enterica, and E. coli O157:H7 within meat processing facilities, and the first to employ statistical methods to define strain persistence in the meat processing environment. The results of this study also help fill a critical need for small and very small meat processors, who often have limited knowledge of pathogen persistence.

Materials and Methods Selection of Participating Facilities and Sampling Sites Four small or very small meat production facilities (designated A-D) were selected from a convenience sample of 51 establishments contacted via U.S. mail. The 93

Texas Tech University, Alex Brandt, May 2014 four facilities were asked to complete questionnaires at the outset of the study (Supplemental Files 1-4), and were categorized as RTE (n = 2) or fresh (n = 2) meat processing operations. Initial visits were made to determine process flows and to identify environmental sampling sites. In each facility a total of 54 environmental sites were chosen, and sampling site maps were constructed (Figs 2.1-2.4). Sites were grouped together in four different categories: i) drains and floors, ii) non-food contact environmental sites, iii) employee surfaces and employee contact sites, and (D) food contact surfaces and food samples. Additionally, for each RTE facility (A and B), three sites were located in the facility’s retail area. These retail areas were either located off- site from the facility (A), or were directly attached to the facility production area (B). Sponge Sample Collection Facilities were visited monthly for six months from May 2011 to December 2011. Each facility had only one work shift. Each of these shifts were during daytime hours, and samples were collected at the approximate midpoint of this single shift. All 54 sites were sampled for L. monocytogenes, S. enterica, and E. coli O157:H7 using separate sterile sponge-handle sampling apparatuses (commonly referred to as “spongecicles”) pre-moistened by the manufacturer with 10 ml of Neutralizing Buffer (SH-10NB, Solar Biolgicals, Ogdensburg, NY). Sponges were passed over three adjacent surface areas (for the three pathogens) of 2 ft X 2 ft using vertical, horizontal, and diagonal movements. In instances where a 2 ft X 2 ft surface area was unavailable (e.g. drains and cracks), the sampling area was roughly divided into thirds. Pressure was applied while passing sponges over sampling areas, and sponges were flipped during sampling in order to allow both sides to come into contact with the sampling area. Because sponge-handle sampling apparatuses preclude the need for hand contact with the sponge portion, sterile gloves were not needed, and gloves were not changed between samples. However, gloves were sanitized with 70% ethanol when they became lightly soiled, and were changed when they became heavily soiled, or when moving from any drain and floor sites or non-food contact environmental sites to food contact surface sites, employee surfaces, or employee contact sites. Sponges were placed into sterile plastic bags after collection, and were transported to the laboratory on coolant material (ice or cold packs). At the laboratory sponges were held at 4°C, and were processed within 48 h. 94

Texas Tech University, Alex Brandt, May 2014

In addition to environmental sites, composite samples of RTE products (from Facilities A and B) and carcasses (from Facilities C and D) were collected on each sampling day. For RTE product composites, three sets (one per pathogen) of five different RTE products were aseptically removed from packaging, placed into sterile foil- lined bins, and surface-sampled using three separate sponges. RTE product samples were taken from products that were available at each RTE facility: wieners, frankfurters, German sausage, chipolatas, jadgwurst, bratwurst, Debriziner sausage, Spicy Polish sausage, chicken apple sausage, or Swedish potato sausage for Facility A, and summer sausage, snack sticks, hot dogs, beef jerky, pepperoni sticks, frankfurters, hot links, salami, or Andouille sausage for Facility B. Carcass composite samples were collected by sponge-sampling the flank, brisket, neck, and front shank portions of three sets (one per pathogen) of three carcass halves in the facility cooler. The focus of carcass sampling methodology was toward presence/absence detection rather than quantification; thus the need to use templates to delineate exact surface areas of carcass samples was precluded. Primary Enrichment for L. monocytogenes, S. enterica, and E. coli O157:H7 Sponge samples were enriched for L. monocytogenes, S. enterica, and E. coli O157:H7 by adding 90 ml of either UVM Modified Listeria Enrichment Broth (UVM; Becton, Dickinson and Company, Franklin Lakes, NJ), Buffered Peptone Water (BPW; Becton, Dickinson and Company), or Modified Tryptone Soya Broth (mTSB; Oxoid, Thermo Fisher Scientific, Waltham, MA) with 10 g/L Casamino Acids (Becton, Dickinson and Company) supplemented with 20 mg/L Novobiocin Sodium Salt (N; MP Biomedicals, Santa Ana, CA), respectively. Sponge samples were mechanically homogenized for 2 min and were incubated at 30±2°C for 22±2 h (UVM), 35±2°C for 22±2 h (BPW), or 42±1ºC for 18.5±3.5 h (mTSB+N). Isolation of L. monocytogenes and non-pathogenic Listeria species A 100 µl aliquot of incubated UVM was transferred to a Modified Oxford Agar (MOX; Becton, Dickinson and Company) plate and streaked for isolation. A second 100 µl aliquot was transferred to 10 ml of Morpholinepropane Sulfonic Acid Buffered- Listeria Enrichment Broth (MOPS-BLEB; made in-house from Becton, Dickinson and Company and Sigma-Aldrich, St. Louis, MO ingredients). MOX plates and MOPS-BLEB tubes were incubated at 35±2°C for 52±4 h and 21±3 h, respectively. A 100 µl aliquot of 95

Texas Tech University, Alex Brandt, May 2014 incubated MOPS-BLEB was transferred to a second MOX plate, streaked for isolation, and incubated as above. MOX plates were examined for colonies typical of Listeria. Up to eight colonies per sample (from a combination of both MOX plates) were sub-streaked to Listeria monocytogenes Chromogenic Plating Medium (LMPM; R&F Laboratories, Downers Grove, IL) plates, and incubated at 35±2°C for 24±2 h to determine the preliminary species classification. On LMPM L. monocytogenes and Listeria ivanovii appear blue, while other Listeria species appear white. LMPM plate growth was substreaked to Brain Heart Infusion Agar (BHI Agar; Becton, Dickinson and Company) and incubated at 37±2°C for 24±2 h. A ∆actA mutant strain of L. monocytogenes and Listeria innocua ATCC 33090 were used to inoculate positive control samples. Isolation of S. enterica S. enterica selective enrichment was performed by transferring 500 µl and 100 µl of incubated BPW to 10 ml of Tetrathionate Broth, Hajna (TTB; Becton, Dickinson and Company) and 10 ml of Rappaport-Vassiliadis Soya Broth (RVS; made in-house according to EMD Science recipe), respectively. TTB and RVS were incubated in a 42±0.5°C water bath for 21±3 h. Aliquots (10 µl) of each medium were streaked for isolation on Brilliant Green Sulfa Agar (BGS; Becton, Dickinson and Company) and Xylose Lysine Tergitol 4 Agar (XLT4; Becton, Dickinson and Company). BGS and XLT4 plates were incubated at 35±2°C for 42±6 h, and were examined for colonies typical of S. enterica. If present, up to eight typical colonies (combined together from the four plates) were sub-streaked to CHROMAgar Salmonella (CHROMAgar, Paris, France), incubated at 35±2°C for 24±2 h, and examined for growth typical of S. enterica. Presumptive positive colonies were sub-streaked to BHI Agar plates and incubated at 37±2°C for 24±2 h. A green fluorescent protein-expressing mutant of S. enterica serovar Typhimurium ATCC 700408 (48) was used to inoculate positive control samples. Isolation of E. coli O157:H7 Aliquots of mTSB+N were combined with anti-E. coli O157 immunomagnetic beads (Dynabeads, Life Technologies, Thermo Fisher Scientific) and subjected to immunomagnetic separation (IMS) according to either manual (Samplings 1-2) or automated (Samplings 3-6) processing procedures (Bead Retriever System; Applied Biosystems, Life Technologies, Thermo Fisher Scientific). For the first sampling 96

Texas Tech University, Alex Brandt, May 2014 occasions, a 100 µl aliquot of IMS bead solution was streaked for isolation on Sorbitol MacConkey Agar (Becton, Dickinson and Company) supplemented with 20 mg/L Novobiocin Sodium Salt and 2.5 mg/L Potassium Tellurite (MP Biomedicals). Plates were incubated at 37±2°C for 21±3 hours and examined for typical E. coli O157:H7 colonies. Presumptive positive colonies were sub-streaked to CHROMAgar O157 (CHROMAgar) plates, incubated at 37±2°C for 21±3 h, and examined for mauve-colored growth typical of E. coli O157:H7. On all subsequent sampling days (2-6) IMS bead solutions were streaked for isolation on CHROMAgar O157. To reduce background microflora, 20 mg/L Novobiocin Sodium Salt and 2.5 mg/L Potassium Tellurite were added to CHROMAagar O157 in Samplings 5 and 6. In all instances, typical E. coli O157:H7 growth on CHROMAgar O157 was sub-streaked to BHI Agar and incubated at 37±2°C for 24±2 h. An ATCC 43895 green fluorescent protein-expressing E. coli O157:H7 mutant strain (48) was used to inoculate positive control samples. Confirmation of L. monocytogenes, non-pathogenic Listeria species, S. enterica, and E. coli O157:H7 by organism-specific PCR Assays PCR assays were used to confirm presumptive positive colonies from each isolation procedure. For L. monocytogenes confirmation, a PCR which amplifies a region of the listeriolysin-O gene, hlyA (based on (50)), was used. Confirmation of non- pathogenic Listeria species was accomplished with a PCR that amplifies a region of the stress response regulator gene, sigB (based on (47)). Presumptive S. enterica were confirmed via a PCR that amplifies a section of the invA gene (based on (32)), and presumptive E. coli O157:H7 were confirmed using a six-gene multiplex PCR targeting regions of the hlyE, eaeA, rfbE, fliCH7, stx1, and stx2 genes (40). For each assay, cells from BHI Agar were transferred to a 96-well microtiter PCR plate using a sterile toothpick; lysis was peformed by microwaving for either 4 min for Gram-positives or 30 s for Gram-negatives. PCR master mix volumes (25 µl) contained 12.5 µl of 2X GoTaq Green Master Mix (Promega, Madison, WI) and forward and reverse primers (Table 2.1) at concentrations specified below; nuclease free water accounted for the remaining volume. For the hlyA, sigB, and invA assays, 1 µl of each primer was added at concentrations of 20 µM, 12.5 µM, and 10 µM, respectively. For the E. coli O157:H7 six-gene multiplex assay, 0.5 µl each of primers stx1-F, stx1-R, stx2-F, 97

Texas Tech University, Alex Brandt, May 2014 stx2-R, EC hly F, and EC hly R at 10 µM, 0.25 µl each of primers rfb-F and rfb-R at 10 µM, 0.19 µl each of primers int-F and int-R at 10 µM, and 0.15 µl each of flich7-F and flich7-R at 10 µM were added. Thermal cycling (conditions in Table 2.1) were carried out an Applied Biosystems 2720 Thermal Cycler (Life Technologies, Thermo Fisher Scientific) and PCR products were electrophoresed for 1 h at 110 V on a 2.0% Agarose gel stained with SYBR Safe DNA Gel Stain (Invitrogen, Life Techonologies, Thermo Fisher Scientific). Gel images were visualized using an ultraviolet transilluminator, and captured using gel imaging software. For L. monocytogenes, non-pathogenic Listeria species, and S. enterica, the presence of single bands with sizes of 858, 840, or 677 bp, respectively, was considered indicative of a positive identification for the isolate. For the E. coli O157:H7 six-gene multiplex assay, presence of 772, 625, and 292 bp bands corresponding to hlyE, fliCH7, and rfbE, respectively, was considered criteria for positive identification. The presence of 484, 368, and 210 bp bands, corresponding to virulence genes stx2, eaeA, and stx1, respectively, was also evaluated for each isolate to indicate its virulence profile. Up to four confirmed L. monocytogenes, two confirmed non-pathogenic Listeria species, eight confirmed S. enterica, and eight confirmed E. coli O157:H7 isolates from each sample were selected for preservation. The remainder of colonies used for PCR confirmations were transferred to BHI Broth (Becton, Dickinson, and Company) and incubated overnight at 37°C with agitation. Cultures were mixed with 15% glycerol and frozen at -80°C for long-term preservation and subsequent use in molecular subtyping. Molecular Subtyping of L. monocytogenes All preserved L. monocytogenes isolates were subjected to a modified multiplex PCR assay (13) to place the isolate into one of four serogroups, which each contain some subset of the 13 L. monocytogenes serotypes: I.1 (serotypes 1/2a and 3a), I.2 (serotypes 1/2c and 3c), II.1 (serotypes 4b, 4d, and 4e), and II.2 (serotypes 1/2b, 3b, and 7). Briefly, cells were prepared and lysed as noted above. The PCR master mix composition followed the 25 µl GoTaq Green Master Mix setup described above with 1.25 µl each of the primers lmo0737-F, lmo0737-R, ORF2110-F, ORF-2110-R, ORF2819-F, and ORF2819- R at 20 µM, 1.875 µl each of primers lmo1118-F and lmo1118-R at 20 µM, and 0.25 ul each of primers prs-F and prs-R at 20 µM (Table 2.1). Thermal cycling conditions (Table 98

Texas Tech University, Alex Brandt, May 2014

2.1) and gel electrophoresis were carried out as above, except that 1.5% agarose gels were used and products were electrophoresed for 2.5 h at 80 V. Patterns were evaluated as previously described (13), and L. monocytogenes strains of each serogroup were used as positive controls alongside L. innocua ATCC 33090 as a prs band-only control. One L. monocytogenes isolate per positive sample was further subtyped via automated EcoRI ribotyping (5) using a Riboprinter (DuPont Qualicon, Wilmington, DE) at the Cornell University Laboratory for Molecular Typing (Ithaca, NY). In agreement with previously-published methods (58), ribotype designations encompassing several different patterns differing by a single minor band were further distinguished by adding a single letter as a suffix to the ribotype designation (i.e. DUP-1062B and DUP-1062E). Molecular Subtyping of Non-Pathogenic Listeria Species A sigB allelic typing scheme (based on (47)) was used to subtype one non- pathogenic Listeria species isolate from each positive sample. A sigB allelic type (AT) is defined as a unique combination of sequence polymorphisms within sigB. Briefly, cell preparations, lysis, and PCR master mix formulations were performed as described above for the sigB confirmation PCR with the exception that GoTaq Colorless Master Mix (Promega, Madison, WI) was used in place of GoTaq Green Master Mix. The volume of PCR product remaining after gel confirmation (20 µl) was purified with a GenElute PCR Clean-Up Kit (Sigma-Aldrich), checked for purity and concentration on a NanoDrop 2000 UV-Vis Spectrophotometer (Thermo Fisher Scientific), and frozen at -20°C. For sequencing submission, purified products were combined with either SigB 15 primer or SigB 16 primer. Automated Sanger sequencing was performed at the Cornell University Biotechnology Resource Center (Ithaca, NY) using Big Dye Terminator Chemistry and AmpliTaq-FS DNA Polymerase (Life Technologies, Thermo Fisher Scientific), and PCR products were analyzed on an Applied Biosystems Automated 3730xl DNA Analyzer (Life Technologies, Thermo Fisher Scientific). sigB sequences were assembled, trimmed, and proofread using SeqMan Pro, and consensus sequences were aligned and trimmed using MegAlign. Both programs are part of the Lasergene Core Suite Version 9 (DNASTAR, Inc., Madison, WI). Sequences were aligned with a 660 bp alignment of sigB sequences corresponding to currently identified

Texas Tech University, Alex Brandt, May 2014 sigB ATs (Supplemental File 5). Alignments were uploaded into DnaSP Software Version 5 (University of Barcelona, Barcelona, Spain) to assign sigB ATs to each isolate. Pulsed Field Gel Electrophoresis Typing of S. enterica and E. coli O157:H7 One isolate of S. enterica and E. coli O157:H7 from each positive sample was subjected to XbaI pulsed field gel electrophoresis (PFGE) based on a standardized protocol from PulseNet International (29). Briefly, cell suspensions combined with Proteinase K (Sigma-Aldrich) were embedded in molten 1% SeaKem Gold agarose (Cambrex Bio Science, Rockland, ME) to form plugs, which were later lysed in Cell Lysis Buffer and Proteinase K at 55°C for 5 h. Plugs were washed with ultrapure water and TE Buffer, and kept submerged in TE Buffer at 4°C for storage. Plug slices were restricted in a solution of Sure Cut H Buffer, Bovine Serum Albumin (New England Biolabs, Ipswich, MA), and XbaI restriction enzyme (New England Biolabs) at 37°C for 5 h. Restricted slices were then embedded in a 1.0% SeaKem Gold agarose gel, and electrophoresis was performed using a CHEF Mapper (Bio-Rad Laboratories, Inc., Hercules, CA) programmed with S. enterica and E. coli O157:H7-specific paramaters provided in the PulseNet International protocol. Gels were stained with ethidium bromide, visualized using an ultraviolet transilluminator, and captured as .tif files. For all PFGE runs, plugs containing digested DNA from S. enterica serovar Braenderup strain H9812 were included as a size standard (28). All gel images were formatted and normalized using the BioNumerics Version 6.6 Software (Applied Maths, Sint-Martens-Latem, Belgium), with bands sizes assigned based on the pattern of the S. Braenderup strain. Dendrograms were created for groups of isolates from each facility using the Dice Correlation Coefficient in conjunction with the Unpaired Groupwise Matching Algorithm set to 1.5% optimization and 1.5% tolerance. Traditional Kaufman-White-LeMinor Serotyping of S. enterica Isolates One S. enterica isolate representing each unique XbaI PFGE pattern (i.e. pulsotype) was further characterized by traditional serotyping according to the White- Kaufman-LeMinor scheme (20) at the USDA Animal and Plant Health Inspection Service National Veterinary Services Laboratories (Ames, IA). All isolates of identical PFGE patterns were presumed to be of the same serotype based on previous evidence

100

Texas Tech University, Alex Brandt, May 2014 describing the high degree of correlation between pulsotype and serotype (79). Thus, herein S. enterica subtypes will be referred to as a combined serotype-pulsotype. Online Storage of Sampling and Subtyping Data For all isolates of Listeria, S. enterica, and E. coli O157:H7, pertinent sample collection information, subtyping and serotyping information, and current holding and storage information was entered into the open-access Food Microbe Tracker database (www.foodmicrobetracker.com) (76). All isolate entries generated for the study were grouped under the project title “TTU NIFSI Persistence 2010.” Statistical Analyses for Subtype Persistence Molecular subtyping data for L. monocytogenes, non-pathogenic Listeria species, and S. enterica were subjected to two previously-described statistical tests to determine subtype persistence at the facility level (39). One is a binomial test for subtype frequency (herein referred to as the “binomial frequency test”) that compares the facility-level frequency of a subtype with the frequency of the subtype in a reference distribution. Because this test is based on conditional probability that a subtype is isolated on a subsequent instance after its introduction into a facility, subtype frequencies were calculated as the number of isolates of the subtype minus one divided by the total number of isolates of the species minus one. For L. monocytogenes, facility-level frequency of a particular EcoRI ribotype was compared to two reference distributions derived from Food Microbe Tracker: i) a subset of 5000 L. monocytogenes entries encompassing all sources (e.g. food, environmental, and clinical) (ALL LM; Supplemental File 6) and ii) a smaller subset of 1411 L. monocytogenes entries specifically from food environments (ENV LM; Supplemental File 7). Similarly, for non-pathogenic Listeria species, facility-level frequency of a particular sigB AT was compared to two Food Microbe Tracker reference distributions: i) a subset of 1928 non-pathogenic Listeria species entries from all sources (ALL LS; Supplemental File 8) and ii) a smaller subset of 773 non-pathogenic Listeria species entries specifically from food environments (ENV LS; Supplemental File 9). Because XbaI PFGE patterns are not consistently named in Food Microbe Tracker, comparisons based on the XbaI patterns of S. enterica and E. coli O157:H7 were precluded. However, frequency calculations for S. enterica were able to be based on the serotype of the combined serotype-pulsotype subtype. Thus, for S. enterica, two 101

Texas Tech University, Alex Brandt, May 2014 reference distributions were derived from Food Microbe Tracker: i) a reference distribution of 7273 S. enterica isolates from all sources (ALL SE; Supplemental File 10) and ii) a reference distribution of 672 S. enterica isolates from environmental sources (farms, food processing environments) (ENV SE; Supplemental File 11). All analyses for this statistical test were carried out with the binom.test function in R i386 version 2.15.1 (www.r-project.org). Bonferroni corrections were applied to α values based on the number of subtypes of a particular species obtained in a facility. For example, for Facility B, four EcoRI ribotypes of L. monocytogenes were observed, and the Bonferroni-corrected α was 0.05/4 or 1.25 X 10-2. When the P value corresponding to a certain subtype was less than α, the subtype was considered a persistent strain at the facility level due to its overrepresentation as compared to the reference distributions. This same test based on subtype frequency was also applied to analysis of site- level frequencies for species subtypes. All sampling sites in a facility that tested positive for a certain species on more than one instance were included in the analysis. Thus, Bonferroni corrections to α values were based on the number of sites with more than one positive result. For instance, Facility B had 12 sites with more than one positive instance of L. monocytogenes and the Bonferroni-corrected α was 0.05/12 or 4.17 X 10-3. Site- level frequencies were calculated in the same fashion as facility level frequencies, considering the first instance of isolation as the introduction of a subtype into the site. The same reference distributions described above (ALL LM, ENV LM, ALL LS, ENV LS, ALL SE, and ENV SE) were also used for comparison with the binom.test function of R. As above, when the P value for a subtype was less than α, the subtype was considered persistent at the site, and the site was defined as a site of peristence. The other statistical test used is a binomial test that considers a previous positive result for a specific subtype as a risk factor for a current positive result of the same subtype. Herein this test will be referred to as the “binomial risk factor test.” Subtype transitions were noted at each site on consecutive sampling dates whereas a site was either i) currently positive for a subtype after being previously positive for the subtype (++), ii) currently positive after being previously negative (-+), iii) currently negative after being previously positive (+-), or iv) currently negative after being previously negative (--). Transitions for a subtype within a facility were compiled into a 2 X 2

102

Texas Tech University, Alex Brandt, May 2014 contingency table and analyzed using Fisher’s Exact Test with the fisher.test function in R. Bonferrroni corrections to the α were based on the number of subtypes in the facility. Relative risks and the corresponding 95% confidence intervals were calculated using the MedCalc online calculator for Relative Risk (www.med calc.org/calc/relative _risk.php). As in the binomial frequency test, when the P value for a subtype was less than α, the subtype was considered a persistent subtype in the corresponding facility.

Results Characteristics of Facilities Enrolled Facility A is a USDA-inspected facility located in an urban area, and exclusively processes ready-to-eat products in the range of 400,000 lbs per year. Products produced primarily include sausages and hot dogs. The facility operates under Alternative 3 of the Listeria Control Rule (64), and practices separation of raw and RTE with different colored frocks and different areas of the facility devoted to each processing stage. Facility B is a state department of agriculture-inspected custom-exempt processing facility located in an urban area, which produces approximately 72,300 lbs of ready-to-eat product per year primarily in the form of ham, bacon, smoked brats, snack sticks, summer sausage and salami and beef jerky. Some of this RTE product work is done as custom curing work for four other facilities that contract with the operation, while other product is offered for sale in the facility under local health department inspection. The facility also has a harvesting component and custom slaughters cattle and hogs. Unlike Facility A, Facility B does not comply with a Listeria Control Rule Alternative, and does not require separation of equipment and clothing based on raw or RTE processing. Facility C is also a state department of agriculture-inspected custom-exempt facility located in a rural area, and though it irregularly peformed small amounts of custom pork and turkey smoking, the primary volume of production is in custom beef slaughter with approximately 750,000 lbs of carcass weight per year. Facility D is a USDA-inspected facility located in a suburban area, and has slaughter production that amounts to over 2,000,000 lbs of product per year. The facility also produces ready-to-eat product under Alternative 3 of the Listeria Control Rule. Though three of the four facilities have mixed operations, Facilities A and B were treated as primarily RTE meat production facilities, 103

Texas Tech University, Alex Brandt, May 2014 while Facilities C and D were classified as primarily fresh-slaughter and cutting facilities. Table 2.2 also provides a synoposis of pertinent facility characteristics such as number of shifts, facility age, and number of employees that were captured via manager questionnaires. Facility A Sampling, Subtyping, and Persistence Test Outcomes In Facility A, 11.21% of the 330 samples collected were positive for L. monocytogenes, and individual sampling date prevalence values ranged from 1.82 to 21.82% (Fig. 2.5). Additionally, 9.39% of the 330 Listeria samples contained non- pathogenic Listeria species, where individual sampling date prevalence values ranged from 5.45 to 14.55% (Fig. 2.5). No samples from Facility A collected and processed for S. enterica or E. coli O157:H7 produced a positive result. Employee surfaces and employee contact sites had the highest prevalence of L. monocytogenes among all categories, followed closely by drain and floor sites (Table 2.3). Non-food contact environmental sites and food contact sites and food samples both had lower prevalence values (Table 2.3). Non-pathogenic Listeria species followed a different trend whereby drain and floor sites had the highest prevalence, followed by employee surfaces and employee contact sites, then food contact sites and food samples, and finally non-food contact environmental sites (Table 2.3). Three L. monocytogenes EcoRI ribotypes 116-239-S-2, DUP-1052A, and DUP- 1048A, were observed among the isolates collected from Facility A, and DUP-1052A was most predominant. All 116-239-S-2 and DUP-1052A isolates belonged to serogroup II.2, while the single DUP-1048A ribotype isolate was categorized in the I.2 serogroup. Six sigB allelic types were observed among the non-pathogenic Listeria species isolates and consisted of L. innocua ATs 56, 37, 27, 19, and 32 and Listeria welshimeri AT 69. Most contamination for both L. monocytogenes and non-pathogenic Listeria species was limited to the raw processing areas of the facility (i.e. raw product storage, the grinding room, and the stuffing room). However, both L. monocytogenes and non- pathogenic Listeria species were isolated from the drains and gap between smokehouses in the cooking room, and from smokehouse truck rods in the packaging room. L. monocytogenes was also isolated from employee boots in the packaging room, while non- pathogenic Listeria species were additionally isolated from the floor in the cooking room 104

Texas Tech University, Alex Brandt, May 2014 and the drains of the packaging room, indicating spread beyond the raw processing area into areas of the facility where products are exposed post-lethality. Of the three ribotypes identified, only two, 116-239-S-2 and DUP-1052A, were persistent according to the binomial frequency test (Table 2.4), and only two non- pathogenic Listeria species strains sigB ATs 56 and 69 were persistent based on the test (Table 2.5). When applying the binomial frequency test to L. monocytogenes ribotype data on the individual site level, two sites emerged as sites of L. monocytogenes 116-239- S-2 persistence in Facility A: Sites 3 and 10 (Table 2.6). Site-level application of the test to non-pathogenic Listeria species sigB AT data yielded only one site of persistence, Site 25, where the L. welshimeri AT 69 strain was the persistent colonizer (Table 2.7). Both L. monocytogenes strains that were persistent according to the binomial frequency test (DUP-1052A and 116-239-S-2) were also considered persistent by the binomial risk factor test (Table 2.8). However, for non-pathogenic Listeria species, only AT 69 was deemed persistent according to the binomial risk factor test (Table 2.9). AT 56 along with all others identified were defined as non-persistent subtypes. Facility B Sampling, Subtyping, and Persistence Test Outcomes Results for Facility B indicated that 15.01% of 325 samples were positive for L. monocytogenes, and that individual sampling date prevalence values ranged from 7.41% to 25.45% (Fig. 2.5). Non-pathogenic Listeria species had an overall prevalence of 25.85% with a range of 14.81% to 36.36% for each individual sampling date (Fig. 2.5). Similar to Facility A, no E. coli O157:H7 was isolated, although S. enterica was isolated from 1.54% of the 326 samples. Drain and floor sites had the highest prevalence for L. monocytogenes, followed by employee surfaces and employee contact sites, then non- food contact sites, and finally food contact sites and food samples (Table 2.3). An identical trend in site category prevalence was also observed for non-pathogenic Listeria species (Table 2.3). All S. enterica originated from drain and floor sites (Table 2.3). Unlike Facility A, contamination was not confined to a certain area of the facility. The kill floor area, coolers, and processing room of Facility B all produced positive samples, likely due to the lack of barriers such as boot baths, floor foamers, and designated attire. L. monocytogenes isolates belonged to four EcoRI ribotypes: DUP-1062E, DUP- 1062B, DUP-1053E, and DUP-1042C. All isolates belonging to DUP-1602E, DUP- 105

Texas Tech University, Alex Brandt, May 2014

1062B, and DUP-1053E were assigned to serogroup I.1 while DUP-1042C was placed into serogroup II.2. Twelve unique ATs were observed among non-pathogenic Listeria species isolates (Table 2.5). XbaI PFGE revealed a total of three S. enterica pulsotypes differing by more than three bands each; two PFGE patterns differed by only one band, but were still grouped together (Fig. 2.6). Serotypes identified by Kaufman-White- LeMinor serotyping for the three pulsotypes were Liverpool, Chailey, and Heidelberg. According to the binomial frequency test, three out of the four L. monocytogenes ribotypes identified (DUP-1062E, DUP-1062B, and DUP-1053E) were persistent (Table 2.4). Applying this test to non-pathogenic Listeria species identified four of the twelve sigB ATs as persistent subtypes (AT 11, AT 6, AT 23, and AT 38) (Table 2.5). One of the three combined S. enterica serotype-pulsotype combinations, corresponding to serotype Heidelberg, was persistent (Table 2.10). When applying the binomial frequency test to individual sites in Facility B, only one site was classified as a site of persistence for an L. monocytogenes strain, Site 44, which was persistently colonized by ribotype DUP-1062E (Table 2.6). Two sites were sites of persistence for specific non-pathogenic Listeria species subtypes: Sites 10 and 12 (Table 2.7); Site 10 was persistently colonized with AT 6, while Site 12 was persistently colonized with AT 23. No sites produced S. enterica on more than one occasion, and thus were not available for analysis via this test. The binomial risk factor test defined only one of the four EcoRI ribotypes (DUP- 1062E) as a persistent subtype, while the other three, including two strains that were persistent based on the binomial frequency test (DUP-1062B and DUP-1053E), were non-persistent (Table 2.8). Likewise, the test only defined AT 23 as a persistent non- pathogenic Listeria species AT out of all twelve ATs identified (Table 2.9). No S. enterica subtypes were considered persistent according to this test (Table 2.11). Facility C Sampling and Subtyping Results L. monocytogenes prevalence in Facility C over the six-month study period was 16.97% among the 330 Listeria samples, with individual sampling date prevalences varying between 5.45 and 27.27% (Fig. 2.7). Non-pathogenic Listeria species prevalence was 28.18%, varying from 16.36 to 38.18% on individual sampling dates (Fig. 2.7). S. enterica prevalence in Facility C was the highest among all four facilities at 6.97%, and 106

Texas Tech University, Alex Brandt, May 2014 was as high as 18.18% on one sampling occasion (Fig. 2.7). Though E. coli O157:H7 was isolated on two different sampling occasions, the overall prevalence remained low at 1.21% (Fig. 2.7). Drain and floor sites demonstrated the highest prevalence for L. monocytogenes, followed by the other categories in a fashion identical to that observed in Facility B (Table 2.3). This same trend was also observed for both non-pathogenic Listeria species and S. enterica (Table 2.3). The four sites where E. coli O157:H7 was isolated on the two different sampling dates included two non-food contact environmental sites and two drain and floor sites. Like Facility B, Listeria contamination was not predominantly associated with a particular area of the facility. However, no S. enterica or E. coli O157:H7 contamination was ever detected in the processing room. EcoRI ribotyping revealed the presence of two ribotypes, DUP-1042B and DUP- 1057B. DUP-1042B comprised all Facility C L. monocytogenes isolates except for two, and all DUP-1042B isolates were categorized as serogroup II.2, while DUP-1057B isolates were of serogroup I.1. sigB allelic typing revealed 13 ATs among the non- pathogenic Listeria isolates from Facility C, including a previously unobserved AT of L. welshimeri, herein referred to as LW AT NEW A. The L. innocua AT 141 colonies had a light blue appearance on LMPM plates, and a positive result for the hlyA confirmation PCR, leading them to be initially classified as L. monocytogenes. However, only the prs gene PCR product was detected via molecular serogrouping PCR, indicating that the isolates belonged to a species other than L. monocytogenes. Upon further investigation it was noted that AT 141 belongs to a group of rare hemolytic L. innocua. This strain has also been isolated from soil associated with a poultry production environment (43). XbaI PFGE for S. enterica revealed six pulsotypes that differed from one another by more than three bands (Fig. 2.8). One isolate of each pulsotype was subjected to Kaufman-White-LeMinor serotyping, and the serotypes identified included Muenchen, Infantis, Rough O:z29:-, Uganda, Anatum, and the 50:r:z serotype of the rare S. enterica subsp. arizonae (subsp. III). E. coli O157:H7 isolates from Sampling 2 shared an identical PFGE pattern, while isolates of Sampling 5 shared a different pattern (Fig 2.9). Isolates from Sampling 5 harbored all typical E. coli O157:H7 virulence genes (eaeA, stx1, and stx2), while those obtained during Sampling 2 notably lacked the stx1 gene.

107

Texas Tech University, Alex Brandt, May 2014

Facility-level analysis for L. monocytogenes persistence via the binomial frequency test indicated that only the DUP-1042B ribotype strain was persistent (Table 2.4). The strain was also persistent in six different sites (Table 2.6): Sites 15, 19, 26, 27, 39, and Site 48. Analysis of facility-level sigB allelic typing data via the binomial frequency test indicated that both L. innocua AT 70 and AT 71 strains were persistent (Table 2.5). Three sites of persistence, all persistently colonized by AT 70, were identified: Sites 18, 19, and 26 (Table 2.7). However, the frequency of AT 70 at these sites was only significantly greater than its frequency in the ALL LS distribution. Using the binomial frequency test to compare serotype frequencies for S. enterica, only three of the six serotype-pulsotype combinations (those belonging to serotypes Infantis, Uganda, and the S. enterica subsp. arizonae 50:r:z serotype) were persistent at the facility level (Table 2.10). However, of these three, only the Infantis serotype- pulsotype was found on multiple sampling occasions. Thus, the overrepresentation (and persistent result) of serotypes Uganda and III 50:r:z seemed to be due to their spread to multiple sites on single sampling dates rather than to long-term persistence. In agreement with the binomial frequency test, the binomial risk factor test demonstrated that L. monocytogenes DUP-1042B strain was persistent (Table 2.8). AT 70 was considered a persistent sigB AT based on this test (similar to the outcome for the binomial frequency test), but AT 71 was not (Table 2.9). Rather, AT 6 was included as a persistent subtype alongside AT 70. The S. enterica Infantis serotype-pulsotype did produce one ++ transition giving it a Relative Risk of 14.94, but did not produce a significant result to identify it as a persistent subtype (Table 2.11). Facility D Sampling and Subtyping Results Facility D had the lowest contamination level for L. monocytogenes with an overall prevalence of 1.21% and individual sampling date prevalence values in the range of 0.00% to 3.64% (Fig. 2.7). All L. monocytogenes isolates originated from drain and floor sites (Table 2.3). Similarly, non-pathogenic Listeria species contamination levels were low at 6.07% with individual sampling date prevalences ranging from 3.64% to 9.09% (Fig. 2.7). Site category prevalence for non-pathogenic Listeria species was identical to that in Facilities B and C (Table 2.3). E. coli O157:H7 was not isolated, but S. enterica was isolated from eight sites on Sampling 3 (individual date prevalence of 108

Texas Tech University, Alex Brandt, May 2014

14.55%), giving it a total prevalence of 2.42% (Fig. 2.7). Only drain and floor sites and non-food contact environmental surfaces were contaminated with S. enterica (Table 2.3). L. monocytogenes ribotypes observed included DUP-1030A and DUP-1030B, and isolates of both ribotypes were placed in the I.1 serogroup. Non-pathogenic Listeria species sigB allelic types identified included L. innocua ATs 6, 31, 109, 124, and 153, L. welshimeri AT 129, Listeria seeligeri AT 121, and a previously unobserved L. welshimeri AT, herein referred to as LW AT NEW B. Most Listeria-positive samples were obtained from either the kill floor area or the carcass cooler, and only one L. monocytogenes isolate was obtained from the processing room at Site 18. PFGE typing demonstrated that all S. enterica isolates from the eight positive sites on Sampling 3 belonged to the same XbaI pulsotype (Fig. 2.10). A representative isolate of the pulsotype was Kaufman-White-LeMinor serotyped as Typhimurium Var 5-. Interestingly, one could see a logical progression of contamination from one site to the next based on the process flow. Contamination presumably orginated near Site 36 and passed to Sites 37, 38, 39, 41, 42, and 44. One positive sample was also found in the processing room at Site 18. This latter result was discussed with the manager who mentioned that the drain (Site 18) was connected to the drainage system from the kill floor, and likely could have backed up, causing the S. enterica to be isolated from it. Analysis of L. monocytogenes subtype data via the binomial frequency test indicated that ribotype the DUP-1030B strain was persistent (Table 2.4). No site produced a positive result for L. monocytogenes on more than one occasion, and thus no sites of persistence for an L. monocytogenes ribotype were identified. According to the binomial frequency test, non-pathogenic Listeria species strains of ATs 6 and 129 were persistent (Table 2.5). Site 25 was persistently colonized by AT 129 based on the site- level outcomes of the binomial frequency test (Table 2.9). Since the S. enterica Typhimurium Var 5- serotype-pulsotype was the only one identified in Facility D, the serotype-pulsotype was by default persistent according to the binomial frequency test (Table 2.10). However, similar to the situation for III 50:r:z and Uganda in Facility C, this overrepresentation likely resulted from its spread to multiple sites in the same facility on the same sampling date. This single-date isolation of S. enterica also precluded site- level analysis of persistence via the binomial frequency test.

109

Texas Tech University, Alex Brandt, May 2014

The binomial risk factor test did not identify either Facility D L. monocytogenes ribotype as persistent (Table 2.8). Likewise, the lone S. enterica serotype-pulsotype was not considered persistent (Table 2.11). However, the L. welshimeri AT 129 strain found repeatedly from Site 25 was considered persistent by the binomial risk factor test (P = 2.38 X 10-3) and had a Relative Risk of 33.50 (Table 2.9).

Discussion S. enterica and E. coli O157:H7 Exhibited Lower Prevalence Among Processing Facility Environmental Sites Compared to Listeria, and a Substantial Range in Listeria Prevalence Was Observed Across the Four Facilities To our knowledge this study is the first to simultaneously test for Listeria, S. enterica, and E. coli O157:H7 among environmental sites of RTE and fresh meat processing facilities. Several studies have investigated the presence of L. monocytogenes and non-pathogenic Listeria species in a wide array of facilities including RTE seafood facilities (e.g.(50, 56)), dairies and cheese processing environments (e.g. (30, 36)), chilled food processing environments (e.g.(25, 42)), and meat and poultry production facilities (e.g.(46, 52, 78)). However, food environmental sampling for S. enterica has been infrequently performed, and studies on food environment surveillance for E. coli O157:H7 have been limited to Turkish kasar cheese facilities (6) and beef abbattoirs (21). Facility C had the highest S. enterica prevalence at 6.97%, but this was lower than the facility’s prevalence of both L. monocytogenes (16.97%) and non-pathogenic Listeria species (28.18%) (Fig. 2.7). This same trend was also observed for Facility B where S. enterica prevalence was 13.47% and 24.31% lower than that of L. monocytogenes and non-pathogenic Listeria species, respectively. However, in Facility D the S. enterica prevalence was 1.21% greater than the L. monocytogenes prevalence, although it was 3.65% lower than the prevalence of non-pathogenic Listeria species. The isolation of E. coli O157:H7 on only two occasions in Facility C (Fig. 2.7) demonstrates that its contamination of sites in the facility environment is likely infrequent. Gun et al. similiarly found a low prevalence of E. coli O157:H7 in the environments of four Turkish cattle abbattoirs with isolates obtained from floors, aprons, knives, and worker hands

110

Texas Tech University, Alex Brandt, May 2014

(21). Likewise, Cagri-Mehmetoglu et al. noted an E. coli O157:H7 prevalence of only 2.7% among environmental samples taken from two kasar cheese processing plants (6). The substantial range in L. monocytogenes prevalence from 1.21% in Facility D to 16.97% in Facility C was similar to that observed in a previous study of RTE meat facilities where values ranged from 1.70% to 10.80% (78). However, higher prevalences for L. monocytognes have been observed in other studies where meat production facilities have been sampled. Bērziᶇš et al. demonstrated a prevalence of 23.4% for L. monocytogenes among enviromental sites at a cold-smoked pork processing plant (3), and Chasseignaux et al. demonstrated prevalence of 23.7% in poultry and pork meat processing plants (8). Thévenot et al. sampled 13 dried sausage processing plants and found a mid-shift prevalence of 47.3% (59). Similar to the aforementioned study (78), non-pathogenic Listeria species prevalence also varied substantially between the four facilities included in this study, with prevalences ranging from 6.07 to 28.18%. Drain and Floor Sites Had Higher Levels of Listeria and S. enterica Contamination as Compared to Other Meat Processing Plant Sites Site category prevalence across all facilities showed that drains and floors were the most highly contaminated sites for L. monocytogenes, non-pathogenic Listeria species, and S. enterica (Table 2.3). E. coli O157:H7 was an exception with a prevalence evenly split between drain and floor sites and non-food contact environmental sites. This contamination trend for certain site categories was similar to that observed for L. monocytogenes in a study by Hoffman et al. where drain sites among all areas (raw product, in-process, and finished product) of one smoked fish processing plant had a prevalence of 62.50%, compared with 32.30% for other environmental sites, and 3.10% for samples collected from food contact surfaces (24). A similar trend for L. monocytogenes in smoked fish processing plants was recorded by Thimothe et al. who showed that 23.7% of drain samples (from raw product, in-process, and finished product areas) were positive for the pathogen, followed by 12.3% of non-food contact environmental surfaces, 10.4% of employee contact surfaces, and 4.8% of food contact surfaces (60). Likewise, Kabuki et al. also demonstrated that drain and floor samples were second and third in order of L. monocytogenes prevalence among sites in a fresh cheese processing facility (30). In addition, the majority of the sites of persistence 111

Texas Tech University, Alex Brandt, May 2014 defined by the binomial frequency test were drain and floor sites (8 of the 14 unique sites), and most other sites of persistence (4 of 6) were still associated in some way with floors (i.e. boots, squeegees). These observations, along with the fact that all managers indicated that drain and floor sites were cleaned and sanitized daily, seems to suggest that sanitation procedures for drains and floors were either: i) ineffective at eliminating contamination, ii) not being performed at the same frequency in all areas of the facility (i.e. processing room drains as opposed to cooler drains), or iii) effective but were being negated by inadequate floor- focused control measures (i.e. boot baths or floor foamers). Indeed, in Facility A, which used floor foamers and boot baths, the prevalence for L. monocytogenes in drain and floor sites was actually the second highest categorically. While persistent strains may have certain characteristics that allow them to withstand routine control measures within drain and floor sites, the high degree of contamination and persistence associated with these sites further re-emphasizes their problematic nature. Thus, when suggesting interventions and changes for facilities to adopt, a heavy focus should be provided on proper management of drain and floor contamination. Molecular Subtyping Revealed Unique Molecular Ecologies for All Facilities, Including Rare and Novel Subtypes, and Subtypes Associated with Illness Molecular subtyping revealed that the bacterial population of interest in each procesing facility was comprised of a unique set of resident microflora, and that subtype richness and evenness varied. The diversity in L. monocytogenes ecology was best observed by comparing the composition of Facility C, which was heavily predominated by the persistent L. monocytogenes DUP-1042B strain, with that of Facility B, which was colonized by three less predominant, yet persistent, ribotypes (DUP-1062B, DUP-1062E, and DUP-1053E). The same was true for S. enterica where only one subtype was identified in Facility D, while six serotype-pulsotypes were observed Facility C. Of all groups, the non-pathogenic Listeria species population was the richest in composition with the largest amount of subtypes observed in each facility. Interestingly, no EcoRI ribotypes or XbaI pulsotypes were found across facilities. However, strains of non- pathogenic Listeria species were less exclusive, and identical ATs were often found

112

Texas Tech University, Alex Brandt, May 2014 across multiple establishments. For example, L. innocua ATs 6 and 109 were found across Facilities B, C, and D, and Facilities B and C shared eight ATs between them. The isolation of the rare subspecies arizonae (III) 50:r:z strain in Facility C was interesting due to the fact that the subspecies is usually restricted to gut of reptiles (38). However, a strain of the same serotype was previously isolated in irrigation ditch water of an agricultural environment in previous study performed by our group (40). Hemolytic L. innocua strains, although rare, have been previously isolated from processing facilities (44, 78). Yet, this study is the first known instance where an AT 141 strain of hemolytic L. innocua was found in a processing facility. Also, the identification of novel L. welshimeri ATs in both Facility C and Facility D was interesting as these novel ATs may represent previously uncharacterized strains with unique characteristics. The persistence of the L. monocytogenes DUP-1042B strain in Facility C was concerning since the DUP-1042B ribotype is also shared with L. monocytogenes Epidemic Clone (EC) IV. EC IV is a strain of a particular multi-virulence locus sequence type that has been associated with several foodborne illness outbreaks (2, 31, 45). However, EC IV strains that have caused disease in the past have belonged to the II.1 serogroup, while all isolates collected in this study belonged to the II.2 serogroup. Serotype 4b isolates (part of serogroup II.1) are most commonly associated with human illness, and based on their II.2 serogrouping, all DUP-1042B isolates in this study belonged to either serotype 1/2b, 3b or 7. Also, among the ten S. enterica serotypes identified in this study, three (Muenchen, Infantis, and Heidelberg) were among the ten serotypes most commonly associated with human salmonellosis in the U.S. in 2011 (73). Serotype Heidelberg was also recently implicated in a large salmonellosis outbreak associated with chicken processed in California (74). The Binomial Frequency Test Identified Persistent Subtypes of Foodborne Pathogens at the Facility and Individual Site Levels The binomial test for ribotype frequency developed by Malley et al. identified persistent strains and sites of persistence in all facilities (39). The outcomes of Facility A were especially interesting since they were able to be compared with the results of a previous study by Williams et al. who also enrolled the facility (78). The persistence of the L. monocytogenes DUP-1052A strain in this study corresponded well with the 113

Texas Tech University, Alex Brandt, May 2014 expansive time intervals over which it has been isolated from the facility. Williams et al. first noted the presence of the strain in Facilty A in May 2007. This suggests that the strain persisted in the environment of the facility for at least four years up until the start of this study in May 2011. Unpublished data from the second year of sampling conducted by Williams et al. (77) also demonstrated that the ribotype 116-239-S-2 strain, which also persisted in this study, was also isolated from the facility in March 2008. This seems to suggest that the 116-239-S-2 strain also persisted in the facility for up to three years until it was isolated during this study in May 2011. Furthermore, the persistent L. innocua AT 56 and L. welshimeri AT 69 strains were also found in Facility A in May 2007, demonstrating over four years of persistence for these non-pathogenic listeriae (77). Though no prior sampling data is available for the other three facilities, other studies have provided evidence of persistence of strains belonging to their resident subtypes in other food processing operations. For instance, ribotype DUP-1062B, the ribotype of one persistent Facility B L. monocytogenes strain, was the ribotype of another L. monocytogenes strain that was highly predominant and persistent in a Norwegian fish processing facility (33). A strain of ribotype DUP-1053E, which was the subtype of another persistent Facility B strain, also persisted in a raw fish slaughter and processing plant and an adjacent, downstream, salmon smoking operation (34). Similar to this study, a ribotype DUP-1042B strain was also persistent and widespread in a farmstead cheese processing plant (10). An L. innocua AT 70 strain also persistently colonized the raw/in- process area of a RTE meat production plant sampled by Williams et al. (77). A continuity of subtype persistence at specific sites in Facility A was also observed by comparing the results of this study with that of Williams et al. Both sites that were persistently colonized with L. monocytogenes 116-239-S-2 during this study (Table 2.6) were also positive for the strain during the second year of sampling by Williams et al. Further analysis of the physical characteristics of these two sites and the 116-239-S-2 isolates obtained from the sites would be interesting from the standpoint of whether longitudinal changes in cleaning and santitation or physical structure of the site are reflected by adaptations within the strain. Regardless, these results seem to indicate the existence of a relationship between the strain and the specific conditions present at these two sites, which has allowed it to persist in them for over three years.

114

Texas Tech University, Alex Brandt, May 2014

Identification of two sites in Facility C as sites of co-persistence for L. monocytogenes DUP-1042B and L. innocua AT 70 strains (Sites 19 and 26), also seemed to indicate that the sites possessed characteristics that contributed to the persistence of both subtypes. Investigation of the characteristics shared between the DUP-1042B strain and AT 70 strain would also be interesting as both strains were co-isolated from the same Facility C samples on a large number of occasions (23 of 330 samples). Interestingly, several sites that were contaminated with either L. monocytogenes or non-pathogenic Listeria species on all six sampling occasions (e.g. Facility B Sites 17, 35, 39, and 42) were not classified as sites of persistence by the binomial frequency test. This was due to the fact that a specific strain was not isolated with enough frequency at any of these sites such that it was overrepresented and deemed persistent. One plausible explanation for this is that the sites may have been locations where contaminating strains were frequently deposited during the course of operations, but did not establish persistence. Thus, although some Listeria may have always been present at the sites, various strains may have been randomly reintroduced on a regular basis, and the composition of strains found at the sites may have changed often. This explanation seems logical given that the sites were drain and floor sites located near to high traffic areas, and it seems to account for why such a diversity of strains were isolated from the sites. An alternative explanation is that these sites may actually have been locations of of persistence for multiple strains rather than just one, and because only one isolate from each sampling occasion was chosen for molecluar subtyping, this co-persistence may not have been revealed. For the facilities with a diverse set of strains present among the resident population, this very easily could have been possible. Indeed, of all sites of persistence for a specific Listeria strain identified by the binomial frequency test among all facilities, 11 sites contained other Listeria strains in addition to the one that persisted at the site. Thus, it is possible that by subtyping a larger number of isolates from each sample, additional sites where more than one strain had established persistence could have been revealed. This is a key limitation of the methodology, and in future studies this limitation should be addressed by subtyping greater numbers of isolates from each sample (especially those from sites that are frequently contaminated) in order to ensure that all strains persisting at the site are revealed and accounted for in the calculation.

115

Texas Tech University, Alex Brandt, May 2014

The Binomial Risk Factor Test Identified Persistent Subtypes of Foodborne Pathogens at the Facility Level Relative to the binomial frequency test, the binomial risk factor test identified fewer subtypes as persistent. Only four and three of the seven and ten persistent L. monocytogenes strains and non-pathogenic Listeria species strains identified by the binomial frequency test, respectively, were also persistent according to the binomial risk factor test. A fourth non-pathogenic Listeria strain, an AT 6 strain that was consecutively isolated three times at Facility C Site 5, was not deemed persistent according to the binomial frequency test, though the binomial risk factor test denoted that it was persistent. This observation demonstrates that a strain need not be prevalent within a facility to be defined as persistent via the binomial risk factor test – in fact, AT 6 comprised only 5.37% of Facility C non-pathogenic Listeria species isolates. No S. enterica strains were persistent according to the binomial risk factor test, mainly owing to the fact that, with the exception of serotype-pulsotype Infantis in Facility C, none were isolated from the same location on more than one instance. For example, the S. enterica Typhimurium Var 5- serotype-pulsotype was persistent in Facility D according to the binomial frequency test, but because it was only observed on one occasion and never more than once in the same site, it was not persistent according to the binomial risk factor test. Thus, the binomial risk factor test seemed to be more robust with regard to the influence that dissemination to multiple sources on a single sampling day could have in the determination of a strain’s persistence. Instances such as these highlight how caution may need to be exercised in interpretation of the outcomes of the binomial frequency test. Though the test takes into account the historical prevalence of a subtype and thus aims to differentiate whether the subtype is actually isolated more frequently in a facility because of its persistence, and not just due to chance alone, the test does not account for spread to multiple sites on a single sampling date. While the binomial risk factor test does take this into account, it relies on the selection of sites where a strain persists. If these sites are not explicitly included in the sampling scheme, subtypes which may still be persisting in other areas of the facility, may not be counted among the persistent strains. In these situations, the binomial frequency test would still likely identify the persistent strain due to their 116

Texas Tech University, Alex Brandt, May 2014 significant overrepresentation in the facility. Thus, a combined definition of persistence, which takes into account the outcomes of both tests, may provide an even more succinct definition of a truly persistent strain, and warrants future consideration in development.

Conclusions The integrated approach of longitudinal sampling, molecular subtyping and statistical analysis employed in this study identified unique molecular ecologies of each processing facility with regard to the organisms of interest, and aided identification of subtype persistence on the facility and individual site levels. Furthermore, the longitudinal sampling results re-emphasized that food processing facility drains and floors continue to be problematic areas with regard to pathogen contamination, similar to that observed in previous studies. The data also demonstrate that S. enterica and E. coli O157:H7 have lower prevalence among environmental sites in RTE and fresh meat processing facilities as compared to L. monocytogenes and non-pathogenic Listeria species. Also, S. enterica strains, even when repeatedly isolated, exhibit much lower degrees of persistence at the facility and individual site levels than that observed for Listeria. The outcomes of the binomial frequency test for persistence corresponded well with the results of four year spans of environmental sampling in Facility A, but caution should be exercised when interpreting binomial frequency test calculations that include instances where a single subtype is distributed to multiple sites in the processing facility on a single date. In contrast, the binomial risk factor test appeared to be more robust with regard to these single-date strain overreprestation scenarios. However, unless sites where strains are recurrently isolated are included in the sampling scheme, the binomial risk factor test may fail to identify certain subtypes as persistent. Therefore, a combined definition of persistence based the outcomes of both tests may be worthwhile to develop as efforts continue to create a standardized definition of pathogen persistence. Also, consideration should be given to including sites with characteristics other than those commomly associated with Listeria (i.e. cool, damp sites) in future studies. This will help ensure that S. enterica and E. coli O157:H7 are not persisting in sites with other characteristics (i.e. those that are dry and dusty). Furthermore, several areas of facilities where pathogens could have possibly persisted were not included in the sampling 117

Texas Tech University, Alex Brandt, May 2014 schemes (i.e. the spice room and dry storage areas of Facility A, the dry storage room and food preparation areas of Facility B, and the laundry room of Facility D). Thus, more careful consideration to initial sample site selection should be given in future studies so as to include these potential outer-lying areas where strains may also persist. Lastly, in order to elucidate strain and site-specific risk factors for persistence, future efforts should focus on studying characteristics shared between persistent strains that are concurrently isolated from the same samples on multiple occasions, and should study the interactions between the conditions of certain sites of persistence and the strains that persist there.

2. Aureli P, Fiorucci GC, Caroli D, Marchiaro G, Novara O, Leone L, Salmaso S. 2000. An outbreak of febrile gastroenteritis associated with corn contaminated by Listeria monocytogenes. N. Eng. J. Med. 342:1236-1241.

3. Bērziᶇš A, Hellström S, Siliᶇš I, Korkeala H. 2010. Contamination patterns of Listeria monocytogenes in cold-smoked pork processing. J. Food Prot. 73:2103- 2109.

4. Borucki MK, Peppin JD, White D, Loge F, Call DR. 2003. Variation in biofilm formation among strains of Listeria monocytogenes. Appl. Environ. Microbiol. 69:7336-7342.

5. Bruce J. 1996. Automated system rapidly identifies and characterizes microorganisms in food. Food Technol. 50:77-81.

6. Cagri-Mehmetoglu A, Yaldirak G, Bodur T, Simsek M, Bozkir H, Eren NM. 2011. Incidence of Listeria monocytogenes and Escherichia coli O157:H7 in two kasar cheese processing environments. Food Contr. 22:762-766.

7. Carpentier B, Cerf O. 2011. Review - persistence of Listeria monocytogenes in food industry equipment and premises. Int. J. Food Microbiol. 145:1-8.

8. Chasseignaux E, Gérault P, Toquin M-T, Salvat G, Colin P, Ermel G. 2002. Ecology of Listeria monocytogenes in the environment of raw poultry meat and raw pork meat processing plants. FEMS Microbiol. Lett. 210:271-275.

118

Texas Tech University, Alex Brandt, May 2014

9. Cruz CD, Fletcher GC. 2011. Prevalence and biofilm-forming ability of Listeria monocytogenes in New Zealand mussel (Perna canaliculus) processing plants. Food Microbiol. 28:1387-1393.

10. D'Amico DJ, Donnelly CW. 2008. Enhanced detection of Listeria spp. in farmstead cheese processing environments through dual primary enrichment, PCR, and molecular subtyping. J. Food Prot. 71:2239-2248.

11. Dauphin G, Ragimbeau C, Malle P. 2001. Use of PFGE typing for tracing contamination with Listeria monocytogenes in three cold-smoked salmon processing plants. Int. J. Food Microbiol. 64:51-61.

12. Djordjevic D, Wiedmann M, McLandsborough LA. 2002. Microtiter plate assay for assessment of Listeria monocytogenes biofilm formation. Appl. Environ. Microbiol. 68:2950-2958.

13. Doumith M, Buchrieser C, Glaser P, Jacquet C, Martin P. 2004. Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. J. Clin. Microbiol. 42:3819-3822.

14. Earnshaw AM, Lawrence LM. 1998. Sensitivity to commercial disinfectants, and the occurrence of plasmids within various Listeria monocytogenes genotypes isolated from poultry products and the poultry processing environment. J. Appl. Microbiol. 84:642-648.

15. Ferreira V, Wiedmann M, Teixeira P, Stasiewicz MJ. 2014. Listeria monocytogenes persistence in food-associated environments: epidemiology, strain characteristics, and implications for public health. J. Food Prot. 77:150-170.

16. Fox EM, Leonard N, Jordan K. 2011. Physiological and transcriptional characterization of persistent and nonpersistent Listeria monocytogenes isolates. Appl. Environ. Microbiol. 77:6559-6569.

17. U.S. Department of Agriculture Food Safety and Inspection Service. 2014. Data collection and reports: microbiology. Available At: http://www. fsis.usda.gov /wps /portal /fsis /topics /data-collection-and-reports /microbiology. Accessed On: 2/18/14

18. Gannon VPJ, Dsouza S, Graham T, King RK, Rahn K, Read S. 1997. Use of the flagellar H7 gene as a target in multiplex PCR assays and improved specificity in identification of enterohemorrhagic Escherichia coli strains. J. Clin. Microbiol. 35:656-662.

119

Texas Tech University, Alex Brandt, May 2014

19. Gale Group. 2006. Bar-S Foods company: international directory of company histories. Available At: http://www.answers.com/topic/bar-s-foods-co. Accessed On: 2/12/14

20. Guibourdenche M, Roggentin P, Mikoleit M, Fields PI, Bockemuhl J, Grimont PA, Weill FX. 2010. Supplement 2003-2007 (no. 47) to the White- Kauffmann-Le Minor scheme. Res. Microbiol. 161:26-29.

21. Gun H, Yilmaz A, Turker S, Tanlasi A, Yilmaz H. 2003. Contamination of bovine carcasses and abattoir environment by Escherichia coli O157:H7 in Istanbul. Int. J. Food Microbiol. 84:339-344.

22. Harvey J, Keenan KP, Gilmour A. 2007. Assessing biofilm formation by Listeria monocytogenes strains. Food Microbiol. 24:380-392.

23. Heir E, Lindstedt B-A, Røtterud O-J, Vardund T, Kapperud G, Nesbakken T. 2004. Molecular epidemiology and disinfectant susceptibility of Listeria monocytogenes from meat processing plants and human infections. Int. J. Food Microbiol. 96:85-96.

24. Hoffman AD, Gall KL, Norton DM, Wiedmann M. 2003. Listeria monocytogenes contamination patterns for the smoked fish processing environment and for raw fish. J. Food Prot. 66:52-60.

25. Holah JT, Bird J, Hall KE. 2004. The microbial ecology of high-risk, chilled food factories; evidence for persistent Listeria spp. and Escherichia coli strains. J. Appl. Microbiol. 97:68-77.

26. Holah JT, Taylor JH, Dawson DJ, Hall KE. 2002. Biocide use in the food industry and the disinfectant resistance of persistent strains of Listeria monocytogenes and Escherichia coli. J. Appl. Microbiol. 92S:111S-120S.

27. Hu Y, Zhang Q, Meitzler JC. 1999. Rapid and sensitive detection of Escherichia coli O157:H7 in bovine faeces by a multiplex PCR. J. Appl. Microbiol. 87:867-876.

28. Hunter SB, Vauterin P, Lambert-Fair MA, Van Duyne MS, Kubota K, Graves L, Wrigley D, Barrett T, Ribot E. 2005. Establishment of a universal size standard strain for use with the PulseNet standardized pulsed-field gel electrophoresis protocols: converting the national databases to the new size standard. J. Clin. Microbiol. 43:1045-1050.

120

Texas Tech University, Alex Brandt, May 2014

29. PulseNet International. 2013. Standard operating procedure for PulseNet PFGE of Escherichia coli O157:H7, Escherichia coli Non-O157 (STEC), Salmonella serotypes, Shigella sonnei, and Shigella flexneri. Available At: http://www.pulse netinternational.org/protocols/. Accessed On: 3/1/14

30. Kabuki DY, Kuaye AY, Wiedmann M, Boor KJ. 2004. Molecular subtyping and tracking of Listeria monocytogenes in latin-style fresh-cheese processing plants. J. Dairy Sci. 87:2803-2812.

31. Kathariou S. 2002. Listeria monocytogenes virulence and pathogenicity, a food safety perspective. J. Food Prot. 65:1811-1829.

32. Kim JS, Lee GG, Park JS, Jung YH, Kwak HS, Kim SB, Nam YS, Kwon ST. 2007. A novel multiplex PCR assay for rapid and simultaneous detection of five pathogenic bacteria: Escherichia coli O157:H7, Salmonella, Staphylococcus aureus, Listeria monocytogenes, and Vibrio parahaemolyticus. J. Food Prot. 70:1656-1662.

33. Klæboe H, Lunestad BT, Borlaug K, Paulauskas A, Rosef O. 2010. Persistence and diversity of Listeria monocytogenes isolates in Norwegian processing plants. Vet. Med. Zoot. 50:42-47.

34. Klæboe H, Rosef O, Fortes E, Wiedmann M. 2006. Ribotype diversity of Listeria monocytogenes isolates from two salmon processing plants in Norway. Int. J. Environ. Health Res. 16:375-383.

35. Koohmaraie M, Arthur TM, Bosilevac JM, Guerini M, Shackelford SD, Wheeler TL. 2005. Post-harvest interventions to reduce/eliminate pathogens in beef. Meat Sci. 71:79-91.

36. Latorre AA, Van Kessel JA, Karns JS, Zurakowski MJ, Pradhan AK, Boor KJ, Adolph E, Sukhnanand S, Schukken YH. 2011. Increased in vitro adherence and on-farm persistence of predominant and persistent Listeria monocytogenes strains in the milking system. Appl. Environ. Microbiol. 77:3676- 3684.

37. Lundén J, Autio T, Markkula A, Hellström S, Korkeala H. 2003. Adaptive and cross-adaptive responses of persistent and non-persistent Listeria monocytogenes strains to disinfectants. Int. J. Food Microbiol. 82:265-272.

38. Mahajan RK, Khan SA, Chandel DS, Kumar N, Hans C, Chaudhry R. 2003. Fatal case of Salmonella enterica subsp. arizonae gastroenteritis in an infant with microcephaly. J. Clin. Microbiol. 41:5830-5832.

121

Texas Tech University, Alex Brandt, May 2014

39. Malley TJ, Stasiewicz MJ, Grohn YT, Roof S, Warchocki S, Nightingale K, Wiedmann M. 2013. Implementation of statistical tools to support identification and management of persistent Listeria monocytogenes contamination in smoked fish processing plants. J. Food Prot. 76:796-811.

40. Manuel CS. 2011. Molecular ecology of foodborne pathogens in primary production agriculture environments in Northern Colorado and investigations in Internalin A diversity in Listeria monocytogenes DUP-1039C strains. M.S. Thesis: Department of Animal Sciences, Colorado State University, 189 p.

41. Meng J, Zhao S, Doyle MP, Mitchell SE, Kresovich S. 1997. A multiplex PCR for identifying shiga-like toxin-producing Escherichia coli O157:H7. Lett. Appl. Microbiol. 24:172-176.

42. Miettinen M. 1999. Characterization of Listeria monocytogenes from an ice cream plant by serotyping and pulsed-field gel electrophoresis. Int. J. Food Microbiol. 46:187-192.

43. Milillo SR, Stout JC, Hanning IB, Clement A, Fortes ED, den Bakker HC, Wiedmann M, Ricke SC. 2012. Listeria monocytogenes and hemolytic Listeria innocua in poultry. Poultry Science 91:2158-2163.

44. Moreno LZ, Paixão R, Gobbi DD, Raimundo DC, Ferreira TP, Hofer E, Matte MH, Moreno AM. 2012. Characterization of atypical Listeria innocua isolated from swine slaughterhouses and meat markets. Res. Microbiol. 163:268- 271.

45. Morris IJ, Ribeiro CD. 1989. Listeria monocytogenes and pâté. Lancet 334:1285-1286.

46. Nesbakken T, Kapperud G, Caugant DA. 1996. Pathways of Listeria monocytogenes contamination in the meat processing industry. Int. J. Food Microbiol. 31:161-171.

47. Nightingale KK, Windham K, Wiedmann M. 2005. Evolution and molecular phylogeny of Listeria monocytogenes isolated from human and animal listeriosis cases and foods. J. Bacteriol. 187:5537-5551.

48. Noah CW, Shaw CI, Ikeda JS, Kreuzer KS, Sofos JN. 2005. Development of green fluorescent protein-expressing bacterial strains and evaluation for potential use as positive controls in sample analyses. J. Food Prot. 68:680-686.

122

Texas Tech University, Alex Brandt, May 2014

49. Nocera D, Bannerman E, Rocourt J, Jaton-Ogay K, Bille J. 1990. Characterization by DNA restriction endonuclease analysis of Listeria monocytogenes strains related to the Swiss epidemic of listeriosis. J. Clin. Microbiol. 28:2259-2263.

50. Norton DM, McCamey MA, Gall KL, Scarlett JM, Boor KJ, Wiedmann M. 2001. Molecular studies on the ecology of Listeria monocytogenes in the smoked fish processing industry. Appl. Environ. Microbiol. 67:198-205.

51. Orsi RH, Borowsky ML, Lauer P, Young SK, Nusbaum C, Galagan JE, Birren BW, Ivy RA, Sun Q, Graves LM, Swaminathan B, Wiedmann M. 2008. Short-term genome evolution of Listeria monocytogenes in a non-controlled environment. BMC Genom. 9:539.

52. Peccio A, Autio T, Korkeala H, Rosmini R, Trevisani M. 2003. Listeria monocytogenes occurrence and characterization in meat-producing plants. Lett. Appl. Microbiol. 37:234-238.

53. Public Health Agency of Canada. 2008. Lessons learned: public health agency of Canada's response to the 2008 listeriosis outbreak. Available At: http://www. phac-aspc.gc.ca/fs-sa/listeria/2008-lessons-lecons-eng.php. Accessed On: 2/12/14

54. Porsby CH, Vogel BF, Mohr M, Gram L. 2008. Influence of processing steps in cold-smoked salmon production on survival and growth of persistent and presumed non-persistent Listeria monocytogenes. Int. J. Food Microbiol. 122:287-295.

55. Ringus DL, Ivy RA, Wiedmann M, Boor KJ. 2012. Salt stress-induced transcription of σB- and CtsR-regulated genes in persistent and non-persistent Listeria monocytogenes strains from food processing plants. Foodborne Pathog. Dis. 9:198-206.

56. Rørvik LM, Caugant DA, Yndestad M. 1995. Contamination pattern of Listeria monocytogenes and other Listeria spp. in a salmon slaughterhouse and smoked salmon processing plant. Int. J. Food Microbiol. 25:19-27.

57. Russo ET, Biggerstaff G, Hoekstra RM, Meyer S, Patel N, Miller B, Quick R. 2013. A recurrent, multistate outbreak of Salmonella serotype Agona infections associated with dry, unsweetened cereal consumption, United States, 2008. J. Food Prot. 76:227-230.

123

Texas Tech University, Alex Brandt, May 2014

58. Sauders BD, Sanchez MD, Rice DH, Corby J, Stich S, Fortes ED, Roof SE, Wiedmann M. 2009. Prevalence and molecular diversity of Listeria monocytogenes in retail establishments. J. Food Prot. 72:2337-2349.

59. Thévenot D, Delignette-Muller ML, Christieans S, Vernozy-Rozand C. 2005. Prevalence of Listeria monocytogenes in 13 dried sausage processing plants and their products. Int. J. Food Microbiol. 102:85-94.

60. Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Tracking of Listeria monocytogenes in smoked fish processing plants. J. Food Prot. 67:328- 341.

61. Tompkin RB. 2002. Control of Listeria monocytogenes in the food-processing environment. J. Food Prot. 65:709-725.

62. U.S. Department of Agriculture, Food Safety and Inspection Service. 1996. Pathogen reduction; hazard analysis and critical control point (HACCP) systems; final rule. Fed. Reg. 61:38805-38989.

63. U.S. Department of Agriculture, Food Safety and Inspection Service. 2001. Oklahoma firm recalls meat and poultry products for possible Listeria contamination. Available At: http://www.fsis.usda.gov/Oa/recalls/prelease/pr021- 2001.htm?redirecthttp=true. Accessed On: 2/12/14

64. U.S. Department of Agriculture, Food Safety and Inspection Service. 2003. Control of Listeria monocytogenes in ready-to-eat meat and poultry products; final rule. Fed. Reg. 68:34207-34254.

65. U.S. Department of Agriculture, Food Safety and Inspection Service. 2014. FSIS compliance guideline: controlling Listeria monocytogenes in post-lethality exposed ready-to-eat meat and poultry products. Available At: http:// www.fsis.usda.gov / shared / PDF / Controlling_LM_RTE_guideline_0912.pdf. Accessed On: 2/12/14

66. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1995. Escherichia coli O157:H7 outbreak linked to commercially distributed dry-cured salami - Washington and California, 1994. Morb. Mortal. Wkly. Rep. 44:157.

67. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1998. Multistate outbreak of Salmonella serotype Agona infections linked to toasted oats cereal--United States, April-May, 1998. J. Am. Med. Assoc. 280:411.

124

Texas Tech University, Alex Brandt, May 2014

68. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 1999. Update: multistate outbreak of listeriosis - United States, 1998-1999. Morb. Mortal. Wkly. Rep. 47:1117-1118.

69. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2000. Multistate outbreak of listeriosis - United States, 2000. Morb. Mortal. Wkly. Rep. 49:1129-1130.

70. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2008. Multistate outbreak of Salmonella Agona infections linked to rice & wheat puff cereal (final update). Available At: http://www. cdc.gov/salmonella/agona. Accessed On: 1/15/14

71. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2011. Multistate outbreak of human Salmonella Heidelberg infections linked to ground turkey. Available At: http://www.cdc.gov/salmonella /heidelberg/index.html. Accessed On: 1/29/14

72. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2012. Multistate outbreak of Salmonella Enteritidis infections linked to ground beef (final update). Available At: http://www.cdc.gov/salmonella /enteritidis-07-12/index.html. Accessed On: 1/29/14

73. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2013. Foodborne diseases active surveillance network (FoodNet). Available At: http://www.cdc.gov/foodnet/about.html. Accessed On: 1/20/14

74. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2014. Multistate outbreak of multidrug-resistant Salmonella Heidelberg infections linked to Foster Farms brand chicken. Available At: http://www.cdc.gov / salmonella / heidelberg-10-13 / index.html. Accessed On: 1/29/14

75. Unnerstad H, Bannerman E, Bille J, Danielsson-Tham ML, Waak E, Tham W. 1996. Prolonged contamination of a dairy with Listeria monocytogenes. Neth. Milk Dairy J. 50:493-500.

76. Vangay P, Fugett EB, Sun Q, Wiedmann M. 2013. Food Microbe Tracker: a web-based tool for storage and comparison of food-associated microbes. J. Food Prot. 76:283-294.

125

Texas Tech University, Alex Brandt, May 2014

77. Williams SK. 2010. Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. M.S. Thesis: Department of Animal Sciences, Colorado State University, 168 p.

78. Williams SK, Roof S, Boyle EA, Burson D, Thippareddi H, Geornaras I, Sofos JN, Wiedmann M, Nightingale K. 2011. Molecular ecology of Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. J. Food Prot. 74:63-77.

79. Zhao S, McDermott PF, Friedman S, Abbott J, Ayers S, Glenn A, Hall- Robinson E, Hubert SK, Harbottle H, Walker RD, Chiller TM, White DG. 2006. Antimicrobial resistance and genetic relatedness among Salmonella from retail foods of animal origin: NARMS retail meat surveillance. Foodborne Pathog. Dis. 3:106-117.

126

Texas Tech University, Alex Brandt, May 2014

TABLE 2.1 Polymerase chain reaction primer sequences and thermal cycling conditions Primer Primer Assay PCR Conditions Reference Name Sequence EC hly F 5’-CCCTGGCAGACC (Forward) TTTGATG-3’ (55) EC hly R 5’-CCGTGTCTTTTC (Reverse) TGATACTCA-3’ flich7-F 5’-GCGCTGTCGAGT (Forward) TCTATCGAGC-3’ (33) flich7-R 5’-CAACGGTGACTT (Reverse) ATCGCCATTCC-3’ int-F 5’-GACTGTCGATGC Initial: 94°C, 2 min (Forward) ATCAGGCAAAG-3’ 20 Cycles: 94°C, 30 s; int-R 5’-TTGGAGTATTAAC 59-49°C (-0.5°C per E. coli O157:H7 (Reverse) ATTAACCCCAGG-3’ cycle), 1 min;72°C, 1 min Multiplex PCR (42) rfb-F 5’-GTGTCCATTTATA 20 Cycles: 94°C, 30 s; Confirmation (Forward) CGGACATCCATG-3’ 49°C, 1 min; 72°C, 1 min rfb-R 5’-CCTATAACGTCA Final: 72°C, 7 min (Reverse) TGCCAATATTGCC-3’ Hold: 4°C stx1-F 5’-TGTAACTGGAAA (Forward) GGTGGAGTATAC-3’ stx1-R 5’-GCTATTCTGAGTC (Reverse) AACGAAAAATAAC-3’ (56) stx2-F 5’-GTTTTTCTTCGG (Forward) TATCCTATTCCG-3’ stx2-R 5’-GATGCATCTCTG (Reverse) GTCATTGTATTAC-3’ lmo1118-F 5’-AGGGGTCTTAAA (Forward) TCCTGGAA-3’ lmo1118-R 5’-CGGCTTGTTCGG (Reverse) CATACTTA-3’ lmo0737-F 5’-AGGGCTTCAAGG (Forward) ACTTACCC-3’ lmo0737-R 5’-ACGATTTCTGCTT (Reverse) GCCATTC-3’ Initial: 94°C, 3 min L. monocytogenes ORF2110-F 5’-AGTGGACAATTG 35 Cycles: 94°C, 24 s; Multiplex PCR for (Forward) ATTGGTGAA-3’ 53°C, 69 s; 72°C, 69 s (22) Molecular ORF2110-R 5’-CATCCATCCCTTA Final: 72°C, 7 min Serogrouping (Reverse) CTTTGGAC-3’ Hold: 4°C ORF2819-F 5’-AGCAAAATGCCA (Forward) AAACTCGT-3’ ORF2819-R 5’-CATCACTAAAGC (Reverse) CTCCCATTG-3’ prs-F 5’-GCTGAAGAGATT (Forward) GCGAAAGAAG-3’ prs-R 5’-CAAAGAAACCTT (Reverse) GGATTTGCGG-3’ CAA1- 5’-GAATCCTCAGTT Initial: 94°C, 2 min invAF TTTCAACGTTTC-3’ 35 Cycles: 94°C, 30 s; S. enterica (Forward) 60°C, 30 s; 72°C, 30 s (47) invA Confirmation CAA2- 5’-AGCCGTAACAA Final: 72°C, 7 min invAR CCAATACAAATG-3’ Hold: 4°C (Reverse)

127

Texas Tech University, Alex Brandt, May 2014

TABLE 2.1, Cont. Polymerase chain reaction primer sequences and thermal cycling conditions Primer Primer Assay PCR Conditions Reference Name Sequence

Initial: 94°C, 5 min LM hly-α 5’-CCTAAGACGCCA 20 Cycles: 94°C, 1 min; (Forward) ATCGAAAAGAAA-3’ 54-44°C (-0.5°C per L. monocytogenes cycle), 1 min; 72°C, 1 min hlyA (65) 20 Cycles: 94°C, 1 min; Confirmation LM hly-β 5’-TAGTTCTACATC 44°C, 1 min; 72°C, 1 min (Reverse) ACCTGAGACAGA-3’ Final: 72°C, 7 min Hold: 4°C

5’-AATATATTAATG Initial: 94°C, 5 min SigB 15 AAAAGCAGGTGGA 20 Cycles: 94°C, 1 min; (Forward) 54-44°C (-0.5°C per Listeria spp. G-3’ cycle), 1 min; 72°C, 1 min sigB (62) 20 Cycles: 94°C, 1 min; Confirmation SigB 16 5’-ATAAATTATTTG 44°C, 1 min; 72°C, 1 min (Reverse) ATTCAACTGCCTT-3’ Final: 72°C, 7 min Hold: 4°C

128

Texas Tech University, Alex Brandt, May 2014

TABLE 2.2 Synopsis of pertinent characteristics for facilities enrolled in the study Operation Inspection Facility Square Full Time Number of Facility ID Typea Programd Age (Years) Footage Employees Shifts USDA- A RTE 27 7800 20 1 FSIS Custom- 50; 6 yr old B RTEb 5500 11 1 Exempt remodel Custom- 49; 18 yr old C FRESHc 2500 7 1 Exempt cutting room USDA- D FRESHc 7 5000 15 1 FSIS

a Primary operations conducted by the facility: RTE, ready-to-eat meat processing; FRESH, animal slaughter and fresh meat cutting and grinding. b Slaughter component also included as part of facility operations. c Ready-to-eat processing component also included as part of facility operations. d Inspection program used in facility: USDA-FSIS, federal inspection by U.S. Department of Agriculture Food Safety and Inspection Service; Custom-Exempt, annual inspection by the Colorado Department of Agriculture as prescribed by the 1970 Curtis Amendment to the Wholesome Meat Act of 1967.

129

Texas Tech University, Alex Brandt, May 2014

TABLE 2.3 Prevalence of each organism by sampling site category in each facility Sampling Site Category Listeria monocytogenes Other Listeria Species Salmonella enterica Facility A Drain and 19.23% 23.08% 0.00% Floor Sites Non-Food Contact 2.90% 2.17% 0.00% Environmental Sites Employee Surfaces and 22.22% 11.11% 0.00% Employee Contact Sites Food Contact Sites 12.82% 7.69% 0.00% and Food Samples Facility B Drain and Floor Sites 42.31% 56.41% 6.41% Non-Food Contact Environmental Sites 6.25% 14.06% 0.00% Employee Surfaces and Employee Contact Sites 13.79% 31.03% 0.00% Food Contact Sites and Food Samples 3.33% 13.33% 0.00% Facility C Drain and Floor Sites 51.39% 72.22% 20.83% Non-Food Contact Environmental Sites 8.64% 15.43% 3.09% Employee Surfaces and Employee Contact Sites 13.89% 25.00% 5.56% Food Contact Sites and Food Samples 0.00% 11.67% 1.67%

Facility D Drain and Floor Sites 5.63% 14.08% 5.63% Non-Food Contact Environmental Sites 0.00% 6.00% 2.67%

Employee Surfaces and 0.00% 4.17% 0.00% Employee Contact Sites Food Contact Sites 0.00% 0.00% 0.00% and Food Samples All Facilities Drain and 29.77% 41.47% 8.03% Floor Sites Non-Food Contact 4.50% 9.52% 1.56% Environmental Sites Employee Surfaces and 13.60% 18.40% 1.60% Employee Contact Sites Food Contact Sites and Food Samples 4.17% 8.01% 0.32%

130

Texas Tech University, Alex Brandt, May 2014

TABLE 2.4 Facility-level persistence metrics for L. monocytogenes, using a binomial test for ribotype frequency

Reference Distribution of All Ribotypes Reference Distribution of Ribtoypes from Food Environments

Facility-Level Ribotype Frequency Ribotype Frequency c c Ribotype Frequencya [Binomial 99.5% CI]b Significance [Binomial 99.5% CI]b Significance

Facility A DUP-1052A 0.639 (23/36) 0.076 (380/5000) [0.065, 0.087] < 2.20 X 10-16* 0.122 (172/1411) [0.099, 0.148] 4.38 X 10-13* -15 116-239-S-2 0.306 (11/36) 0.008 (41/5000) [0.005, 0.012] 5.61 X 10 * 0.014 (20/1411) [0.007, 0.025] 2.01 X 10-12*

DUP-1048A 0.000 (0/36) 0.001 (5/5000) [1.83 X 10-4, 0.003] 1.00 X 100 0.003 (4/1411) [3.92 X 10-4, 0.010] 1.00 X 100 Facility B

-16 -16 DUP-1062E 0.383 (18/47) 0.008 (38/5000) [0.004, 0.012] < 2.20 X 10 * 0.016 (23/1411) [0.008, 0.028] < 2.20 X 10 * DUP-1062B 0.361 (17/47) 0.019 (95/5000) [0.014, 0.025] < 2.20 X 10-16* 0.060 (85/1411) [0.044, 0.080] 8.61 X 10-10*

DUP-1053E -11 -10 0.191 (9/47) ]0.006 (31/5000) [0.004, 0.010] 1.49 X 10 * 0.009 (13/1411) [0.004, 0.019] 4.75 X 10 * DUP-1042C 0.000 (0/47) 0.021 (103/5000) [0.015, 0.027] 1.00 X 100 0.018 (26/1411) [0.010, 0.031] 1.00 X 100 Facility C

-16 DUP-1042B 0.964 (53/55) 0.105 (526/5000) [0.093, 0.118] < 2.20 X 10 * 0.089 (126/1411) [0.069, 0.113] < 2.20 X 10-16* DUP-1057B 0.018 (1/55) 0.002 (11/5000) [0.001, 0.005] 1.14 X 10-1 0.006 (8/1411) [0.002, 0.014] 2.69 X 10-1 ] ]

131

Texas Tech University, Alex Brandt, May 2014

TABLE 2.4, Cont. Facility-level persistence metrics for L. monocytogenes, using a binomial test for ribotype frequency

Reference Distribution of All Ribotypes Reference Distribution of Ribtoypes from Food Environments

Facility-Level Ribotype Frequency Ribotype Frequency c c Ribotype Frequencya [Binomial 99.5% CI]b Significance [Binomial 99.5% CI]b Significance

Facility D

DUP-1030B 0.667 (2/3) 0.009 (45/5000) [0.006, 0.013] 2.42 X 10-4* 0.003 (4/1411) [3.92 X 10-4, 0.010] 2.41 X 10-5* 0 0 DUP-1030A 0.000 (0/3) 0.020 (98/5000) [0.015, 0.026] 1.00 X 10 0.006 (8/1411) [0.002, 0.014] 1.00 X 10 a Frequency at the facility level is the number of samples positive for a particular L. monocytogenes ribotype (excluding the first positive sample) divided by ] the total number of samples positive for L. monocytogenes in the facility (excluding the first positive sample) for the six-month sampling period. This strategy for calculating the frequency] was employed because the statistical methods used assess conditional probabilities; that is, the probability that a sample is positive for a given ribotype, given that the ribotype was introduced into the facility. b Frequency of a ribotype was calculated as the proportion of entries of that particular ribotype in the reference distribution that was used. For the reference distribution of all L. monocytogenes, the reference distribution contained 5000 entries with a ribotype designation (ALL LM; Supplemental File 6). For the reference distribution consisting of L. monocytogenes from food environments, the reference distribution contained 1411 entries (ENV LM; Supplemental File 7); (Number of entries of that ribotype in the particular reference distribution/Total number of entries in the reference distribution); the 99.5% binomial confidence interval is provided for each frequency in square brackets. c P values are calculated from the binom.test function in R where facility level frequency for a ribotype is compared with the frequency in the particular reference distribution of interest. Values with a * were significant after applying a Bonferroni correction to the α; P value for Facility A was 1.67 X 10-2 (0.05/3). P value for Facility B was 1.25 X 10-2 (0.05/4). P value for Facility C was 2.50 X 10-2 (0.05/2). P value for Facility D was 2.50 X 10-2 (0.05/2).

132

Texas Tech University, Alex Brandt, May 2014

TABLE 2.5 Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test for sigB ATa frequency

Reference Distribution of All ATs Reference Distribution of ATs from Food Environments Facility-Level AT Frequency AT Frequency AT Frequencyb [Binomial 99.5% CI]c Significanced [Binomial 99.5% CI]c Significanced

Facility A

LI AT 56 0.533 (16/30) 0.048 (92/1928) [0.035, 0.063] 5.53 X 10-14* 0.054 (42/773) [0.034, 0.081] 4.03 X 10-13* LW AT 69 0.233 (7/30) 0.047 (91/1928) [0.035, 0.062] 4.06 X 10-4* 0.054 (42/773) [0.034, 0.081] 9.38 X 10-4* -1 -1 LI AT 37 0.033 (1/30) 0.041 (71/1928) [0.029, 0.055] 7.15 X 10 0.087 (67/773) [0.061, 0.119] 9.34 X 10 LW AT 27 0.033 (1/30) 0.062 (120/1928) [0.048, 0.079] 8.55 X 10-1 0.038 (29/773) [0.021, 0.061] 6.38 X 10-1 LW AT 19 0.000 (0/30) 0.016 (30/1928) [0.009, 0.025] 1.00 X 100 0.001 (1/773) [3.24 X 10-6, 1.06 X 10-2] 1.00 X 100

-4 0 0 LW AT 32 0.000 (0/30) 0.003 (6/1928) [6.91 X 10 , 0.009] 1.00 X 10 0.006 (5/773) [0.001, 0.019] 1.00 X 10 Facility B -11 -4 LI AT 11 0.305 (25/82) 0.066 (127/1928) [0.051, 0.083] 5.39 X 10 * 0.154 (119/773) [0.119, 0.194] 4.33 X 10 * LI AT 6 0.244 (20/82) 0.042 (81/1928) [0.030, 0.056] 1.46 X 10-10* 0.076 (59/773) [0.052, 0.107] 2.68 X 10-6* LI AT 23 0.171 (14/82) 0.032 (62/1928) [0.022, 0.045] 3.54 X 10-7* 0.032 (25/773) [0.017, 0.054] 3.79 X 10-7*

-3 -4 LI AT 38 0.061 (5/82) 0.011 (22/1928) [0.006, 0.020] 2.55 X 10 * 0.008 (6/773) [0.002, 0.021] 4.68 X 10 * LI AT 45 0.049 (4/82) 0.013 (25/1928) [0.007, 0.022] 2.22 X 10-2 0.028 (22/773) [0.014, 0.050] 8.49 X 10-2 -1 -1 LI AT 37 0.024 (2/82) 0.041 (79/1928) [0.029, 0.055] 8.54 X 10 0.087 (67/773) [0.061, 0.119] 9.95 X 10 LI AT 31 0.012 (1/82) 0.027 (52/1928) [0.018, 0.039] 8.94 X 10-1 0.034 (26/773) [0.018, 0.056] 7.67 X 10-1 LI AT 94 0.000 (0/82) 0.003 (6/1928) [6.91 X 10-4, 0.009] 1.00 X 100 0.001 (1/773) [3.23 X 10-6, 1.06 X 10-2] 1.00 X 100

-4 0 -5 -2 0 LI AT 44 0.000 (0/82) 0.003 (6/1928) [6.91 X 10 , 0.009] 1.00 X 10 0.003 (2/773) [9.38 X 10 , 1.30 X 10 ] 1.00 X 10 LI AT 109 0.000 (0/82) 0.007 (14/1928) [0.003, 0.015] 1.00 X 100 0.013 (10/773) [0.004, 0.029] 1.00 X 100

LI AT 26 0.000 (0/82) 0.031 (60/1928) [0.021, 0.044] 1.00 X 100 0.012 (9/773) [0.004, 0.027] 1.00 X 100

LI AT 30 0.000 (0/82) 0.013 (25/1928) [0.007, 0.022] 1.00 X 100 0.008 (6/773) [0.002, 0.021] 1.00 X 100

133

Texas Tech University, Alex Brandt, May 2014

TABLE 2.5, Cont. Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test for sigB ATa frequency

Reference Distribution of All ATs Reference Distribution of ATs from Food Environments Facility-Level AT Frequency AT Frequency

AT Frequencyb [Binomial 99.5% CI]c Significanced [Binomial 99.5% CI]c Significanced

Facility C

LI AT 70 0.533 (49/92) 0.042 (81/1928) [0.030, 0.056] < 2.20 X 10-16* 0.103 (80/773) [0.075, 0.138] <2.20 X 10-16* LI AT 71 0.141 (13/92) 0.030 (58/1928) [0.020, 0.043] 3.99 X 10-6* 0.035 (27/773) [0.019, 0.058] 1.94 X 10-5* -1 -1 LI AT 23 0.043 (4/92) 0.032 (62/1928) [0.022, 0.045] 3.43 X 10 0.032 (25/773) [0.017, 0.054] 3.47 X 10 LI AT 6 0.043 (4/92) 0.042 (81/1928) [0.030, 0.056] 5.44 X 10-1 0.076 (59/773) [0.052, 0.107] 9.27 X 10-1 -1 -2 LI AT 26 0.043 (4/92) 0.031 (60/1928) [0.021, 0.044] 3.22 X 10 0.012 (9/773) [0.004, 0.027] 2.29 X 10

LI AT 109 0.033 (3/92) 0.007 (14/1928) [0.003, 0.015] 2.98 X 10-2 0.013 (10/773) [0.004, 0.029] 1.17 X 10-1 LI AT 30 0.022 (2/92) 0.013 (25/1928) [0.007, 0.022] 3.35 X 10-1 0.008 (6/773) [0.002, 0.021] 1.60 X 10-1 -1 0 LI AT 37 0.011 (1/92) 0.041 (79/1928) [0.029, 0.055] 9.79 X 10 0.087 (67/773) [0.061, 0.119] 1.00 X 10 LI AT 53 0.000 (0/92) 0.011 (21/1928) [0.005, 0.019] 1.00 X 100 0.016 (12/773) [0.006, 0.033] 1.00 X 100 LI AT 31 0.000 (0/92) 0.027 (52/1928) [0.018, 0.039] 1.00 X 100 0.034 (26/773) [0.018, 0.056] 1.00 X 100

-6 -3 0 -6 -2 0 LW AT NA 0.000 (0/92) 0.001 (1/1928) [1.30 X 10 , 4.25 X 10 ] 1.00 X 10 0.001 (1/773) [3.24 X 10 , 1.06 X 10 ] 1.00 X 10 LI AT 11 0.000 (0/92) 0.066 (127/1928) [0.051, 0.083] 1.00 X 100 0.154 (119/773) [0.119, 0.194] 1.00 X 100 -4 0 -6 -2 0 HLI AT 141 0.000 (0/92) 0.002 (3/1928) [1.37 X 10 , 0.006] 1.00 X 10 0.001 (1/773) [3.24 X 10 ,1.06 X 10 ] 1.00 X 10 Facility D LI AT 6 0.316 (6/19) 0.042 (81/1928) [0.030, 0.056] 9.28 X 10-5* 0.076 (59/773) [0.052, 0.107] 2.25 X 10-3*

-8 -7 LW AT 129 0.263 (5/19) 0.005 (10/1928) [0.002, 0.012] 4.11 X 10 * 0.008 (6/773) [0.002, 0.021] 2.99 X 10 * LI AT 53 0.053 (1/19) 0.011 (21/1928) [0.005, 0.019] 1.88 X 10-1 0.016 (12/773) [0.006, 0.033] 2.57 X 10-1 -4 0 -6 -2 0 LI AT 124 0.000 (0/19) 0.002 (3/1928) [1.37 X 10 , 0.006] 1.00 X 10 0.001 (1/773) [3.24 X 10 , 1.06 X 10 ] 1.00 X 10 LS AT 121 0.000 (0/19) 0.004 (8/1928) [0.001, 0.010] 1.00 X 100 0.001 (1/773) [3.24 X 10-6, 1.06 X 10-2] 1.00 X 100

134

Texas Tech University, Alex Brandt, May 2014

TABLE 2.5, Cont. Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test for sigB ATa frequency

Reference Distribution of All ATs Reference Distribution of ATs from Food Environments Facility-Level AT Frequency AT Frequency c d c d AT Frequencyb [Binomial 99.5% CI] Significance [Binomial 99.5% CI] Significance

Facility D LI AT 109 0.000 (0/19) 0.007 (14/1928) [0.003, 0.015] 1.00 X 100 0.013 (10/773) [0.004, 0.029] 1.00 X 100 LW AT NB 0.000 (0/19) 0.001 (1/1928) [1.30 X 10-6, 4.25 X 10-3] 1.00 X 100 0.001 (1/773) [3.24 X 10-6, 1.06 X 10-2] 1.00 X 100

LI AT 31 0.000 (0/19) 0.027 (52/1928) [0.018, 0.039] 1.00 X 100 0.034 (26/773) [0.018, 0.056] 1.00 X 100 a AT is an abbreviation for “allelic type.” Organism species are abbreviated according to the following: Listeria innocua (LI), hemolytic Listeria innocua (HLI), Listeria welshimeri (LW), and Listeria seeligeri (LS). b Frequency at the facility level is the number of samples positive for a particular Listeria sigB AT (excluding the first positive sample) divided by the total number of samples positive for non-pathogenic Listeria species in the facility (excluding the first positive sample) for the six-month sampling period. This strategy for calculating the frequency was employed because the statistical methods used assess conditional probabilities; that is, the probability that a sample is positive for a given sigB AT, given that the sigB AT was introduced into the facility. c Frequency of a sigB AT was calculated as the proportion of entries of that particular sigB AT in the reference distribution that was used. For the reference distribution of all non-pathogenic Listeria species, the reference distribution contained 1928 entries with a sigB AT designation (ALL LS; Supplemental File 8). For the reference distribution consisting of non-pathogenic Listeria species from food environments, the reference distribution contained 773 entries (ENV

LS; Supplemental File 9); (Number of entries of that sigB AT in the particular reference distribution/Total number of entries in the reference distribution); the 99.5% binomial confidence interval is provided for each frequency in the square brackets. d P values are calculated from the binom.test function in R where facility level frequency for a sigB AT is compared with the frequency in the particular reference distribution of interest. Values with a * were significant after applying a Bonferroni correction to the α; P value for Facility A was 1.00 X 10-2 (0.05/5). P value for Facility B was 4.17 X 10-3 (0.05/12). P value for Facility C was 3.85 X 10-3 (0.05/13). P value for Facility D was 6.25 X 10-3 (0.05/8).

135

Texas Tech University, Alex Brandt, May 2014

TABLE 2.6 Metrics for facility sites demonstrating significant levels of L. monocytogenes persistence, using a binomial test for ribotype frequency

Ribotype of Site-Level Significance for Reference Distribution Significance for Reference Distribution Site for Persistencea Interestb Frequencyc of All Ribotypesd of Ribotypes from Food Environmentsd Facility A

-5 -4 Site 3 - Grinding Room Drains and Floor 116-239-S-2 1.000 (2/2) 6.72 X 10 * 2.01 X 10 * Site 10 - Stuffing Room Drains and Floor 116-239-S-2 0.667 (2/3) 2.01 X 10-4* 5.97 X 10-4*

Facility B

Site 44 - Processing Employee Boots DUP-1062E 1.000 (2/2) 5.78 X 10-5* 2.66 X 10-4* Facility C -3 -4 Site 15 - Floor Next to Closed Square DUP-1042B 1.000 (3/3) 1.16 X 10 * 7.12 X 10 * Drain Site 19 - Smoking Area Floor and Wall Jct. DUP-1042B 1.000 (3/3) 1.16 X 10-3* 7.12 X 10-4* -3 -4 Site 26 - Exsanguination Blood Catch DUP-1042B 1.000 (3/3) 1.16 X 10 * 7.12 X 10 *

Drain -4 -5 Site 27 - Kill Floor Area Floor DUP-1042B 1.000 (4/4) 1.23 X 10 * 6.36 X 10 * Site 39 - Middle Cooler Drain DUP-1042B 1.000 (4/4) 1.23 X 10-4* 6.36 X 10-5* -3 -4 Site 48 - Processing Employee Boots DUP-1042B 1.000 (3/3) 1.16 X 10 * 7.12 X 10 * Facility D None Identified a All sites where a given ribotype was isolated on at least two different instances were included in the original analysis. However, the only sites reported here are those where a comparison between site level frequency and the frequency of a particular reference distribution produced a statistically significant result. b The ribotype whose frequency among all isolates at the particular site was statistically significant when compared to the reference distributions of interest. c Frequency at the site level is the number of samples positive for a particular L. monocytogenes ribotype (excluding the first positive sample) divided by the total number of samples positive for L. monocytogenes at that site (excluding the first positive sample) for the six-month sampling period. For instance, the designation 2/3 refers to three samples positive for a given ribotype out of four samples that produced an L. monocytogenes isolate of any ribotype. d Frequencies of each ribotype in the particular reference distribution denoted can be found in Table 2.6. P values are calculated from the binom.test function in R where site level frequency for a ribotype is compared with the frequency in the particular reference distribution of interest. Values with a * were significant after applying a Bonferroni correction to the α, whereby the α is divided by the total number of sites in a facility which produced a sample positive for L. monocytogenes on at least two instances; P value for Facility A was 5.00 X 10-3 (0.05/10). P value for Facility B was 4.17 X 10-3 (0.05/12). P value for Facility C was 3.33 X 10-3 (0.05/15). P value for Facility D is not applicable. 136

Texas Tech University, Alex Brandt, May 2014

TABLE 2.7 Metrics for facility sites demonstrating significant levels of non-pathogenic Listeria species persistence, using a binomial test for sigB ATa frequency

AT of Site-Level Significance for Reference Significance for Reference Distribution b Site for Persistence Interesta,c Frequencyd Distribution of All ATse of ATs from Food Environmente

Facility A

Site 25 - Packaging Room Drains and Floor LW AT 69 1.000 (2/2) 2.23 X 10-3* 2.95 X 10-3* Facility B

-3 -3 Site 10 - Smokehouse #2 Bottom and Ramp LI AT 6 1.000 (2/2) 1.77 X 10 * 5.83 X 10 Site 12 - Squeegees LI AT 23 1.000 (2/2) 1.03 X 10-3* 1.05 X 10-3*

Facility C Site 18 - Smokehouse Door and Handle LI AT 70 1.000 (2/2) 1.77 X 10-3* 1.07 X 10-2

Site 19 - Smoking Area Floor and Wall Jct. LI AT 70 0.750 (3/4) 2.87 X 10-4* 4.09 X 10-3

Site 26 - Exsanguination Blood Catch Drain LI AT 70 0.600 (3/5) 6.96 X 10-4* 9.44 X 10-3 Facility D

-5 -5 Site 25 - Door from Processing to Cooler LW AT 129 1.000 (2/2) 2.69 X 10 * 6.03 X 10 * a AT is an abbreviation for “allelic type.” Organism species are abbreviated according to the following: Listeria innocua (LI) and Listeria welshimeri (LW) b All sites where a given sigB AT was isolated on at least two different instances were included in the original analysis. However, the only sites reported here are those where a comparison between site level frequency and the frequency of a particular reference distribution produced a statistically significant result. b The sigB AT isolated at a particular site whose frequency among all isolates at that site was statistically significant when compared to the reference distributions of interest. c Frequency at the site level is the number of samples positive for a particular non-pathogenic Listeria species sigB AT (excluding the first positive sample) divided by the total number of samples positive for non-pathogenic Listeria species at that site (excluding the first positive sample) for the six-month sampling period. For instance, the designation 3/4 refers to four samples positive for a given sigB AT out of five samples that produced a non-pathogenic Listeria species isolate of any sigB AT. d Frequencies of each sigB AT in the particular reference distribution denoted can be found in Table 2.7. P values are calculated from the binom.test function in R where site level frequency for a sigB allelic type is compared with the frequency in the particular reference distribution of interest. Values with a * were significant after applying a Bonferroni correction to the α, whereby the α is divided by the total number of sites in a facility which produced a sample positive for non-pathogenic Listeria species on at least two instances; P value for Facility A was 6.25 X 10-3 (0.05/8). P value for Facility B was 2.50 X 10-3 (0.05/20). P value for Facility C was 2.27 X 10-3 (0.05/22). P value for Facility D was 8.33 X 10-3 (0.05/6).

137

Texas Tech University, Alex Brandt, May 2014

TABLE 2.8 Facility-level persistence metrics for L. monocytogenes, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsa Fisher’s Exact Test c Ribotype ++ +- -+ -- Relative Risk [95% CI]b (Right One-Tail) Significance

Facility A DUP-1052A 5 19 14 237 3.74 [1.472, 9.479] 1.67 X 10-2* 116-239-S-2 3 8 5 259 14.40 [3.932, 52.738] 2.41 X 10-3*

DUP-1048A 0 1 1 273 NC [NC, NC]d 1.00 X 100 Facility B -3 DUP-1062E 4 14 8 240 6.89 [2.291, 20.711] 5.23 X 10 * DUP-1062B 2 14 10 240 3.13 [0.747, 13.079] 1.57 X 10-1

DUP-1053E 0 9 9 248 NC [NC, NC] 1.00 X 100

DUP-1042C 0 1 1 264 NC [NC, NC] 1.00 X 100 Facility C

-7 DUP-1042B 21 23 27 204 4.08 [2.551, 6.536] 2.29 X 10 * DUP-1057B 0 2 1 272 NC [NC, NC] 1.00 X 100

138

Texas Tech University, Alex Brandt, May 2014

TABLE 2.8, Cont. Facility-level persistence metrics for L. monocytogenes, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsa Fisher’s Exact Test c Ribotype ++ +- -+ -- Relative Risk [95% CI]b (Right One-Tail) Significance

Facility D DUP-1030B 0 2 2 269 NC [NC, NC] 1.00 X 100 DUP-1030A 0 1 0 272 NC [NC, NC] 1.00 X 100 a Transitions from one status (positive or negative) to another (positive or negative) were recorded at every site for a specific L. monocytogenes ribotype. For instance, ++ indicates a current positive result following a previous positive result on the immediate prior sampling date, while -+ indicates a current positive result following a previous negative result on the immediate prior sampling date. b Relative risk was calculated using the MedCalc online Relative Risk Calculator and represents the probability of a current positive result given a previous positive result divided by the probability of a current positive result given a previous negative result. A 95% confidence interval calculated by the MedCalc relative risk calculator is also provided. c P values are calculated from the fisher.test function in R using the right one-tailed test option. Values with a * were significant after applying a Bonferroni correction to the α; P value for Facility A was 1.67 X 10-2 (0.05/3). P value for Facility B was 1.25 X 10-2 (0.05/4). P value for Facility C was 2.50 X 10-2 (0.05/2). P value for Facility D was 2.50 X 10-2 (0.05/2). d Abbreviation for “not calculated.” Values were not able to be calculated due to a value of 0 in cell A (corresponding to the ++ transition) of the 2 X 2 table.

139

Texas Tech University, Alex Brandt, May 2014

TABLE 2.9 Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsb Fisher’s Exact Test d AT a ++ +- -+ -- Relative Risk [95% CI]c (Right One-Tail) Significance

Facility A

-1 LI AT 56 1 15 16 243 1.01 [0.143, 7.155] 6.50 X 10 LW AT 69 2 4 3 266 29.89 [6.060, 147.430] 3.87 X 10-3* e 0 LI AT 37 0 2 2 271 NC [NC, NC] 1.00 X 10 LW AT 27 0 0 2 273 NC [NC, NC] 1.00 X 100 LW AT 19 0 1 1 273 NC [NC, NC] 1.00 X 100

0 LW AT 32 0 1 1 273 NC [NC, NC] 1.00 X 10 Facility B

LI AT 11 3 21 23 219 1.32 [0.426, 4.062] 4.24 X 10-1

LI AT 6 4 13 14 235 4.18 [1.545, 11.338] 2.02 X 10-2 LI AT 23 5 9 4 247 22.41 [6.753, 74.376] 2.14 X 10-5*

0 LI AT 38 0 5 4 257 NC [NC, NC] 1.00 X 10 LI AT 45 0 4 5 257 NC [NC, NC] 1.00 X 100

LI AT 37 0 1 3 262 NC [NC, NC] 1.00 X 100

LI AT 31 0 2 2 262 NC [NC, NC] 1.00 X 100 LI AT 94 0 1 0 265 NC [NC, NC] 1.00 X 100 0 LI AT 44 0 1 1 264 NC [NC, NC] 1.00 X 10 LI AT 109 0 1 1 264 NC [NC, NC] 1.00 X 100 0 LI AT 26 0 1 1 264 NC [NC, NC] 1.00 X 10

0 LI AT 30 0 0 1 265 NC [NC, NC] 1.00 X 10

140

Texas Tech University, Alex Brandt, May 2014

TABLE 2.9,Cont. Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsb Fisher’s Exact Test d AT a ++ +- -+ -- Relative Risk [95% CI]c (Right One-Tail) Significance

Facility C

LI AT 70 13 26 28 208 2.81 [1.599, 4.938] 1.37 X 10-3* LI AT 71 1 8 13 253 2.27 [0.333, 15.546] 3.80 X 10-1 0 LI AT 23 0 5 5 265 NC [NC, NC] 1.00 X 10 LI AT 6 2 3 3 267 36.00 [7.601, 170.505] 2.60 X 10-3* LI AT 26 0 5 5 265 NC [NC, NC] 1.00 X 100

0 LI AT 109 0 4 1 270 NC [NC, NC] 1.00 X 10 LI AT 30 0 3 2 270 NC [NC, NC] 1.00 X 100 0 LI AT 37 0 2 2 271 NC [NC, NC] 1.00 X 10 LI AT 53 0 1 0 274 NC [NC, NC] 1.00 X 100 LI AT 31 0 1 0 274 NC [NC, NC] 1.00 X 100

0 LW AT NA 0 1 1 273 NC [NC, NC] 1.00 X 10 LI AT 11 0 1 0 274 NC [NC, NC] 1.00 X 100 0 HLI AT 141 0 1 1 273 NC [NC, NC] 1.00 X 10 Facility D LI AT 6 0 5 4 264 NC [NC, NC] 1.00 X 100

-3 LW AT 129 2 2 4 265 33.50 [8.422, 133.259] 2.38 X 10 * LI AT 53 0 2 2 269 NC [NC, NC] 1.00 X 100

LI AT 124 0 1 0 272 NC [NC, NC] 1.00 X 100

LS AT 121 0 1 1 271 NC [NC, NC] 1.00 X 100

141

Texas Tech University, Alex Brandt, May 2014

TABLE 2.9, Cont. Facility-level persistence metrics for non-pathogenic Listeria species, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsb Fisher’s Exact Test d AT a ++ +- -+ -- Relative Risk [95% CI]c (Right One-Tail) Significance

Facility D LI AT 109 0 1 0 272 NC [NC, NC] 1.00 X 100 LW AT NB 0 1 1 271 NC [NC, NC] 1.00 X 100

LI AT 31 0 0 1 272 NC [NC, NC] 1.00 X 100 a AT is an abbreviation for “allelic type.” Organism species are abbreviated according to the following: Listeria innocua (LI), hemolytic Listeria innocua (HLI), Listeria welshimeri (LW), and Listeria seeligeri (LS). b Transitions from one status (positive or negative) to another (positive or negative) were recorded at every site for a specific non-pathogenic Listeria species sigB AT. For instance, ++ indicates a current positive result following a previous positive result on the immediate prior sampling date, while -+ indicates a current positive result following a previous negative result on the immediate prior sampling date. c Relative risk was calculated using the MedCalc online Relative Risk Calculator and represents the probability of a current positive result given a previous positive result divided by the probability of a current positive result given a previous negative result. A 95% confidence interval calculated by the MedCalc relative risk calculator is also provided. d P values are calculated from the fisher.test function in R using the right one-tailed test option. Values with a * were significant after applying a Bonferroni -2 -3 -3 correction to the α; P value for Facility A was 1.00 X 10 (0.05/5). P value for Facility B was 4.17 X 10 (0.05/12). P value for Facility C was 3.85 X 10 (0.05/13). P value for Facility D was 6.25 X 10-3 (0.05/8). e Abbreviation for “not calculated.” Values were not able to be calculated due to a value of 0 in cell A (corresponding to the ++ transition) of the 2 X 2 table.

142

Texas Tech University, Alex Brandt, May 2014

TABLE 2.10 Facility-level persistence metrics for S. enterica, using a binomial test for serotype frequency

Reference Distribution of All ATs Reference Distribution of ATs from Food Environments Facility-Level Serotype Frequency Serotype Frequency b c b c Serotype Frequencya [Binomial 99.5% CI] Significance [Binomial 99.5% CI] Significance

Facility B

Heidelberg 0.500 (2/4) 0.025 (179/7273) [0.020, 0.030] 3.52 X 10-3* 0.006 (4/672) [8.23 X 10-4, 0.020] 2.11 X 10-4* Chailey 0.000 (0/4) 1.37 X 10-4 (1/7273) [3.44 X 10-7, 0.001] 1.00 X 100 0.001 (1/672) [3.72 X 10-6, 0.012] 1.00 X 100

Liverpool 0.000 (0/4) 0.002 (16/7273) [0.001, 0.004] 1.00 X 100 0.001 (1/672) [3.72 X 10-6, 0.012] 1.00 X 100 Facility C

-14 -7 III 50:r:z 0.318 (7/22) 0.002 (14/7273) [0.001, 0.004] 1.63 X 10 * 0.021 (14/672) [0.009, 0.041] 2.21 X 10 * Infantis 0.227 (5/22) 0.015 (108/7273) [0.011, 0.019] 6.52 X 10-7* 0.028 (19/672) [0.014, 0.051] 2.58 X 10-5*

Uganda 0.136 (3/22) 0.002 (11/7273) [0.001, 0.003] 5.21 X 10-6* 0.006 (4/672) [0.001, 0.020] 2.98 X 10-4*

Anatum 0.045 (1/22) 0.010 (76/7273) [0.007, 0.014] 2.06 X 10-1 0.034 (23/672) [0.018, 0.059] 5.35 X 10-1 Muenchen 0.045 (1/22) 0.006 (42/7273) [0.004, 0.009] 1.20 X 10-1 0.003 (2/672) [1.07 X 10-4, 0.015] 6.35 X 10-2

-4 -7 0 -6 0 Rough O:z29:- 0.000 (0/22) 1.37 X 10 (1/7273) [3.44 X 10 , 0.001] 1.00 X 10 0.001 (1/672) [3.72 X 10 , 0.012] 1.00 X 10 Facility D -13 -14 Typhimurium V5- 1.000 (7/7) 0.015 (109/7273) [0.011, 0.019] 1.70 X 10 * 0.012 (8/672) [0.003, 0.029] 3.39 X 10 * a Frequency at the facility level is the number of samples positive for a particular S. enterica serotype (excluding the first positive sample) divided by the total number of samples positive for S. enterica in the facility (excluding the first positive sample) for the six-month sampling period. This strategy for calculating the frequency was employed because the statistical methods used assess conditional probabilities; that is, the probability that a sample is positive for a given serotype, given that the serotype was introduced into the facility. b Frequency of a serotype was calculated as the proportion of entries of that particular serotype in the reference distribution that was used. The reference

distribution of all S. enterica contained 7273 entries with a serotype designation (ALL SE; Supplemental File 10). The reference distribution consisting of S. enterica from environmental sources contained 672 entries (ENV SE; Supplemental File 11); (Number of entries of that serotype in the particular reference distribution/Total number of entries in the reference distribution); the 99.5% binomial confidence interval is provided for each frequency in square brackets. c P values are calculated from the binom.test function in R where facility level frequency for a serotype is compared with the frequency in the particular reference distribution of interest. Values with a * were significant after applying a Bonferroni correction to the α; P value for Facility B was 1.67 X 10-2 (0.05/3). P value for Facility C was 8.33 X 10-3 (0.05/6). P value for Facility D was 5.00 X 10-2 (0.05/1). 143

Texas Tech University, Alex Brandt, May 2014

TABLE 2.11 Facility-level persistence metrics for S. enterica, using a binomial test based on previous positive results as a risk factor

Number of Result Transitionsa Fisher’s Exact Test d Serotype ++ +- -+ -- Relative Risk [95% CI]b (Right One-Tail) Significance

Facility B Heidelberg 0 3 3 262 NC [NC, NC]d 1.00 X 100 Chailey 0 1 1 266 NC [NC, NC] 1.00 X 100

Liverpool 0 1 1 266 NC [NC, NC] 1.00 X 100 Facility C 0 III 50:r:z 0 0 8 267 NC [NC, NC] 1.00 X 10 Infantis 1 5 3 266 14.94 [1.805, 123.718] 8.49 X 10-2

Uganda 0 4 4 267 NC [NC, NC] 1.00 X 100

Anatum 0 1 2 272 NC [NC, NC] 1.00 X 100 Muenchen 0 2 2 271 NC [NC, NC] 1.00 X 100

0 Rough O:z29:- 0 1 1 273 NC [NC, NC] 1.00 X 10 Facility D

0 Typhimurium V5- 0 8 8 255 NC [NC, NC] 1.00 X 10 a Transitions from one status (positive or negative) to another (positive or negative) were recorded at every site for a specific S. enterica serotype. For instance, ++ indicates a current positive result following a previous positive result, while -+ indicates a current positive result following a previous negative result. b Relative risk was calculated using the MedCalc online Relative Risk Calculator and represents the probability of a current positive result given a previous positive result divided by the probability of a current positive result given a previous negative result. A 95% confidence interval calculated by the MedCalc online Relative Risk Calculator is also provided. c P values are calculated from the fisher.test function in R using the right one-tailed test option. Values with a * were significant after applying a Bonferroni correction to the α; P value for Facility B was 1.67 X 10-2 (0.05/3). P value for Facility C was 8.33 X 10-3 (0.05/6). P value for Facility D was 5.00 X 10-2 (0.05/1). d Abbreviation for “not calculated.” Values were not able to be calculated due to a value of 0 in cell A (corresponding to the ++ transition) of the 2 X 2 table.

144

Texas Tech University, Alex Brandt, May 2014

APPENDIX D APPENDIX C APPENDIX D

APPENDIX D

FIG 2.1 Sampling scheme map for Facility A

145

Texas Tech University, Alex Brandt, May 2014

FIG 2.2 Sampling scheme map for Facility B

146

Texas Tech University, Alex Brandt, May 2014

FIG 2.3 Sampling scheme map for Facility C

147

Texas Tech University, Alex Brandt, May 2014

FIG 2.4 Sampling scheme map for Facility D

148

Texas Tech University, Alex Brandt, May 2014

I 50.00% 45.00% 40.00% 35.00% 30.00% 25.00% L.L. monocytogenes monocytogenes 20.00% Prevalence 15.00% NonNon-Pathogenic Listeria Listeria

10.00% S.S. entericaenterica 5.00% E.E. coli coli O157:H7O157:H7 0.00%

Sampling Date

II 50.00% 45.00% 40.00% 35.00% 30.00% 25.00% L.L. monocytogenes monocytogenes 20.00% Prevalence 15.00% NonNon-Pathogenic Listeria Listeria

10.00% S.S. entericaenterica 5.00% E.E. coli coli O157:H7O157:H7 0.00%

Sampling Date

FIG 2.5 Prevalence of each organism by sampling date in ready-to-eat meat Facilities A (I) and B (II).

149

Texas Tech University, Alex Brandt, May 2014

Serotype Sampling Site Sampling Date Liverpool Site 28 07/12/2011 Chailey Site 16 09/09/2011 Heidelberg Site 35 09/09/2011 Heidelberg Site 39 09/30/2011

Heidelberg Site 46 10/28/2011

FIG 2.6 Dendrogram of XbaI PFGE patterns for S. enterica in Facility B. Dendrogram was produced using the Dice Correlation Coefficient with the Unpaired Groupwise Matching Algorithim of

BioNumerics version 6.6 set to 1.5% Optimization and 1.5% Tolerance. Traditional serotyping was performed by the National Veterinary Services Laboratories and serotypes were assigned according to the Kaufman, White, LeMinor Scheme.

150

Texas Tech University, Alex Brandt, May 2014

I 50.00% 45.00% 40.00% 35.00% 30.00% 25.00% L.L. monocytogenes monocytogenes 20.00%

Prevalence NonNon-Pathogenic Listeria Listeria 15.00% 10.00% S.S. entericaenterica

5.00% E.E. coli coli O157:H7O157:H7 0.00%

Sampling Date

II 50.00% 45.00% 40.00% 35.00% 30.00% 25.00% L.L. monocytogenes monocytogenes 20.00% Prevalence 15.00% NonNon-Pathogenic Listeria Listeria

10.00% S.S. entericaenterica 5.00% E.E. coli coli O157:H7O157:H7 0.00%

Sampling Date

FIG 2.7 Prevalence of each organism by sampling date in fresh meat Facilities C (I) and D (II)

151

Texas Tech University, Alex Brandt, May 2014

Serotype Sampling Site Sampling Date Muenchen Site 16 08/24/2011 Muenchen Site 27 08/24/2011 Infantis Site 17 06/03/2011 Infantis Site 26 06/03/2011 Infantis Site 26 07/05/2011 Infantis Site 19 08/24/2011 Infantis Site 36 08/24/2011 Infantis Site 39 08/24/2011

Rough R O:z29:- Site 39 09/23/2011

Uganda Site 17 08/24/2011 Uganda Site 24 08/24/2011 Uganda Site 26 08/24/2011

Uganda Site 40 08/24/2011

Anatum Site 25 08/24/2011 Anatum Site 50 11/16/2011

B III 50:r:z Site 26 11/16/2011 III 50:r:z Site 27 11/16/2011

III 50:r:z Site 29 11/16/2011

III 50:r:z Site 33 11/16/2011

III 50:r:z Site 34 11/16/2011 III 50:r:z Site 39 11/16/2011

III 50:r:z Site 40 11/16/2011

III 50:r:z Site 42 11/16/2011

FIG 2.8 Dendrogram of XbaI PFGE patterns for S. enterica in Facility C. Dendrogram was produced using the Dice Correlation Coefficient with the Unpaired Groupwise Matching Algorithim of BioNumerics version 6.6 set to 1.5% Optimization and 1.5% Tolerance. Traditional serotyping was performed by the National Veterinary Services Laboratories and serotypes were assigned according to the Kaufman, White, LeMinor Scheme.

152

Texas Tech University, Alex Brandt, May 2014

Sampling Site Sampling Date Site 21 10/21/2011 Site 26 10/21/2011 Site 34 07/05/2011 Site 45 07/05/2011

FIG 2.9 Dendrogram of XbaI PFGE Patterns for E. coli O157:H7 in Facility C. Dendrogram was produced using the Dice Correlation Coefficient with the Unpaired Groupwise Matching Algorithim of BioNumerics version 6.6 set to 1.5% Optimization and 1.5% Tolerance.

153

Texas Tech University, Alex Brandt, May 2014

Serotypeb Sampling Site Sampling Date

Typhimurium Var 5- Site 18 09/12/2011

Typhimurium Var 5- Site 36 09/12/2011

Typhimurium Var 5- Site 37 09/12/2011

Typhimurium Var 5- Site 38 09/12/2011

Typhimurium Var 5- Site 39 09/12/2011

Typhimurium Var 5- Site 41 09/12/2011

Typhimurium Var 5- Site 42 09/12/2011

Typhimurium Var 5- Site 44 09/12/2011

FIG 2.10 Dendrogram of XbaI PFGE Patterns for S. enterica in Facility D. Dendrogram was produced using the Dice Correlation Coefficient with the Unpaired Groupwise Matching Algorithim of BioNumerics version 6.6 set to 1.5% Optimization and 1.5% Tolerance. Traditional serotyping was performed by the National Veterinary Services Laboratories and serotypes were assigned according to the Kaufman, White, LeMinor Scheme.

154

Texas Tech University, Alex Brandt, May 2014

CHAPTER III ASSESSING FOODBORNE PATHOGEN PERSISTENCE MITIGATION IN MEAT PROCESSING ESTABLISHMENTS FOLLOWING ENACTMENT OF PHYSICAL AND PERSONNEL-BASED CONTROL MEASURES

Introduction Investigations of recent multi-state outbreaks of foodborne listeriosis and salmonellosis implicated long-term persistence of the outbreak strains within food processing facilities as a key factor leading to product contamination (15, 18, 22). Very little data exists on the persistence of Salmonella enterica and Escherichia coli O157:H7 in food processing environments. However, persistence of Listeria monocytogenes in food processing establishments is a well-recognized hazard (6, 7), and is a topic which has drawn considerable regulatory attention over the past several decades (26). Mitigating L. monocytogenes contamination requires an effective Listeria control program that often entails a multi-pronged strategy of i) raw and ready-to-eat (RTE) product separation, ii) recognizing entry points and harborage sites, iii) sanitary design and maintenance of equipment, iv) effective facility sanitation measures, v) employee personal hygiene and training, and vi) environmental, food contact surface, and food product testing programs (24, 25). However, stringent Listeria control programs still may not fully eliminate persistent strains that can colonize a processing facility for years at a time (16, 17, 19). Several investigators that probed L. monocytogenes ecology in food processing facilities used results they acquired from longitudinal sampling to devise intervention strategies to control the pathogen, but had mixed outcomes. In one study, an L. monocytogenes eradication program, which used hot steam, hot air, and hot water to clean and sanitize environmental sites and equipment, appeared to be successful at eliminating L. monocytogenes due to the fact that no samples were positive for the pathogen during the five month sampling period that followed the eradication program (2). Other studies, conducted in smoked fish and pork processing facilities, also had success in reducing L. monocytogenes prevalence after control measures (e.g. intensified cleaning and sanitation procedures, employee training, and minor equipment redesign) were implemented (9, 13, 20). However, likely due to challenges with overall sanitary 155

Texas Tech University, Alex Brandt, May 2014 design of the facility and equipment, several persistent strains remained present in these facilities even after the control measures were implemented and prevalence was reduced. Furthermore, in yet another instance, very little effect on Listeria prevalence was observed after implementation of trainings and interventions in two crawfish processing operations (10). Collectively, these outcomes highlight that overcoming Listeria contamination is a formidable challenge, and that implementation of control measures is not a complete guarantee that prevalence will be reduced, nor that persistence will be eliminated. It is also noteworthy to mention that most of the above-mentioned studies were conducted in RTE seafood operations. Therefore, much remains to be understood about the relationship between the implementation of control measures (such as revised cleaning and sanitation procedures, improved employee knowledge, and physical changes associated with the facility environment) and the impact on prevalence and persistence of Listeria in other types of operations, including those that process meat and poultry. Thus, the primary objective of this study was to use data generated from a six- month longitudinal assessment of pathogen persistence in four meat production facilities, including two RTE and two fresh meat processing operations, to develop training modules and control measures, and to use targeted follow-up sampling to assess any changes in prevalence and persistence that followed their implementation. The control measures assessed for their collective impact on Listeria prevalence and persistence were i) employee knowledge changes as a result of in-plant trainings, ii) employee behavioral modifications, and iii) physical facility changes. Directional sampling strategies were also enacted to assess distribution of contamination in areas adjacent to sites recurrently contaminated with Listeria, and a unique opportunity was taken to sample a newly- constructed area of one processing facility in order to observe whether persistent strains from older areas of the facility colonized the newer area.

Materials and Methods Facility Enrollment at Various Stages of the Study Four meat production facilities (two predominantly RTE and two predominantly fresh slaughter and cutting) were enrolled in this study. All four facilities had previously undergone a six-month longitudinal environmental sampling program to assess the 156

Texas Tech University, Alex Brandt, May 2014 prevalence and strain persistence of L. monocytogenes, non-pathogenic Listeria species, S. enterica, and E. coli O157:H7. In-plant training sessions covering foodborne pathogen persistence and facility-specific results were conducted after the longitudinal study. After in-plant trainings, three of the facilities (A, B, and C) were selected to continue in a follow-up intervention and environmental sampling program that was focused on mitigating pathogen persistence. These three facilities were selected because each of them had at least one location that was defined as a site of persistence for an L. monocytogenes strain during the longitudinal study period. Facility D was excluded from this group because it did not have a site where an L. monocytogenes strain persisted. Also, none of the facilities had a site where a strain of S. enterica or E. coli O157:H7 persisted; thus no further focus was devoted to controlling S. enterica or E. coli O157:H7 in any of the facilities. Furthermore, follow-up sampling schemes were limited to sites that produced at least one positive result for either L. monocytogenes or non-pathogenic Listeria species over the course of the initial longitudinal study. Because several deficiencies in employee hygienic practices and behaviors (e.g. improper attire, eating and drinking in the facility) were noted for Facilities B and C during the longitudinal study, changes in personnel behavior were also tracked in the two facilities throughout the follow-up intervention and sampling period to assess impact of in-plant trainings. In-Plant Food Safety Training Sessions and Knowledge Assessments The in-plant food safety training sessions were held for all personnel in each facility within six months of the conclusion of the initial longitudinal study period. Training curricula and PowerPoint handouts (Supplemental Files 12-15) were developed in collaboration with food safety extension personnel at Colorado State University who provided review and suggestions for editing on all materials. Portions of the handouts that were used for all facilities were bilingual (English and Spanish) and were designed to include project-pertinent information on: i) general food safety, ii) specific attributes of the three pathogens tracked in the initial longitudinal study, iii) good manufacturing practices, and iv) sites that commonly harbor pathogens in food processing facilities and specific practices to manage them. Facility-specific portions of training handouts covered: i) facility-specific sampling and molecular subtyping results, and ii) facility- specific suggestions for interventions. Subtyping results were provided on sampling site 157

Texas Tech University, Alex Brandt, May 2014 maps using color codes for each subtype in order to provide their spatial representation in each facility (Supplemental Files 16-19). The trainings also included a handwashing demonstration aided by the use of Glo Germ (Glo Germ Company, Moab, UT), and were followed by an English-only question-answer period for employees, managers and trainers to discuss current facility issues and concerns. Bilingual (English and Spanish) fact sheets (Supplemental File 20) on general food safety and each of the three pathogens were also developed by project personnel, reviewed by extension personnel, and copies were provided to facility managers to use to refresh knowledge after the training session. A bilingual pre- and post-training questionnaire with an anonymous matched number code for each respondent was administered to measure changes in knowledge. Knowledge assessments were also reviewed by extension personnel and edits were made based on suggestions. Because the focus was only on general improvements in knowledge among the entire population of personnel in each facility, and not on groups with specific characteristics, no demographic information was collected. Scores were assigned with no penalty for blank answers, one point per correct answer, and minus one point for a wrong answer, with partial credit given for multiple choice questions. Also, questions 15, 16, and 19 were eliminated post-hoc due to misleading wording. All methodology involving human subjects in this study was submitted for review by the Texas Tech University Institutional Review Board (Application 503360). After the initial submission of the project proposal, exemption from further review was granted. Targeted Post-Training Follow-Up Sampling and Subtyping for Listeria Follow-up sampling for Listeria in the three facilities enrolled in the follow-up intervention and environmental sampling program was conducted at pre-operational (pre- op) and mid-shift stages, with two occasions corresponding to each stage. Table 3.1 provides dates and time intervals post-training for the different sampling occasions. On each 2012 follow-up sampling occasion for Facility A, a subset of 20 sites that were positive at least once for Listeria during the initial longitudinal study was sampled (Fig 3.1). Site 10, a pooled site, which during the longitudinal sampling period included both the north and south drains located in the stuffing area, was divided into two sites, 10.1 and 10.2, corresponding to each drain, respectively, during the follow-up sampling period. This was done to assess the individual contamination patterns at each drain during 158

Texas Tech University, Alex Brandt, May 2014 the follow-up period. However, the two sites were still treated as a combined site for comparison of initial and follow-up prevalence data in order to be consistent with other sites included in the prevalence analysis. Additionally, six raw meat swab samples were collected in Facility A during the November 2012 mid-shift follow-up sampling occasion to assess incoming raw material contamination. Thirty-five sites that were Listeria- positive at least once during the initial study period were sampled in Facility B on each 2012 follow-up sampling occasion (Fig. 3.2), and 37 sites were sampled in Facility C on each 2012 sampling occasion (Fig. 3.3). For each facility on each occasion in 2013 all original sites used in the initial longitudinal period were sampled, except for: i) food samples, ii) sites in the off-site retail area of Facility A, and iii) those sites that were no longer present due to changes implemented in the facilities (Figs. 3.4-3.6). Sponge samples were collected from the sites and transported back to the laboratory as described in Chapter II. Procedures for isolation and preservation of L. monocytogenes and non-pathogenic Listeria species, subtyping of L. monocytogenes via EcoRI ribotyping and molecular serogrouping, subtyping of non-pathogenic Listeria species by sigB allelic typing, and deposition of sampling and subtyping data into Food Microbe Tracker were also performed as previously described in Chapter II. Facility Physical Changes and Intervention Implementation Interventions and physical changes that were implemented by the three facilities included in the follow-up intervention and environmental sampling program were recorded and grouped together based on whether they were suggested in trainings or were autonomously chosen by the facilities. They were also further grouped together based on the time point of their implementation: i) after training and before the first mid-shift follow-up sampling occasion and ii) between the first mid-shift follow-up sampling occasion and the second mid-shift follow-up sampling occasion. Furthermore, changes suggested in trainings, but not instituted by the facilities, were also noted. Self-Reported Behavioral Changes and Direct Observation For Facilities B and C, eight behaviors (e.g. use of facility-designated footwear, use of clean aprons for cleaning and sanitation, etc.) were targeted for the follow-up behavioral assessments. During each facility’s 2012 mid-shift follow-up sampling, paper questionnaires were administered to personnel asking them to self-report their pre- and 159

Texas Tech University, Alex Brandt, May 2014 post-training practices with regard to the certain behaviors targeted. Answer choices were “YES” and “NO,” and comment areas were provided. Project personnel also directly observed facility personnel for at least 2 h during the sampling occasion to note whether self-reported behaviors reflected actual behaviors. Observer handouts listed the targeted behaviors, and the observer was to circle “complete compliance” if all personnel were properly following the targeted behavior, “partial compliance” if only some of the personnel were properly following the targeted behavior, and “no compliance” if no personnel were properly following the targeted behavior. The same procedures were carried out during the November 2013 mid-shift follow up sampling with the exception that personnel were not asked about pre-training behaviors, but rather only current behaviors. Questionnaires from 2013 were matched to 2012 questionnaires using a numbered code for each respondent. Direct observation was also performed using observer forms similar to those used during the 2012 observation. Directional Sampling in Sites of Recurrent Contamination Select sites in Facilities A, B, and C, including those that were persistently colonized with certain subtypes of Listeria, were recurrently positive for Listeria over the initial six-month sampling period but had no persistent subtype, or were positive for Listeria during the initial pre-op follow-up sampling in 2012, were selected for directional sampling during the mid-shift follow-up sampling in 2012 to determine the distribution of contamination in areas directly adjacent to the original sample site. Sites included for directional sampling are noted in the 2012 follow-up sampling schematics (Figs. 3.1-3.3). Areas of approximately 2 ft X 2 ft in each cardinal direction (north, south, east, and west) in relation to the original site were sponge-sampled as previously described, and all directional samples were collected at the same time that the original site area was sampled for the 2012 mid-shift follow-up sampling occasion. All directional sponge samples were processed for Listeria isolation as described in Chapter II. Isolates were also subjected to the Listeria subtyping methods previously described in Chapter II, and subtyping info was deposited in Food Microbe Tracker as previously described. Sampling of New Sites in Smokehouse and Cooler Area of Facility B A new smokehouse room, cooler area, and inedible storage area were constructed in Facility B in 2012, providing a unique opportunity to observe whether persistent 160

Texas Tech University, Alex Brandt, May 2014

Listeria strains from older areas of the facility also colonized the newly constructed areas. Ten sites (designated 56-65) in this new area (including trench drains, cooler walls, drains in the cooler and inedible storage area, and areas around the newly-installed smokehouse) (Fig. 3.5) were sampled during the November 2012 mid-shift follow-up sampling occasion and again in both the November 2013 pre-op and mid-shift follow-up sampling occasions. Sponge samples of sites were collected, shipped, and processed for Listeria as described previously in Chapter II. Subtyping of L. monocytogenes and non- pathogenic Listeria isolates obtained also followed previously-described methods. Comparison of Listeria Prevalence from Initial and Follow-Up Periods Only the subset of sites in each facility that were i) positive at least once during the initial longitudinal sampling period and ii) not lost to facility physical changes at the last mid-shift sampling date, were included in the analysis of Listeria (L. monocytogenes and non-pathogenic Listeria species combined) prevalence between the initial and follow-up stages. Only mid-shift sampling occasions were included in the analysis. Additionally, during the 2013 mid-shift sampling, two sites (one in Facility A and one in Facility C) were positive for Listeria though they had never been positive during the initial longitudinal period. Because these positives did not originate from the subsets of sites indicated above, they were precluded from analysis. Initial and follow-up prevalences were also compared among a subset of processing room sites in Facilities B and C that were a part of these larger subsets described above. Statistical Analysis Reliability analysis on the knowledge assessment instrument (i.e. the degree of consistency with which it measured personnel knowledge) was performed by subjecting the pre-training knowledge scores for all individuals to the Split-Halves Method (1) with the two halves comprised of even- and odd-numbered questions. A Spearman-Brown Coefficient of Reliability (4, 23) was calculated in SPSS version 22.0 (IBM Corporation, Armonk, NY). Individual facility and overall mean pre- and post-training knowledge assessment scores were compared using the Paired t-Test option of the TTEST procedure in SAS version 9.3 (SAS Institute, Cary, NC) with α=0.05. Responses to self-reported behavior assessments were analyzed using a McNemar’s test in R version i386 2.15.1 (www.r-project.org) with α=0.05; McNemar’s test was chosen for analysis to account for 161

Texas Tech University, Alex Brandt, May 2014 the paired pre- and post- nature of the categorical yes/no responses for an individual. Changes in Listeria prevalence were assessed using repeated measures regression models (based either on binomial or Poisson distributions), with sample site as the repeated subject, in the GENMOD procedure of SAS. LSMEANS differences for initial and follow-up prevalence values were compared with α=0.05.

Results Knowledge Assessment Reliability and Pre- and Post-Training Outcomes The Spearman-Brown Coefficient of Reliability calculated for the knowledge assessment instrument was 0.453. While coefficients reflecting modest reliability (in the range of 0.50 to 0.60) are generally acceptable for research purposes and in cases where erroneous initial decisions can easily be corrected (e.g. through personnel training sessions) (1), the value calculated was lower than this acceptable range. This indicates that the consistency of its measurement of personnel knowledge was not as optimal as desired. However, it should be noted that this was the first instance in which this assessment instrument was used, and future users of the instrument can take this low value into consideration when developing revisions that will help improve its reliability. Despite the lower-than-desired reliability of the instrument, increases in mean knowledge assessment scores were observed between the pre-training time point and post-training time point in each facility (Fig. 3.7). The largest increase was observed for Facility B, where the mean score increased by 14.86%. The increase in mean assessment scores for all other facilities ranged from 7.75-9.06%. When combining all test scores into an overall mean for pre- and post-training time points across all facilities, an increase of 9.76% was observed. Paired t-Test analysis indicated that the increases in means were significant (P < 0.05) for Facilities A, B, and D, as well as the overall mean for all facilities. The mean increase for Facility C was not significant (P = 0.1633) though it had the highest mean score of all facilities at both time points. Implementation of Facility Physical Changes and Interventions Between in-plant training sessions and the first mid-shift follow-up sampling in Fall 2012, all three facilities enrolled in the follow-up intervention and environmental sampling program implemented several physical changes and interventions that were 162

Texas Tech University, Alex Brandt, May 2014 either recommended during the in-plant trainings or were autonomously implemented by the facility managers. Facility A implemented a total of six changes, four that had been suggested during the training, and two that were autonomously implemented by the facility manager (Table 3.2). Facility B also had six changes implemented, four that were suggested during trainings, and two that were autonomously implemented by the facility manager (Table 3.3). Facility C implemented the most extensive amount of changes in the interval between the in-plant training and the 2012 mid-shift follow-up sampling with 12 changes recorded (Table 3.4). All changes implemented by Facility C were suggested in the training and no autonomous changes had been implemented before the September 2012 mid-shift sampling. Between the 2012 mid-shift sampling and the 2013 follow-up samplings, Facility C underwent a renovation project and autonomously implemented numerous changes that were not suggested during the in-plant training session. During this period, the facility implemented 19 changes, two that were suggested during the in-plant training session, and 17 that were autonomously implemented by the facility managers (Table 3.4). During this same period, Facility B began using a newly-constructed smokehouse room and ready-to-eat cooler area, and implemented nine new changes, four that were suggested during the in-plant training session and five that were autonomously implemented (Table 3.3). Facility A implemented three autonomous changes during this period (Table 3.2). Facility A implemented all changes suggested during the in-plant training by the end of the study (Table 3.2), while Facility B chose not to implement 11 suggested changes (Table 3.3), and Facility C chose not to implement eight suggested changes (Tables 3.4). Self-Reported Behavioral Changes and Direct Observation Behavior self-reporting in Facility B indicated that only one significant (P = 0.0077) behavioral change was adopted during the six months between the in-plant training session and the 2012 mid-shift follow up sampling (Table 3.5). This change entailed the use of hygienic items such as hairnets and gloves. However, when direct observation was performed, only partial compliance for the behavior was observed among the personnel (Table 3.5). The seven other behaviors discussed with the personnel during the training session had not been significantly changed, and when personnel were directly observed, none of these targeted behaviors were being followed with complete 163

Texas Tech University, Alex Brandt, May 2014 compliance (Table 3.5). Assessment of long-term behavior changes at the 18-month time point revealed that hygienic item use was still the only significantly-changed behavior from the original pre-training baseline (P = 0.0133). Although direct observation revealed that some behaviors had improved in status from no compliance to partial compliance over the 12-month interval, none had improved to complete compliance (Table 3.5). In Facility C only two of the eight targeted behaviors were significantly changed between the pre-training baseline and 2012 mid-shift follow-up sampling time point: i) use of facility-designated footwear in place of street shoes (P = 0.0412) and ii) regular cleaning and sanitizing of boots throughout the day (P = 0.0412) (Table 3.6). Regular cleaning and sanitation of boots was likely facilitated by the installation of boot baths in the facility. Also, complete compliance was noted for these two behaviors during direct observation of facility personnel, while all other targeted behaviors had either complete or partial compliance noted for them (Table 3.6). At the 19-month follow-up point, the same two behaviors were significantly changed compared to the baseline with P = 0.0233 and P = 0.0412, respectively, while no other behaviors demonstrated significant changes. Direct observation also revealed that employees were still in complete compliance with regard to the use of facility-designated footwear, but that regular cleaning and sanitation of footwear had dropped from complete compliance in 2012 to partial compliance (Table 3.6). Several other behaviors with complete compliance status during the 2012 observation also had partial compliance status during the 2013 observation (Table 3.6). Listeria Contamination Patterns During Follow-Up Sampling The initial and follow-up period contamination patterns for Facility A sites with at least one positive Listeria result during the initial period are summarized in Table 3.7. During the Facility A September 2012 pre-op follow-up sampling, only Site 16 was positive for L. monocytogenes, and was contaminated with a DUP-1052A ribotype strain. This strain was persistent based on the analysis described in Chapter II. Isolates of Listeria welshimeri sigB allelic type (AT) 69, another persistent Listeria strain, were also collected during this sampling occasion at four sites. Three of these sites were floor and drain sites, and one was a non-food contact environmental site. During the November 2012 mid-shift follow-up sampling, both persistent Listeria strains (DUP-1052A and AT 69) were isolated again, along with an L. welshimeri AT 89 164

Texas Tech University, Alex Brandt, May 2014 strain not previously observed. As in the pre-op sampling occasion, L. monocytogenes DUP-1052A was found at Site 16. Additionally, the DUP-1052A strain was found at Sites 10, 11 and 12, all of which had been positive for the subtype at least once during the initial period (Table 3.7). The AT 69 strain was again isolated from two sites that were positive during the September pre-operational sampling occasion: Sites 3 and 10. Both L. monocytogenes DUP-1052A and L. welshimeri AT 69 strains were isolated again one year later during the Facility A 2013 pre-op and mid-shift follow-up samplings (Table 3.7). The DUP-1052A strain was only isolated from one site during the pre-op sampling, Site 45, while AT 69 was found at two sites, Sites 10 and 11. Several hours later, during the mid-shift sampling, both Sites 10 and 11 were again positive for AT 69. The L. monocytogenes DUP-1052A stain was detected mid-shift at Sites 10 and 11, as well as at Site 44, and L. welshimeri AT 69 was also found in a cluster of samples in the grinding room area at Sites 4, 5, and 6. Two L. monocytogenes strains not previously isolated in Facility A (of ribotypes DUP-1062D and DUP-18042) were isolated mid-shift at Sites 16 and 13, respectively. Site 1 was also positive for the DUP- 1062D strain. However, Site 1 was not included in the subset of sites used for calculation of Listeria prevalence, and thus the result is not included in the follow-up prevalence measure. Other persistent strains of Facility A observed in the initial longitudinal study, L. monocytogenes 116-239-S-2 and Listeria innocua AT 56, were absent from samples collected during the entire follow-up sampling period. Table 3.8 contains initial and follow-up period results for Facility B sites that were Listeria positive at least once during the initial longitudinal period. During the September 2012 pre-op follow-up sampling, all eight L. monocytogenes isolates belonged to either ribotype DUP-1062B (seven isolates) or ribotype DUP-1062E (one isolate). Both strains had persisted during the initial longitudinal period. Seven of the eleven non- pathogenic Listeria species isolates belonged to persistent L. innocua strain AT 6, while two belonged to AT 45, a previously-isolated, but non-persistent, strain. Two isolates of L. innocua AT 108, a strain not previously isolated from Facility B, were also collected. During mid-shift sampling in November 2012, all twelve L. monocytogenes isolates were of the persistent DUP-1062B strain, and all but two originated from drain and floor sites in the facility (Table 3.8). Four sites that were positive for the DUP-1062B

165

Texas Tech University, Alex Brandt, May 2014 strain on this occasion had also been positive during the September pre-op sampling occasion. L. innocua strains AT 6 and 45 were also isolated again, with four and three isolates apiece, and all but one of these isolates were obtained from drain and floor sites. The AT 108 strain was isolated again, and two strains not previously observed in Facility B were isolated: L. innocua AT 70, and Listeria seeligeri AT 24. During the November 2013 pre-op sampling, three of the four L. monocytogenes isolates were of the persistent DUP-1062B strain, and one was of the persistent DUP- 1062E strain (Table 3.8). L. innocua AT 45 was isolated again along with an L. innocua strain of a novel AT, which is herein referred to as LI AT NEW A. The mid-shift sampling yielded similar results with only two L. monocytogenes isolates, one of the persistent DUP-1062B strain, and one of the persistent DUP-1062E strain. The DUP- 1062B isolate was from Site 35, which was also positive for the DUP-1062B strain during pre-op. The DUP-1062E-positive site, Site 46, was in very close proximity to Site 45, which was positive for the DUP-1062E strain during pre-op. L. innocua AT 6 and 70 strains were also isolated mid-shift along with the LI AT NEW A strain. Several Facility B strains that persisted during the initial longitudinal period were absent during follow-up sampling. These included L. innocua AT 11, 23, and 38 strains. The results for initial and follow-up period Listeria contamination patterns for Facility C sites positive at least once during the initial longitudinal period are shown in Table 3.8. During the July 2012 pre-op follow-up sampling, four sites were positive for the L. monocytogenes DUP-1042B strain that was predominant and persistent in the initial longitudinal period (Table 3.8). Two were drain and floor sites, one was a non- food contact environmental site, and one sample came from employee kill floor boots. An L. innocua AT 70 strain, which was also prevalent and persistent during the initial period, was isolated from two drain and floor sites. The September 2012 mid-shift sampling produced two isolates of L. monocytogenes. One isolate was of the DUP-1042B strain and the other was of a DUP- 1057B strain, which was sporadically isolated during the initial longitudinal period. Of the six non-pathogenic Listeria species isolates, four were of the persistent AT 70 strain, while the others were L. innocua AT 23 and 109 strains. Both of the latter were previously, but only sporadically, isolated from Facility C.

166

Texas Tech University, Alex Brandt, May 2014

During the November 2013 pre-op sampling, only one L. monocytogenes isolate was obtained, and belonged to ribotype DUP-1046A. The DUP-1046A strain had not been previously isolated in Facility C. L. innocua strains of ATs 71 (four isolates), 70, 31, and 91 (one isolate each) were also isolated. AT 71 was persistent in the initial sampling period, while AT 91 was a newly-observed strain. Site 45, which had not previously produced a Listeria isolate, was contaminated with an L. innocua AT 30 strain. However, as for Site 1 contaminated with DUP-1062D in Facility A, this site was not included in the subset of sites for prevalence calculations, and the data point was excluded. The samples collected several hours later at the mid-shift stage produced one L monocytogenes DUP-1042B isolate, and six L. innocua isolates belonging to ATs 31, 70, 71, and 91. Site 17 was contaminated with the AT 70 strain at the pre-op stage, and was also positive for the strain during this mid-shift sampling round. The site was also positive for AT 70 during the mid-shift sampling one year earlier. A similar scenario was seen for AT 71 at Site 40, which was positive for the strain during both the pre-op and mid-shift rounds of sampling. Unlike Facilities A and B, all strains that persisted during the initial longitudinal period in Facility C were still isolated during follow-up. Outcomes of Directional Sampling at Sites of Recurrent Listeria Contamination Directional sampling was performed at eight sites in Facility A (Fig. 3.8), 13 sites in Facility B (Fig. 3.9), and five sites in Facility C (Fig. 3.10). In 12 instances, both the original sample area and all four directional samples were either all negative or all positive for Listeria. In 14 instances, at least one of the five sample areas had a different Listeria presence/absence result from the rest. Furthermore, strains found in the original sample area were frequently different from those found in adjacent directional samples. In Facility A (Fig. 3.8), the original sample areas at Sites 8 and 25 were negative for Listeria along with all corresponding directional areas. Sites 10.1 and 11 were positive at the original site as well as at all adjacent sites. Sites 16 and 3 were positive at the original site but had only one and two directional samples that yielded positive results, respectively. Sites 14 and 10.2 were both negative at the original sample site, but had one and two directional samples that were positive, respectively. In Sites 3 and 16, the strains found at directional sites (AT 69 and DUP-1052A, respectively) matched those found in the original site area. In Site 10.2, L. welshimeri AT 167

Texas Tech University, Alex Brandt, May 2014

69 was found in both the north and east directional samples. DUP-1052A was the only strain found in the west directional sample for Site 14. Site 10.1 harbored DUP-1052A at the original site, as well as in all directions, but the AT 69 strain was also found in the south and west directional samples. Site 11 was positive for the DUP-1052A and AT 69 strains at the original area, and in three directional areas (north, east, and south). However, at the Site 11 west directional site, only the AT 69 strain was isolated. In Facility B (Fig. 3.9), only Site 1 was negative at the original site area and in all directional areas. Sites 16, 17, 35, 39, and 42 were positive at the original site area as well as in all directional sites. Sites 10 and 11 were negative at the original site area, but had at least one positive directional sample. Sites 5, 18, 20, 45, and 46 were all positive at the original site, but had at least one directional sample that was negative for Listeria. The set of strains present at the original site areas never matched that of all four of their corresponding directional areas. For example, at Site 5, L. monocytogenes DUP- 1062B and L. innocua AT 45 were present at the original site area, but only AT 45 was found at the three positive directional sites (south, east, and west). At Site 17, the DUP- 1062B strain was found in the original area and all four directional samples, but only directional samples had non-pathogenic Listeria species. All four of these isolates were of a different subtype (ATs 45, NEW A, 6, and 108). At Site 35, all five areas were positive for DUP-1062B. However, non-pathogenic Listeria from each area belonged to different ATs, except for the north and east samples, which both belonged to AT 45. In Facility C (Fig 3.10), Site 15 was negative at the original site area and in all directional areas. Site 39 was positive at the original site area and in all corresponding directional areas. Only Site 26 had positive directional areas, while the original site was negative for Listeria. Sites 19 and 27 were positive at the original site area, but had at least one negative directional sample. As in Facility B, no original site areas harbored the same set of strains as all of their corresponding adjacent directional areas. At Site 19, L. innocua AT 70 was isolated at the original site area, and in the south and west directional areas, but L. monocytogenes DUP-1042B was also isolated from the south sample. At Site 39, both L. monocytogenes DUP-1057B and L. innocua AT 70 were isolated from the original site, the east site, and the west site. However, the north site was positive for

168

Texas Tech University, Alex Brandt, May 2014 only DUP-1057B. The south site was still positive for the AT 70 strain, but was positive for L. monocytogenes DUP-1042B instead of DUP-1057B. Listeria Contamination in Newly Constructed Areas of Facility B The first round of sample collection in the newly-constructed smokehouse, cooler, and inedible storage areas of Facility B area took place during the November 2012 mid- shift follow-up sampling (Table 3.10). During this occasion, five sites were positive for L. monocytogenes. Interestingly, all isolates belonged to ribotype DUP-1062B, which persisted in the originally-sampled areas of the facility during the initial longitudinal period. Non-pathogenic Listeria isolates collected from Sites 61 and 65 belonged to a strain of L. innocua AT 6 (which also persisted during the initial longitudinal period), and a strain of L. welshimeri AT 133, which had not been previously isolated. During the 2013 pre-op and mid-shift collections, L. monocytogenes DUP-1062B and L. innocua AT 6 were again found in this new area, but were also joined by L. monocytogenes DUP-1053E and DUP-1062E strains (two other strains that persisted during the initial longitudinal period), L. innocua AT 26 (a previously-isolated but non- persistent strain), and L. innocua AT 22 (a strain not previously observed in Facility B). Among the sites positive for DUP-1062B during pre-op sampling, only Site 64 was still positive for the strain mid-shift. Sites 56, 59, and 60 were contaminated with the DUP- 1062B strain during pre-op, but were negative for Listeria mid-shift. Site 61 was positive for DUP-1062B at pre-op, but was contaminated with the DUP-1053E strain mid-shift. Interestingly, at Site 62, different L. monocytogenes and L. innocua strains were isolated pre-op (DUP-1053E and AT 22) compared to mid-shift (DUP-1062B and AT 26). Comparison of Initial and Follow-Up Stage Listeria Prevalence In Facilities B and C, the mid-shift prevalences for Listeria among the sites that were: i) positive for Listeria at least once during the initial longitudinal period, and ii) were followed for the entirety of the study, were 47.8% and 49.0% during the initial longitudinal period, respectively (Table 3.11). These values decreased to 27.9% and 20.0%, respectively, during the follow-up sampling period (Table 3.11). Though Facility B prevalence on the November 2012 mid-shift sampling (44.1%) was similar to the overall prevalence of the initial longitudinal period (47.8%), the November 2013 mid- shift prevalence of 11.8% was the lowest prevalence ever recorded for this subset of 169

Texas Tech University, Alex Brandt, May 2014

Facility B sites. Using a repeated measures regression model to compare prevalences, the follow-up prevalence LSMEANS value was significantly (P < 0.0001) different than the initial prevalence LSMEANS value in Facilities B and C (Table 3.11). Facility A presented a slightly different situation where the Listeria prevalence among the subset of sites positive for Listeria during the initial period and followed throughout the entirety of the study was 46.5% during the initial longitudinal period, but remained at 36.8% during the follow-up period (Table 3.11). Though this represented a decrease in prevalence, when analyzed via the repeated measures regression model, the follow-up prevalence LSMEANS value did not significantly differ from the initial prevalence LSMEANS value (P = 0.2687) (Table 3.11). Listeria were mostly absent during the follow-up sampling period among nine processing room sites of Facility C that were followed for the entire study. Site 15 was the only processing room site that produced a follow-up Listeria-positive sample, and did so during the July 2012 pre-op sampling. Though simultaneous implementation of control measures within the April 2012-September 2012 period prevents analysis of individual control measure effects on Listeria prevalence, it is tempting to hypothesize that two key changes made in this area during follow- up (installation of boot baths and use of facility- designated footwear) may have led to the Listeria mid-shift prevalence drop among these sites from 48.1% to 0.0% (Table 3.11). This was still observed even though compliance dropped among personnel from complete compliance in 2012 to partial compliance in 2013. Because the 0.0% follow-up prevalence prevented convergence of a binomial model, a repeated measures model based on a Poisson distribution was used to compare prevalence LSMEANS values for these nine sites, and the difference between initial and follow-up prevalence LSMEANS values was significant (P < 0.0001) (Table 3.11). A similar situation was observed in Facility B where the mid-shift prevalence of Listeria among 22 sites, which were: i) always located in the processing room, and ii) followed through the entirety of the study, was 42.4% during the initial sampling period, and averaged to 25.0% during the follow-up sampling period (Table 3.11). This lower prevalence was mainly due to the 4.5% point prevalence observed among these sites during the 2013 mid-shift sampling. Similar to Facility C, boot baths were installed and facility-specific footwear began to be used during the time interval before this drop in

170

Texas Tech University, Alex Brandt, May 2014 processing room site prevalence was observed in November 2013. Again, though this decrease is unable to be definitively attributed to these practices, it is interesting to note the association. Initial and follow-up prevalence LSMEANS values were significantly different (P = 0.0040) from one another for these sites (Table 3.11).

Discussion In-Plant Training Sessions Promoted Knowledge and Behavioral Changes and Instigated the Implementation of Facility Physical Changes and Interventions Overall, this study demonstrates the utility of in-plant training sessions, which included results from combined testing and molecular subtyping, as a means to increase personnel knowledge and instigate behavioral and physical changes in a facility. Similar outcomes were observed in a previous study performed by our group among a set of small and very small RTE meat production facilities (27). In that study, the mean knowledge assessment score increased significantly (P < 0.0001) across all six facilities following in-plant training sessions. Though increases in assessment score means in this study were not drastic (ranging from 7.75 to 14.86%), and measures other than test scores may be more appropriate indicators of knowledge obtained, the outcomes still seem to suggest that in-plant trainings may have a positive impact on personnel knowledge. It was interesting that Facility C had the highest test score means at both time points. Although Facility C managers never conducted formal trainings with employees, the small number of full-time employees in Facility C seemed to promote a large degree of interaction between employees and managers, which was evident during the training session and during all sampling visits to the facility. The effects of such informal interactions, and the overall food safety climate of a facility (11), may be interesting to investigate in future studies to better understand how these existing cultural aspects among personnel may influence the knowledge they possess and retain. In-plant training sessions also seemed to instigate changes in personnel behavior in Facilities B and C. While self-reporting demonstrated an increase among Facility B personnel in their use of hygienic items such as gloves and hairnets, both at the short- term (6-month) and long-term (18-month) time points, this apparent behavior change was not completely adopted by all employees based on direct observation. This discrepancy 171

Texas Tech University, Alex Brandt, May 2014 seems to indicate that although self-reporting may indicate a behavior change, the behavior may not actually be modified. Observations such as these stress the importance of direct observation of employees when instituting changes in their daily routine. In Facility C, personnel self-reported changes with regard to their use of i) facility- designated footwear and ii) cleaning and sanitation of footwear, at both the 5-month and 19-month time points, and complete compliance for these behaviors was verified via direct observation. It was only during the 2013 (19-month) direct observation that employees were not in complete compliance regarding the sanitation of footwear. Suggestions for improvements provided to facility managers during the in-plant training sessions also instigated facility changes and interventions based on the four, eight, and 14 suggested physical changes or interventions implemented by Facilities A, B, and C, respectively, over the course of the follow-up period. Additionally, this seemed to instigate, at least partially, five, seven, and 16 autonomously-adopted changes in Facilities A, B, and C, respectively. This observation suggests that an initial impetus, such as an in-plant training and related discussion, may initiate a chain effect in some facilities whereby other issues are highlighted. The facility manager may then feel compelled to make other changes to improve the overall food safety measures in the facility. The reasons that managers adopted these autonomous changes (and did not adopt certain suggested changes) were not documented in this study, but warrant consideration in future studies, since aspects such as financial considerations, ease of implementation, and applicability to the facility process flow are all likely to have an influence on whether facility managers choose to implement certain changes in the facility. Follow-Up Sampling Contamination Patterns Were Characterized by Decreases in Listeria Prevalence but Incomplete Elimination of Persistent Strains Among all sites that: i) were positive for Listeria during the initial six months of sampling, and ii) were followed through the entire course of the study in Facilities A, B, and C, decreases in Listeria prevalence were observed between the initial and follow-up sampling periods, and these decreases were significant in Facilities B and C. Though simultaneous implementation of the various control measures makes it difficult to assess each measure’s individual contribution to decreases in Listeria prevalence, the collective Listeria reduction effect of these measures is still noteworthy. Similar results have also 172

Texas Tech University, Alex Brandt, May 2014 been seen in other studies. For instance, Lappi et al. observed a similar effect on Listeria prevalence after the implementation of control measures in four RTE smoked fish plants (9). Two-year monitoring of Listeria contamination patterns with enactment of control measures (such as employee trainings and targeted sanitation procedures) between Years 1 and 2 revealed a decrease in Listeria prevalence among environmental sites from 26.1% in the first year to 19.5% in the second year. Though less pronounced, a decrease in Listeria prevalence among finished product samples was also observed. Likewise, Malley et al. used data from four months of Listeria sampling in two RTE smoked fish facilities as guidelines for development of facility-specific interventions, and observed borderline significant monthly decreases in Listeria prevalence (P = 0.026 and 0.076) during the occasions that followed (13). Ortiz et al. investigated Listeria contamination patterns in an Iberian pork processing facility for three years, and while they did not notice a substantial decrease in Listeria prevalence in environmental and equipment sites after the enactment of improved cleaning and sanitation measures and raw product controls in Year 2, they did notice a decrease in Listeria prevalence among raw materials and products from 29% in Year 1 to 9% in Year 3 (21). Collectively, the results of these studies, along with results observed for this study, reinforce the impact that new control measures can have with regard to reducing Listeria prevalence. Yet, despite these Listeria prevalence decreases, several L. monocytogenes and non-pathogenic Listeria species strains that persisted during the initial longitudinal period were still isolated during the follow-up sampling period. In Facility A, the persistent L. monocytogenes DUP-1052A and L. welshimeri AT 69 strains were isolated during follow-up. However, persistent L. monocytogenes 116-239-S-2 and L. innocua AT 56 strains were not observed in 2012 or in 2013. In Facility B, all three persistent L. monocytogenes strains DUP-1062B, DUP-1062E, and DUP-1053E were isolated during follow-up. However, DUP-1053E was only isolated from new sites in the Facility B smokehouse room and cooler area. Of the persistent Facility B non-pathogenic Listeria species subtypes (ATs 6, 11, 23, and 38), only AT 6 was observed during follow-up. In Facility C, all three persistent subtypes identified during the initial longitudinal study, L. monocytogenes DUP-1042B, and L. innocua ATs 70 and 71, were isolated again during follow-up. Altogether, these observations highlight the extreme difficulty in fully

173

Texas Tech University, Alex Brandt, May 2014 eliminating persistent Listeria strains once established, even when stringent control measures are put into place within the facility. Similar difficulties in fully eliminating persistent Listeria strains were also seen in all three studies mentioned above. Despite decreased Listeria prevalence among environmental sites in Year 2 (after interventions), Lappi et al. noted that persistent L. monocytogenes strains remained present in three of the four facilities (9). Despite borderline significant monthly decreases in Listeria prevalence observed for both facilities sampled by Malley et al., persistent L. monocytogenes ribotypes DUP-1043A and DUP-1039C remained present in one RTE smoked seafood facility during the entire study, and persistent ribotype DUP-1042A remained present in another (13). Similarly, Ortiz et al. noted that three of eight persistent pulsed-field gel electrophoresis types of L. monocytogenes remained present in the pork processing facility they sampled, even though decreases in prevalence were noted between the initial and final study years (20). In Facility A, both sites of L. monocytogenes persistence identified during the initial longitudinal period (Sites 3 and 10) were persistently colonized by the L. monocytogenes 116-239-S-2 strain. However, the strain was not found in Facility A at any point during the follow-up sampling period, and it is tempting to hypothesize that control measures led to the elimination of its persistence. Sites 3 and 10 are both drains. The facility manager installed new floor foamers at different entry points, and increased spray volume on existing foamers in other areas of the facility. These floor-based interventions could have contributed to the elimination of the strain in these drain sites. Though not fully eliminated, L. welshimeri AT 69 was never found again at Site 25, also a drain, where it persisted for a portion of the initial longitudinal period. The absence of the subtype at this site also could have been due to the floor foamer-based change. However, AT 69 was isolated at Site 11 (stress mats) on all four follow-up samplings, and at Sites 3 and 10 (both drains) on at least two follow-up sampling occasions. The inability to eliminate the persistent L. monocytogenes DUP-1052A strain, which persisted in Facility A from 2007 to 2013 (28), despite the implementation of additional control measures, adequate employee hygiene, and adequate sanitation, is a prime example of the difficulty in eliminating persistent strains. Though raw products may in some cases constitute a source of repeated introduction of L. monocytogenes into

174

Texas Tech University, Alex Brandt, May 2014 a facility, this contribution was investigated by collecting six sponge surface samples of raw meat during the 2012 mid-shift follow-up sampling. Neither the DUP-1052A strain, nor any other Listeria strain, was found among these samples. However, the facility uses natural casings for the sausage it produces, and these casings were not sampled during the course of this study or by Williams et al. (28). L. monocytogenes has been isolated from the intestines of swine (8) and previous work has implicated casings as a source of L. monocytogenes in the sausage production process (14). Also, Sites 10 and 11 (both near the stuffer), harbored the strain only at mid-shift time points (and not pre-op). Thus, handling casings during operations may have caused the strain to appear mid-shift. The only persistent site of L. monocytogenes identified for Facility B was Site 44, which was persistently colonized with L. monocytogenes ribotype DUP-1062E. The site was never positive for DUP-1062E during the follow-up sampling period, demonstrating some mitigation of persistence at the site. Sites 10 and 12, which had been sites of persistence for L. innocua ATs 6 and 23, respectively, during the initial sampling period, were never positive for the strains during the follow-up period. Additionally, certain sites, which were recurrently positive for Listeria, but were not persistently colonized by a specific strain (i.e. Sites 17, 39, and 42), were not positive during either 2013 sampling occasion. While control measures may have mitigated contamination at these sites, other sites, such as Site 35, were positive for Listeria for the entirety of the study (including the 2013 sampling occasions) despite the implementation of control measures. In Facility C, among the six sites of persistence that were identified for the L. monocytogenes DUP-1042B strain during the initial sampling period, only Sites 15 and 19 were positive for the strain during the follow-up sampling period. Indeed, Sites 26, 27, 39, and 48 were never positive for the subtype during the follow-up period (though DUP- 1042B was found in an adjacent directional site of Site 39; Fig. 3.10), likely due to some effect of the enacted control measures. Also, Sites 19 and 26, both sites of persistence for L. innocua AT 70 during the initial sampling period, were positive for the strain on either one or both 2012 follow-up sampling occasions. Site 18, the smokehouse, which was also a site of AT 70 persistence, was not positive during the 2012 follow-up samplings, and was removed from the facility in the time interval between September 2012 and November 2013.

175

Texas Tech University, Alex Brandt, May 2014

Taken together, these results reinforce the impact that control measures can have on reducing the overall load of Listeria contamination in a facility, but also demonstrate the great challenge that is associated with full elimination of strain persistence both facility-wide and at certain harborage sites. Even more stringent control measures may completely eliminate Listeria contamination, as has been demonstrated before (2), but the results of this study and others seem to indicate that efforts may need to be diverted from a focus of fully eradicating persistent strains in all environmental sites of a facility, and more toward reducing the risk of their transfer to food contact sites and finished products. Listeria Contamination Patterns at Defined Sampling Sites Differed From Those of Immediately Adjacent Areas The results from directional sampling in each facility demonstrated that the contamination patterns at a particular sampling site may not always be completely reflective of that present in immediately adjacent sites. Furthermore, some sites selected for testing, which appear to be sources of contamination, may actually represent collection points at which the contamination gathers. Malley et al. also performed vector swabbing during months 5, 7 and 8 of their study as part of an adaptive sampling strategy to identify origins of L. monocytogenes contamination in sites of two RTE seafood facilities (13). Similar to the results of this study, the authors found that in most occasions at least one vector swab had a different strain than the one found at the original site it was associated with. Yet, through scrutiny of vector swabbing results, especially at one trench drain that was consistently contaminated with L. monocytogenes DUP-1042A, the authors discovered that the bottom of a nearby sliding door was the harborage site of the persistent strain, and was the site to target with a control strategy. This demonstrates the particular utility of directional sampling in proverbial “seek-and-destroy” strategies that are often used to identify and eliminate Listeria harborage sites (5). While the results of directional swabbing were not used to aid intervention implementation in this study, the results do provide for some interesting hypotheses for future investigation. In Facility A, Sites 10.1 and 11 were in close proximity to one another with Site 11 located to the north of Site 10.1. The overlap of the L. monocytogenes DUP-1052A and L. welshimeri AT 69 subtypes observed from the directional sponge samples seemed to indicate that either Site 11, the stress mats, was the original site of contamination, or 176

Texas Tech University, Alex Brandt, May 2014 that some site above the two sites, i.e. the stuffing room table (Site 12), was the original site of contamination. A similar situation could be hypothesized for Sites 39 and 42 in Facility B where Site 39 was generally located to the north and east of Site 42. Site 42 north and east directional swab results closely matched the contamination profile found at the Site 39 original site. Refrigeration units, which dripped condensation, were located above Site 39, and this condensation was drained into Site 42. While not investigated further, the units, rather than Sites 39 and 42 could have been the actual source of the Listeria strains frequently isolated from these sites. In Site 39 of Facility C, another cooler drain, the contamination was also spread in the various directions around the drain. Condensation from an overhead refrigeration unit was also collected in this drain. Thus, a refrigeration unit again could have been the actual source of contamination, and should be considered during future work that is done with the facility. Newly-Constructed Areas in a Facility May Become Colonized By Strains Which Persist in Older Areas of the Facility All three persistent Facility B L. monocytogenes strains (identified during the initial longitudinal sampling period) were isolated in the newly-constructed smokehouse room and cooler area of the facility during follow-up sampling. This observation implies that they were somehow transferred into these newly-constructed areas from the older areas of the facility. While previous studies have suggested that a number of sources of cross-contamination, such as incoming raw materials, employees, supplies, machines, tools, and construction may all be contributing sources for the introduction of Listeria strains into previously uninhabited areas of processing facilities, definitive conclusions cannot be formed about the means by which these strains were introduced into these new areas of Facility B since routes of Listeria contamination were not comprehensively sampled in this study. However, potential routes are still worthwhile to discuss. A slightly similar situation was documented in a study performed by Berrang et al. where the Listeria colonization of a newly-constructed chicken further processing facility was followed over a 21-month period (3). While environmental sites in the facility were completely free of Listeria before commencement of processing operations, the floor drains of the facility were colonized within four months by several subtypes of L. monocytogenes. Though several potential sources of contamination were investigated 177

Texas Tech University, Alex Brandt, May 2014

(including raw products, personnel, and incoming fresh air), a persistent strain found on eight different occasions in four different drains was only isolated from incoming raw product. Thus, raw product was deemed the likely route of entry. Such a scenario could have also been the case for Facility B, with raw product as the primary mode of transfer. Yet, even though raw product is one potential route of entry, the transfer of equipment into this area could have also promoted the colonization of these new sites. In a study conducted by Lundén et al. the appearance of a persistent L. monocytogenes strain in three different cooked meat production plants was chronologically linked to the transfer of a dicing machine from one plant to the others (12). After moving the machine from Plant A (where L. monocytogenes pulsotype I was initially found) to Plant B, the strain persistently contaminated the Plant B dicing line. Furthermore, when the machine was moved to Plant C, the same phenomenon occurred. Overall, this stressed that equipment can have a role in introduction of persistent strains into a new environment. Yet other components of the Facility B environment, such as personnel, packaging materials, and smokehouse wood shavings, cannot be completely ruled out. Future investigations of similar situations should take approaches like that used by Berrang et al. where comprehensive sampling of all incoming materials and the surrounding environment is conducted to obtain a more clear perspective by which the pathogens are able to enter into a previously uninhabited processing environment.

Conclusions The presence and persistence of foodborne pathogens within a meat processing facility is likely influenced by a complex interplay between a host of factors (e.g. facility and equipment design, production processes, types of products manufactured, employees, sanitation practices, etc.). Accounting for the individual influence of each of these factors is a formidable challenge when assessing the impact of intervention implementation, and fully attributing any changes in pathogen prevalence and persistence to a defined set of interventions is difficult. However, the outcomes of this study seem to suggest that the in- plant trainings, personnel behavioral changes, and physical facility changes that were implemented during the over 1.5 year course of the follow-up intervention and sampling

178

Texas Tech University, Alex Brandt, May 2014 period of this study had at least some part in the reduction of Listeria prevalence in two of the three facilities that were included in this follow-up program. The outcomes of this study also seem to suggest that there is utility in including testing and molecular subtyping results from longitudinal environmental sampling when developing focused in-plant training sessions. These focused training sessions seemed to instigate (at least in part) personnel behavioral changes and facility physical changes and interventions, which may have ultimately helped to mitigate Listeria contamination. However, persistent strains of Listeria, which were first identified during the initial longitudinal period, were still isolated during follow-up in each facility. Also, several sites in Facility C where specific Listeria strains persisted during the initial longitudinal period still produced positive results for the persistent strains on multiple occasions throughout follow-up. Collectively, these results indicate that although new control measures may provide some contribution in mitigating Listeria contamination, the full elimination of persistent strains remains an elusive challenge to the processor. Additionally, directional sampling results discussed in this study demonstrated that contamination profiles at certain sampling sites may not match those found in adjacent sites. In these cases, directional sampling results should be scrutinized in order to decipher patterns which may indicate strain harborage in neighboring sites. Also, results obtained from sampling a newly-constructed area in Facility B demonstrated that persistent strains present in older areas of the facility were transferred to newly- constructed sectors. Future studies that investigate Listeria colonization of previously- uninhabited facility sectors should comprehensively sample an array of potential routes by which the organism is transferred into the newly-constructed area including, but not limited to, raw products, equipment, employees, tools, supplies, and other miscellaneous items (i.e. smokehouse wood chips and liquid smoke). Information garnered from these studies may provide better insight on more effective measures to prevent L. monocytogenes entry and colonization of new facilities. Also, because persistence is difficult to eliminate once established, preventing this initial entry may be crucial to mitigating Listeria persistence over the course of the facility’s life span, and thus may help mitigate contamination of food products that are produced by the establishment.

179

Texas Tech University, Alex Brandt, May 2014

This study also further highlights the need for facility managers to adopt effective and aggressive environmental monitoring programs for foodborne pathogens. While they can be expensive to maintain for small and very small facilities with limited resources, these monitoring programs are critical components in one’s array of tools for assessing the risk that environmental contamination poses to the products produced by the establishment. They also can help identify issues in cleaning and sanitation before they become seriously deficient and begin to pose an immediate threat to product safety. Thus, if a facility’s financial situation at least allows some environmental sampling to be conducted (regardless of how small the number of samples is), it should be considered. Also, the educational materials developed for this study, including the training handouts, facts sheets, and knowledge assessment instruments may be useful for future extension programming, especially if revisions are made to the knowledge assessment to improve its reliability. These materials would be especially valuable for education of other small and very small meat processors who need information about best practices to effectively and efficiently counteract pathogen persistence amidst financial constraints. Thus, efforts will be made to make all materials developed for this project publicly available via the world wide web.

References 1. Ary D, Cheser Jacobs L, Ashgar R, Sorensen C. 2006. Introduction to research in education, 7th ed. Thomson Wadsworth, Belmont, CA.

2. Autio T, Hielm S, Miettinen M, Sjoberg AM, Aarnisalo K, Bjorkroth J, Mattila-Sandholm T, Korkeala H. 1999. Sources of Listeria monocytogenes contamination in a cold-smoked rainbow trout processing plant detected by pulsed-field gel electrophoresis typing. Appl. Environ. Microbiol. 65:150-155.

3. Berrang ME, Meinersmann RJ, Frank JF, Ladely SR. 2010. Colonization of a newly constructed commercial chicken further processing plant with Listeria monocytogenes. J. Food Prot. 73:286-291.

4. Brown W. 1910. Some experimental results in the correlation of mental abilities. Brit. J. Psychol. 3:296-322.

180

Texas Tech University, Alex Brandt, May 2014

5. Butts J. 2003. Seek & destroy: identifying and controlling Listeria monocytogenes growth niches. Available At: http://www.foodsafetymagazine.c om / magazine-archive1 / aprilmay-2003 / seek-destroy-identifying-and-controllin g-listeria-monocytogenes-growth-niches. Accessed On: 2/13/14

6. Carpentier B, Cerf O. 2011. Review - persistence of Listeria monocytogenes in food industry equipment and premises. Int. J. Food Microbiol. 145:1-8.

7. Ferreira V, Wiedmann M, Teixeira P, Stasiewicz MJ. 2014. Listeria monocytogenes persistence in food-associated environments: epidemiology, strain characteristics, and implications for public health. J. Food Prot. 77:150-170.

8. Kanuganti SR, Wesley IV, Reddy PG, McKean J, Hurd HS. 2002. Detection of Listeria monocytogenes in pigs and pork. J. Food Prot. 65:1470-1474.

9. Lappi VR, Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Longitudinal studies on Listeria in smoked fish plants: impact of intervention strategies on contamination patterns. J. Food Prot. 67:2500-2514.

10. Lappi VR, Thimothe J, Walker J, Bell J, Gall K, Moody MW, Wiedmann M. 2004. Impact of intervention strategies on Listeria contamination patterns in crawfish processing plants: a longitudinal study. J. Food Prot. 67:1163-1169.

11. Lemons L, Brashears T, Hartzog A, Echeverry A, Thompson L, Miller M, Garcia L, Brashears M. 2012. A comparison of food safety climate at municipal and private beef slaughter plants in Mexico, International Association for Food Protection 2012 Annual Meeting, Poster P1-146, Providence, RI.

12. Lundén JM, Autio TJ, Korkeala HJ. 2002. Transfer of persistent Listeria monocytogenes contamination between food-processing plants associated with a dicing machine. J. Food Prot. 65:1129-1133.

13. Malley TJ, Stasiewicz MJ, Grohn YT, Roof S, Warchocki S, Nightingale K, Wiedmann M. 2013. Implementation of statistical tools to support identification and management of persistent Listeria monocytogenes contamination in smoked fish processing plants. J. Food Prot. 76:796-811.

14. Martin B, Garriga M, Aymerich T. 2011. Prevalence of Salmonella spp. and Listeria monocytogenes at small-scale Spanish factories producing traditional fermented sausages. J. Food Prot. 74:812-815.

15. Mead PS, Dunne EF, Graves L, Wiedmann M, Patrick M, Hunter S, Salehi E, Mostashari F, Craig A, Mshar P, Bannerman T, Sauders BD, Hayes P,

181

Texas Tech University, Alex Brandt, May 2014

Dewitt W, Sparling P, Griffin P, Morse D, Slutsker L, Swaminathan B. 2006. Nationwide outbreak of listeriosis due to contaminated meat. Epidemiol. Infect. 134:744-751.

16. Miettinen M. 1999. Characterization of Listeria monocytogenes from an ice cream plant by serotyping and pulsed-field gel electrophoresis. Int. J. Food Microbiol. 46:187-192.

17. Nesbakken T, Kapperud G, Caugant DA. 1996. Pathways of Listeria monocytogenes contamination in the meat processing industry. Int. J. Food Microbiol. 31:161-171.

18. Olsen SJ, Patrick M, Hunter SB, Reddy V, Kornstein L, MacKenzie WR, Lane K, Bidol S, Stoltman GA, Frye DM, Lee I, Hurd S, Jones TF, LaPorte TN, Dewitt W, Graves L, Wiedmann M, Schoonmaker-Bopp DJ, Huang AJ, Vincent C, Bugenhagen A, Corby J, Carloni ER, Holcomb ME, Woron RF, Zansky SM, Dowdle G, Smith F, Ahrabi-Fard S, Ong AR, Tucker N, Hynes NA, Mead P. 2005. Multistate outbreak of Listeria monocytogenes infection linked to delicatessen turkey meat. Clin. Infect. Dis. 40:962-967.

19. Orsi RH, Borowsky ML, Lauer P, Young SK, Nusbaum C, Galagan JE, Birren BW, Ivy RA, Sun Q, Graves LM, Swaminathan B, Wiedmann M. 2008. Short-term genome evolution of Listeria monocytogenes in a non-controlled environment. BMC Genom. 9:539.

20. Ortiz S, López V, Martínez-Suárez JV. 2014. Control of Listeria monocytogenes contamination in an Iberian pork processing plant and selection of benzalkonium chloride-resistant strains. Food Microbiol. 39:81-88.

21. Ortiz S, Lopez V, Villatoro D, Lopez P, Davila JC, Martinez-Suarez JV. 2010. A 3-year surveillance of the genetic diversity and persistence of Listeria monocytogenes in an Iberian pig slaughterhouse and processing plant. Foodborne Pathog. Dis. 7:1177-1184.

22. Russo ET, Biggerstaff G, Hoekstra RM, Meyer S, Patel N, Miller B, Quick R. 2013. A recurrent, multistate outbreak of Salmonella serotype Agona infections associated with dry, unsweetened cereal consumption, United States, 2008. J. Food Prot. 76:227-230.

23. Spearman C. 1910. Correlation calculated from faulty data. Brit. J. Psychol. 3:271-295.

182

Texas Tech University, Alex Brandt, May 2014

24. Tompkin RB. 2002. Control of Listeria monocytogenes in the food-processing environment. J. Food Prot. 65:709-725.

25. Tompkin RB, Scott VN, Bernard DT, Sveum WH, Gombas KH. 1999. Guidelines to prevent post-processing contamination from Listeria monocytogenes. Dairy Food Environ. Sanit. 19:551-562.

26. U.S. Department of Agriculture, Food Safety and Inspection Service. 2003. Control of Listeria monocytogenes in ready-to-eat meat and poultry products; final rule. Fed. Reg. 68:34207-34254.

27. Williams SK. 2010. Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. M.S. Thesis: Department of Animal Sciences, Colorado State University, 168 p.

28. Williams SK, Roof S, Boyle EA, Burson D, Thippareddi H, Geornaras I, Sofos JN, Wiedmann M, Nightingale K. 2011. Molecular ecology of Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. J. Food Prot. 74:63-77.

183

Texas Tech University, Alex Brandt, May 2014

TABLE 3.1 Dates of follow-up sampling occasions in Facilities A, B, and C Sampling Occasion Date of Sampling Time Post-Training Facility A Pre-Operational 1 September 2012 4 months Mid-Shift 1 November 2012 6 months Pre-Operational 2 November 2013 18 months Mid-Shift 2 November 2013 18 months Facility B Pre-Operational 1 September 2012 4 months Mid-Shift 1 November 2012 6 months Pre-Operational 2 November 2013 18 months Mid-Shift 2 November 2013 18 months Facility C Pre-Operational 1 July 2012 3 months Mid-Shift 1 September 2012 5 months Pre-Operational 2 November 2013 19 months Mid-Shift 2 November 2013 19 months

184

Texas Tech University, Alex Brandt, May 2014

TABLE 3.2 Facility A physical change and intervention implementation Suggested Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling

 Created better systems to minimize traffic between raw, post-lethality-exposed, storage, and cooler areas. Dry casings and daily materials are retrieved at the beginning of the day prior to the start of production. This reduces the amount of times personnel walk through to get supplies.  In lieu of installing additional floor foamers in the raw to post-lethality-exposed transition area, turned up the volume on existing foamers to account for high traffic in the area.  Replaced old stress mats with new mats and instructed cleaning personnel to clean these regularly.  Instructed cleaning personnel to do a better job of cleaning the gap between the smokehouses during the weekly smokehouse cleaning session. Autonomous Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling  Installed additional floor foamers at entry points into the facility (four total) in order to replace boot baths that were problematic, especially for cart traffic.  Added a floor foamer in dry storage so that people walking past the offices towards the dock also have the bottoms of their shoes sanitized.

Suggested Changes Made Between 2012 and 2013 Mid -Shift Follow-Up Samplings

 None Recorded Autonomous Changes Made Between 2012 and 2013 Mid -Shift Follow-Up Samplings  Closed up doorway between the chopping room (raw) and packaging room (post-lethality- exposed) with a wall in order to eliminate traffic from one area to another.  Gap area between the smokehouses was also added to the list of post-lethality-exposed non-food contact areas that were swabbed randomly by the facility when doing weekly Listeria testing.  Implemented a requirement that employees either wear work-issued boots, or, if bringing in their own (or when taking the boots outside), that they are required to spend time at the floor foamers to prevent tracking contaminants into the facility.

Changes Suggested but Not Implemented  None Recorded

185

Texas Tech University, Alex Brandt, May 2014

TABLE 3.3 Facility B physical change and intervention implementation Suggested Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling

 Began providing hairnets for employees to wear and gloves for handling product.  Began changing out rubber blades on squeegees on a more frequent basis, and informed sanitation crew to do more frequent cleaning and sanitation of squeegees.  Began daily cleaning and sanitation of drains and Cooler 1 (carcass cooler) floor.  Instituted a no-smoking policy for employees (one kill floor employee would occasionally smoke while slaughtering).

Autonomous Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling

 Constructed a new portion of the facility with a dedicated room for smokehouses, a cooler for ready-to-eat product, and a new inedible bin storage area. However, in Fall 2012 the transition into this new area of the facility had just begun, and it was still partitioned from the rest of the facility with only a long and winding hallway through the retail food prep area and equipment room to link it to the rest of the facility.  Purchased a new smokehouse and had begun operating it in this new area of the facility.

Suggested Changes Made Between 2012 and 2013 Mid-Shift Follow-Up Samplings

 Installed plastic doors on east and west ends of the processing room to mitigate air flow from the retail area and the kill floor area.  Had employees begin wearing impermeable and more easily cleaned rubber boots instead of their less sanitary street shoes when working inside of the facility.  Installed footbaths with sanitizer solution at different entry and transition (slaughter to ready-to- eat) points of the processing room.  Improvement in raw-and-ready to eat product separation due to the new ready-to-eat cooler next to the new smokehouse room and away from the carcass cooler (Cooler 1). Some raw product was still being stored in this cooler; the facility manager mentioned that he recognized the problem and that improving this was a priority. Autonomous Changes Made Between 2012 and 2013 Mid -Shift Follow-Up Samplings

 Completed the transition into the newly constructed smokehouse room, ready-to-eat product cooler, and inedible storage area; moved one of the old smaller smokehouses into this area, and moved tumbling and injector equipment into the area.  Installed a new doorway between the existing processing room and the new smokehouse room and ready-to-eat cooler area.  Added refrigeration to the processing room and to the new smokehouse room; the facility manager mentioned that the smokehouse room was still not being efficiently refrigerated because of the heat generated by the smokehouses and that this strategy would have to be re-evaluated.  Installed new hot water system for the facility.  Started using lactic acid spray as an intervention on carcasses. Changes Suggested but Not Implemented

 Have employees that work in both slaughter and in fabrication/ready-to-eat wear separate pairs of boots designated specifically for each of those areas of the facility.  Layout a traffic flow plan for employees to follow when moving from one area of the plant to another. Also use this traffic flow for movement of equipment and carts.  Have employees wear different (and preferentially color-coded or marked) aprons that are only used for either raw or ready-to-eat processing.  Install a floor foamer in areas where cart traffic is high (i.e. the swinging doors from Cooler 2 to the processing area), and/or increase sanitation in this area to prevent transfer by cart wheels.

186

Texas Tech University, Alex Brandt, May 2014

TABLE 3.3, Cont. Facility B physical change and intervention implementation Changes Suggested but Not Implemented  Avoid having customers enter the processing area unless they are properly dressed for entering into the processing area (i.e. rubber boots and apron). If customer entry cannot be prevented at least have them enter from retail instead of the back where livestock enter and are held. Also keep them moving in the general direction from cleaner to dirtier areas.  Increase cleaning under and between the smokehouses.  Seal the crack in Cooler 2 floor and try to prevent water from pooling in the area next to the crack.  Avoid placing employee personal food and drinks under carcasses in the cooler or in ready-to-eat product storage areas.  Use electric or captive bolt stunning instead of rifles.  Use a hot water dip station (180°F for 15 sec) to intermittently sanitize knives, sharpeners, and scabbards while slaughtering.  Create and implement a sanitation plan for the freezer and freezer floor.

187

Texas Tech University, Alex Brandt, May 2014

TABLE 3.4 Facility C physical change and intervention implementation

Suggested Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling

 Provided impermeable and more easily cleaned rubber boots for employees to wear in place of less sanitary street shoes.  Installed footbaths with sanitizer at entry points and transition areas (i.e. kill floor to processing) of the processing room.  Provided separate pairs of boots for employees to wear that were designated for either work on the kill floor or work in the processing/cutting room.  Eliminated customer traffic coming from outdoors through the processing area.  Began using a foamer to apply cleaning detergent.  Began wearing clean aprons for cleaning and sanitation activities at the end of the work day.  Began picking up table tops to do scrubbing and rinsing and began scrubbing legs of tables.  Began using warm water in place of scalding hot water for initial cleaning.  Began using an approved quaternary ammonium sanitizer for sanitizing surfaces.  Began using brushes and low-pressure water to clean out drains.  Stopped cleaning offal barrels indoors with high pressure water.  Had plumbing unclogged to provide better drainage from Cooler 2.

Autonomous Changes Made After In-Plant Training and Prior to the 2012 Mid-Shift Follow-Up Sampling  None Recorded Suggested Changes Made Between 2012 and 2013 Mid -Shift Follow-Up Samplings  Installed three compartment sink for proper utensil sanitation.  Got rid of smokehouse that was problematic to clean around.

Autonomous Changes Made Between 2012 and 2013 Mid-Shift Follow-Up Samplings

 Quit pork slaughter and smoking and went exclusively to fresh beef cutting and grinding.  Provided smocks for employees to wear in addition to the aprons previously worn alone.  Completed an extensive facility expansion project: completed an entirely new addition with a new office area, break room, customer restroom, conference room, locker area, laundry room, milk crate storage area, hot water system, and a new hallway leading into the processing room east doorway (the processing room east doorway was previously open to the exterior of the facility and the only barrier was an air curtain apparatus above the door).  Knocked out old office walls to expand the area for processing, cutting and grinding; this renovation allowed almost all tables and equipment to be moved away from the walls.  Added a new customer-designated entry way on the east side of the freezer so that customers did not have to come inside of the processing room to obtain products from the freezer.  Added refrigeration to processing room.  Added new wall and ceiling material, and resurfaced floors in the processing room.  Resurfaced walls in Cooler 1 and added stainless steel footing to prevent chipping and facilitate cleaning.  Began regular sanitizer application for cooling units.  Purchased two eight foot boning tables with tall backs and large stainless steel decking to prevent poly tops from getting dings and scratches.  Purchased two new stainless steel tables with three tall sides to keep orders separated.  Purchased one new stainless steel table for weighing hamburger.  Purchased one new table for the patty machine.  Retrofitted old wrapping table so that the paper rolls and stamps are attached to table instead of to the walls.  All new equipment items were placed on casters so they could be moved around when cleaning. 188

Texas Tech University, Alex Brandt, May 2014

TABLE 3.4, Cont. Facility C physical change and intervention implementation Autonomous Changes Made Between 2012 and 2013 Mid -Shift Follow-Up Samplings  The setup of the new boning station system made it so that meat was placed in tubs, and bones and scraps placed in barrels, both directly next to the worker, which eliminated the need to throw any bones or material at a wall like that which was being previously done. Changes Suggested but Not Implemented  Replace bump caps with hard hats.  Remove tools and unused chairs from along walls to facilitate cleaning.  Implement a control method for flies.  Incorporate a lactic acid spray intervention while processing carcasses.  Use a hot water dip station (180°F for 15 sec) to intermittently sanitize knives, sharpeners, and scabbards while slaughtering and while doing cutting and fabrication.  Change how milk crates used for holding wrapped meat products are moved around and cleaned.  Employ a third party cleaning and sanitation crew (this was rectified by having an employee dedicated to deep cleaning and sanitation at the end of the day).  Use multiple sanitizers and a sanitizer rotation schedule.

189

Texas Tech University, Alex Brandt, May 2014

TABLE 3.5 Summary of self-reported behavior changes and direct observation of employees in Facility B Question Summarya 6 Month Significanceb 6 Month Observationc 18 Month Significanceb 18 Month Observationc

Do you wear the same pair of shoes/boots 0.4795 PARTIAL 0.2482 PARTIAL at work that you wear when you COMPLIANCE COMPLIANCE leave from home? Do you wear a separate pair of boots 1.000 NO NC PARTIAL when working in the kill floor area than COMPLIANCE COMPLIANCE when working in the processing area?

Do you wear hygienic items such as 0.0077* PARTIAL 0.0133* PARTIAL hairnets and gloves when working? COMPLIANCE COMPLIANCE

Do you wear an apron that is dirty from 1.000 NO 1.000 PARTIAL working with raw meat when you work COMPLIANCE COMPLIANCE with cooked meat? Do you follow proper handwashing 0.4795 NO 1.000 PARTIAL practices (using soap, a nailbrush, and COMPLIANCE COMPLIANCE rubbing hands for at least 20 sec?)

Do you wash your hands after shaking 0.1336 NO 1.000 PARTIAL hands with someone, handling something COMPLIANCE COMPLIANCE dirty, or helping at the cash register? Do you eat, smoke, or drink while 0.1336 PARTIAL 0.6171 PARTIAL working with meat? COMPLIANCE COMPLIANCE

Do you store your personal food or drinks 0.4795 PARTIAL 1.000 PARTIAL for lunch or break in the same cooler as COMPLIANCE COMPLIANCE the meat and other products?

190

Texas Tech University, Alex Brandt, May 2014

TABLE 3.5, Cont. Summary of self-reported behavior changes and direct observation of employees in Facility B Question Summarya 6 Month Significanceb 6 Month Observationc 18 Month Significanceb 18 Month Observationc Do you feel as if you are familiar with the 0.1336 N/A 0.2482 N/A harmful bacteria that can be found in foods? Do you often think of basic things you 0.2482 N/A 1.000 N/A can do to prevent contamination of the meats you work with? a A representative summary of the wording of the question that was posed to the respondent is provided. b Significance for changes in employee self-reported responses with regard to a certain behavior are designated with a P value calculated from a McNemar’s Test performed in R version i386 2.15.1. P values less than 0.05 were considered significant, and significance is indicated with a * next to the P value. NC signifies that the B and C cells in the McNemar’s Test 2X2 table both contained 0’s, and thus the table was precluded from analysis by McNemar’s Test. c Direct observations of employee behavior were performed during the mid-shift follow-up sampling occasions in November 2012 (6-month follow-up) and November 2013 (18-month follow-up). Behaviors reflected by the questions listed in this table were specifically targeted. If all employees were observed to be properly following the behavior, COMPLETE COMPLIANCE was noted. If some employees were observed to follow the behavior, while others were not, then PARTIAL COMPLIANCE was recorded. If no employees were found to be following the behavior in the proper way, then NO COMPLIANCE was

recorded. N/A was recorded in both situations for the last two behaviors since these are not able to be visually observed.

191

Texas Tech University, Alex Brandt, May 2014

TABLE 3.6 Summary of self-reported behavior changes and direct observation of employees in Facility C Question Summarya 5 Month Significanceb 5 Month Observationc 19 Month Significanceb 19 Month Observationc

Do you wear the same pair of shoes/boots 0.0412* COMPLETE 0.0233* COMPLETE at work that you wear when you COMPLIANCE COMPLIANCE leave from home? Do you wear a separate pair of boots 0.2482 COMPLETE 0.1336 PARTIAL when working in the kill floor area than COMPLIANCE COMPLIANCE when working in the processing area?

Do you regularly clean and sanitize your 0.0412* COMPLETE 0.0412* PARTIAL boots throughout the day? COMPLIANCE COMPLIANCE

Do you put on a clean apron and clean 0.2482 PARTIAL 1.000 PARTIAL yourself up before starting cleanup and COMPLIANCE COMPLIANCE sanitation duties at the end of the day? Do you follow proper handwashing 1.000 PARTIAL NC PARTIAL practices (using soap, a nailbrush, and COMPLIANCE COMPLIANCE rubbing hands for at least 20 sec?)

Do you wash your hands after shaking NC PARTIAL 1.000 PARTIAL hands with someone or handling COMPLIANCE COMPLIANCE something dirty?

Do you eat, smoke, or drink while NC PARTIAL 1.000 PARTIAL working with meat? COMPLIANCE COMPLIANCE

Do you use scalding hot water when NC COMPLETE 0.4795 COMPLETE cleaning your knives/utensils or your COMPLIANCE COMPLIANCE work area?

192

Texas Tech University, Alex Brandt, May 2014

TABLE 3.6, Cont.: Summary of self-reported behavior changes and direct observation of employees in Facility C Question Summarya 5 Month Significanceb 5 Month Observationc 19 Month Significanceb 19 Month Observationc Do you feel as if you are familiar with the 0.1336 N/A 0.1336 N/A harmful bacteria that can be found in foods? Do you often think of basic things you 1.000 N/A 1.000 N/A can do to prevent contamination of the meats you work with? a A representative summary of the wording of the question that was posed to the respondent is provided. b Significance for changes in employee self-reported responses with regard to a certain behavior are designated with a P value calculated from a McNemar’s Test performed in R version i386 2.15.1. P values less than 0.05 were considered significant, and significance is indicated with a * next to the P value. NC signifies that the B and C cells in the McNemar’s Test 2X2 table both contained 0’s, and thus the table was precluded from analysis by McNemar’s Test. c Direct observations of employee behavior were performed during the mid-shift follow-up sampling occasions in September 2012 (5-month follow-up) and November 2013 (19-month follow-up). Behaviors reflected by the questions listed in this table were specifically targeted. If all employees were observed to be properly following the behavior, COMPLETE COMPLIANCE was noted. If some employees were observed to follow the behavior, while others were not, then PARTIAL COMPLIANCE was recorded. If no employees were found to be following the behavior in the proper way, then NO COMPLIANCE was

recorded. N/A was recorded in both situations for the last two behaviors since these are not able to be visually observed.

193

Texas Tech University, Alex Brandt, May 2014

TABLE 3.7 Initial and follow-up Listeria contamination patterns in Facility Aa Date Sampledb Sample Site 05/27/2011 06/24/2011 08/16/2011 09/16/2011 10/14/2011 11/11/2011 09/07/2012 11/15/2012 11/14/2013 11/14/2013 Drain and Floor Sites LM 116-239-S-2 LM 116-239-S-2 LM 116-239-S-2 - - LW AT 27 LW AT 69 LW AT 69 - - Site 3 LW AT 69 LW AT 69 LM 116-239-S-2 LM 116-239-S-2 LM DUP-1052A LI AT 56 LM 116-239-S-2 - LW AT 69 LM DUP-1052A LW AT 69 LM DUP-1052A Site 10 LW AT 69 LI AT 56 LW AT 69 LM DUP-1052A LM DUP-1052A LM 116-239-S-2 LM DUP-1052A LM DUP-1052A - LM 116-239-S-2 LW AT 69 LW AT 69 LM DUP-1052A Site 11 LW AT 69 LI AT 56 LI AT 56 LI AT 56 LW AT 69 LW AT 69 Site 17 - - - LI AT 56 ------

Site 19 - LM DUP-1048A LM DUP-1052A LI AT 56 - LI AT 56 - - - - Site 21 - LI AT 56 LM DUP-1052A ------Site 25 - - - LW AT 69 LW AT 69 LW AT 69 - - - - LM DUP-1052A - LW AT 32 - - - - - Site 45 - LW AT 69 -

Non-Food Contact Environmental Sites LM DUP-1052A - - LM DUP-1052A - - - - Site 14 LM DUP-1052A LI AT 56 - Site 16 - - - LI AT 56 LM DUP-1052A - LM DUP-1052A LM DUP-1052A LW AT 69 LM DUP-18042

Site 44 - - - - - LW AT 69 LW AT 69 LW AT 89 - LM DUP-1052A Employee Surfaces and Employee Contact Sites LM DUP-1052A LM DUP-1052A LM 116-239-S-2 - LM DUP-1052A - - - LW AT 69 Site 5 LI AT 37 LI AT 56 - LM DUP-1052A Site 6 LM DUP-1052A - - LM DUP-1052A LW AT 27 - - - LW AT 69 LI AT 56 Site 37 - - - - LM DUP-1052A - - - - -

Food Contact Sites and Food Samples

Site 4 LI AT 37 - - - LM 116-239-S-2 - - - - LW AT 69 LM DUP-1052A LM DUP-1052A - - LM 116-239-S-2 - - - - Site 8 LI AT 56 -

194

Texas Tech University, Alex Brandt, May 2014

TABLE 3.7, Cont. Initial and follow-up Listeria contamination patterns in Facility Aa Date Sampledb Sample Site 05/27/2011 06/24/2011 08/16/2011 09/16/2011 10/14/2011 11/11/2011 09/07/2012 11/15/2012 11/14/2013 11/14/2013 Food Contact Sites and Food Samples LM DUP-1052A Site 12 LM DUP-1052A - LI AT 56 - - - LM DUP-1052A - - LI AT 56 Site 13 LM 116-239-S-2 - LW AT 19 LM DUP-1052A LM DUP-1052A - - - - LM DUP-1062D LM DUP-1052A ------Site 31 LI AT 56 - a A result of “-” corresponds to a negative result for that particular site on the respective sampling date. Samples positive for Listeria are described by two capital letters that correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), Listeria welshimeri (LW). The L. monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1052A) and the Listeria spp. sigB allelic type (e.g. AT 56) follow these species abbreviations. b All dates in 2011 correspond to an initial longitudinal sampling period. The September 2012 sampling date corresponds to a pre-operational follow-up sampling. The November 2012 sampling date corresponds to a mid-shift follow-up sampling. The last two samplings on 11/14/2013 were conducted on the same day but the first from left to right corresponds to a pre-operational follow up sampling, and the second corresponds to a mid-shift follow-up sampling, which occurred during operations.

195

Texas Tech University, Alex Brandt, May 2014

TABLE 3.8 Initial and follow-up Listeria contamination patterns in Facility Ba Date Sampledb Sample Site 06/10/2011 07/12/2011 09/09/2011 09/30/2011 10/28/2011 12/02/2011 09/07/2012 11/15/2012 11/13/2013 11/13/2013 Drain and Floor Sites Site 1 LM DUP-1062B LM DUP-1062E - - - - LM DUP-1062B - - LI AT NEW A LM DUP-1062B LM DUP-1062E LM DUP-1062B - - - LI AT 45 LI AT 45 LI AT NEW A - Site 5 LI AT 23 LI AT 23 LI AT 45 LM DUP-1062E LM DUP-1062B LI AT 38 LI AT 31 LI AT 38 LI AT 6 LI AT 6 - - - Site 16 LI AT 44 LI AT 108 LM DUP-1062B LM DUP-1062B LM DUP-1053E LM DUP-1062B LI AT 6 LI AT 11 LI AT 38 LM DUP-1062B - - Site 17 LI AT 94 LI AT 23 LI AT 17 LI AT 108 LM DUP-1062B LM DUP-1062B - - - LM DUP-1053E - LI AT 45 - - Site 20 LI AT 23 Site 23 - LM DUP-1062B - - - LI AT 11 - LM DUP-1062B - - LM DUP-1062E LM DUP-1062B ------Site 28 LI AT 6 LS AT 24 LM DUP-1062B Site 34 - LI AT 6 - - - - LM DUP-1062B - - LI AT 38 LM DUP-1062E LM DUP-1062E LM DUP-1053E LM DUP-1062B LM DUP-1062E LM DUP-1062B LM DUP-1062B LM DUP-1062B LI AT 45 LM DUP-1062B Site 35 LI AT 23 LI AT 23 LI AT 6 LI AT 11 LI AT 108 LI AT 6 LI AT 45 LI AT 6 LM DUP-1053E LM DUP-1042C LM DUP-1062B LM DUP-1062B LM DUP-1062B LI AT 38 LI AT 45 LI AT 6 - - Site 39 LI AT 6 LI AT 11 LI AT 11 LI AT 6 LI AT 6 LM DUP-1062E LM DUP-1062B LM DUP-1053E LM DUP-1062E LM DUP-1062B LM DUP-1062B LI AT 6 LI AT 11 - - Site 42 LI AT 38 LI AT 45 LI AT 45 LI AT 11 LI AT 6 LI AT 70 LM DUP-1062B LM DUP-1062B LI AT 26 LI AT 11 LM DUP-1062E LM DUP-1053E - LI AT 6 LM DUP-1062E LI AT 70 Site 45 LI AT 6 LI AT 6 LM DUP-1062B LM DUP-1053E LM DUP-1062B - LI AT 6 LM DUP-1062E LM DUP-1062B LI AT 6 - LM DUP-1062E Site 46 LI AT 11 LI AT 11 LI AT 6 Non-Food Contact Environmental Sites Site 2 LI AT 6 LM DUP-1053E ------LM DUP-1062E Site 10 LI AT 6 - LI AT 6 ------LI AT 6 Site 11 - LI AT 23 LI AT 6 LI AT 6 LI AT 6 - LI AT 6 - NS NS LM DUP-1062B LM DUP-1062B LI AT 23 LI AT 23 LM DUP-1053E - LI AT 6 LI AT 45 LM DUP-1062B - Site 12 LI AT 23 LM DUP-1062B LM DUP-1062B - - Site 18 LM DUP-1062E - - - - - LI AT 45

196

Texas Tech University, Alex Brandt, May 2014

TABLE 3.8, Cont. Initial and follow-up Listeria contamination patterns in Facility Ba Date Sampledb Sample Site 06/10/2011 07/12/2011 09/09/2011 09/30/2011 10/28/2011 12/02/2011 09/07/2012 11/15/2012 11/13/2013 11/13/2013 Non-Food Contact Environmental Sites LM DUP-1062E Site 19 LM DUP-1062E - LI AT 11 ------LI AT 109 Site 24 - - - - - LI AT 30 - - - -

Site 33 - - LI AT 11 ------

Site 49 LI AT 23 - LI AT 11 - - - LI AT 6 - - -

Site 51 - LI AT 37 ------Employee Surfaces and Employee Contact Sites Site 43 LI AT 23 - LI AT 11 LI AT 11 ------LM DUP-1062E LM DUP-1062E Site 44 LM DUP-1062E LI AT 11 - - - LM DUP-1062B - - LI AT 6 LI AT 31 Site 47 - LI AT 11 ------LM DUP-1062E LM DUP-1062B LI AT 23 NS - - - - LM DUP-1062B - Site 48 LI AT 23 LI AT 6 Food Contact Sites and Food Samples

Site 7 - - LI AT 11 - - LI AT 37 - - - -

Site 26 LM DUP-1062E LI AT 23 ------Site 27 - - - LI AT 11 ------Site 29 - LI AT 6 LI AT 11 - LI AT 11 - Site 31 - - - LI AT 11 - LI AT 37 - - - -

Site 32 - - - LI AT 11 ------

Site 37 - - LI AT 11 LM DUP-1053E ------

197

Texas Tech University, Alex Brandt, May 2014

TABLE 3.8, Cont. Initial and follow-up Listeria contamination patterns in Facility Ba

b Date Sampled Sample Site 06/10/2011 07/12/2011 09/09/2011 09/30/2011 10/28/2011 12/02/2011 09/07/2012 11/15/2012 11/13/2013 11/13/2013 Food Contact Sites and Food Samples Site 40 - - LM DUP-1062B LI AT 11 ------a A result of “-” corresponds to a negative result for that particular site on the respective sampling date. “NS” denotes that the sample site was not sampled on that particular sampling date. Samples positive for Listeria are described by two capital letters that correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI) and Listeria seeligeri (LS). The L. monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1062B) and the Listeria spp. sigB allelic type (e.g. AT 6) follow these species abbreviations. b All dates in 2011 correspond to an initial longitudinal sampling period. The September 2012 sampling date corresponds to a pre-operational follow-up sampling. The November 2012 sampling date corresponds to a mid-shift follow-up sampling. The last two samplings on 11/13/2013 were conducted on the same day but the first from left to right corresponds to a pre-operational follow up sampling, and the second corresponds to a mid-shift follow-up sampling, which occurred during operations.

198

Texas Tech University, Alex Brandt, May 2014

TABLE 3.9 Initial and follow-up Listeria contamination patterns in Facility Ca Date Sampledb Sample Site 06/03/2011 07/05/2011 08/24/2011 09/23/2011 10/21/2011 11/16/2011 07/14/2012 09/07/2012 11/15/2013 11/15/2013 Drain and Floor Sites LM DUP-1042B Site 5 LI AT 70 LI AT 71 LI AT 6 LI AT 6 LM DUP-1042B - - - - LI AT 6 Site 10 - LM DUP-1042B - - LI AT 70 - - - - - LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 6 - LM DUP-1042B - - - Site 15 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1057B LI AT 71 LI AT 71 - LI AT 70 LI AT 70 LI AT 70 Site 17 LI AT 70 LI AT 71 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 70 LI AT 26 LI AT 70 LI AT 31 LM DUP-1042B Site 19 LI AT 70 LI AT 70 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 53 LI AT 70 LI AT 70 - LI AT 91 LI AT 31 Site 26 LI AT 71 LI AT 70 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 70 LM DUP-1042B - LI AT 23 - - Site 27 LI AT 37 LI AT 26 LI AT 70 LI AT 70 - - LM DUP-1042B LM DUP-1042B LI AT 70 LI AT 70 - - LM DUP-1046A - Site 34 Site 38 - - - LI AT 109 - LI AT 70 - - - - LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1057B LI AT 109 - LI AT 71 LI AT 31 Site 39 LI AT 71 LI AT 23 LI AT 70 LI AT 70 LI AT 71 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 109 LI AT 71 LI AT 70 - LM DUP-1042B LI AT 71 LI AT 71 Site 40 LI AT 70 LI AT 26 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 109 - LI AT 70 - - - LI AT 91 Site 43 LI AT 70 HLI AT 141 LI AT 70 Non-Food Contact Environmental Sites Site 7 - - LI AT 70 ------LM DUP-1042B - - - LI AT 70 - - - NS NS Site 11 LI AT 70 LM DUP-1042B LM DUP-1042B - - LM DUP-1042B LI AT 70 - - - - Site 13 LI AT 70 LI AT 71 LM DUP-1042B LI AT 70 - - LI AT 70 LI AT 70 - LI AT 70 LI AT 71 - Site 16 LI AT 23 LM DUP-1042B LM DUP-1042B - - LI AT 70 LI AT 70 - - NS NS Site 18 LI AT 70 Site 20 - - LM DUP-1042B ------

199

Texas Tech University, Alex Brandt, May 2014

TABLE 3.9, Cont. Initial and follow-up Listeria contamination patterns in Facility Ca

b Date Sampled Sample Site 06/03/2011 07/05/2011 08/24/2011 09/23/2011 10/21/2011 11/16/2011 07/14/2012 09/07/2012 11/15/2013 11/15/2013 Non-Food Contact Environmental Sites Site 21 LM DUP-1057B ------LI AT 71

Site 28 - - - LI AT 70 ------

Site 29 - LI AT 30 ------LM DUP-1042B ------Site 30 LI AT 70 Site 32 LI AT 11 - - LI AT 71 - - - - LI AT 71 - Site 36 - LM DUP-1042B LM DUP-1042B - - - LM DUP-1042B LI AT 109 - - Site 44 - - LI AT 70 ------

LM DUP-1042B LI AT 70 - LM DUP-1042B ------Site 47 LI AT 71

Site 50 LI AT 31 - - LI AT 70 - LI AT 70 - - - -

Site 54 - - LI AT 6 ------

Employee Surfaces and Employee Contact Sites

LM DUP-1042B - LI AT 71 LI AT 26 LI AT 70 - LM DUP-1042B - - - Site 22 LI AT 70 Site 31 - - - LW AT NEW ------Site 33 - - LI AT 71 ------Site 42 LI AT 30 ------

LM DUP-1042B LM DUP-1042B LI AT 70 LM DUP-1042B - LM DUP-1042B - - - - Site 48 LI AT 70

Food Contact Sites and Food Samples

Site 1 - - LI AT 23 ------

Site 2 - - LI AT 23 ------

200

Texas Tech University, Alex Brandt, May 2014

TABLE 3.9, Cont. Initial and follow-up Listeria contamination patterns in Facility Ca

b Date Sampled Sample Site 06/03/2011 07/05/2011 08/24/2011 09/23/2011 10/21/2011 11/16/2011 07/14/2012 09/07/2012 11/15/2013 11/15/2013 Food Contact Sites and Food Samples Site 8 - - LI AT 23 ------

Site 24 - - LI AT 26 - - LI AT 71 - - - -

Site 55 - - LI AT 30 - LI AT 37 - NS NS NS NS a A result of “-” corresponds to a negative result for that particular site on the respective sampling date. “NS” denotes that the sample site was not sampled on that particular sampling date. Samples positive for Listeria are described by two or three capital letters that correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), hemolytic Listeria innocua (HLI), and Listeria welshimeri (LW). The L. monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1042B) and the Listeria spp. sigB allelic type (e.g. AT 70) follow these species abbreviations. b All dates in 2011 correspond to an initial longitudinal sampling period. The July 2012 sampling date corresponds to a pre-operational follow-up sampling. The September 2012 sampling date corresponds to a mid-shift follow-up sampling. The last two samplings on 11/15/2013 were conducted on the same day, but the first from left to right corresponds to a pre-operational follow up sampling, and the second corresponds to a mid-shift follow-up sampling, which occurred during operations.

201

Texas Tech University, Alex Brandt, May 2014

TABLE 3.10 Listeria contamination patterns in new smokehouse room and cooler area of Facility Ba Date Sampledb Sample Site 11/15/2012 11/13/2013 11/13/2013

c - - Site 56 - Bottom/Ramp of New Smokehouse LM DUP-1062B Site 57 - Door of New Smokehouse - - - Site 58 - Interior of New Smokehouse - - - Site 59 - East-West Trench Drain LM DUP-1062B LM DUP-1062B - Site 60 - North-South Trench Drain - LM DUP-1062B - LM DUP-1062B LM DUP-1062B LM DUP-1053E Site 61 - Cooler Door Track LI AT 6 LM DUP-1053E LM DUP-1062B LM DUP-1062B Site 62 - New Cooler Floor LI AT 22 LI AT 26

Site 63 - North Floor-Wall Junction in New Cooler - - -

Site 64 - Trench Drain in New Cooler LM DUP-1062B LM DUP-1062B LM DUP-1062B LM DUP-1062B LI AT 6 LM DUP-1062E Site 65 - Trench Drain in New Inedible Area LW AT 133 a A result of “-” corresponds to a negative result for that particular site on the respective sampling date. Samples positive for Listeria are described by two capital letters which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), and Listeria welshimeri (LW). The L. monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1062B) and the Listeria spp. sigB allelic type (e.g. AT 6) follow these species abbreviations. b The November 2012 sampling date corresponds to a mid-shift follow-up sampling. The last two samplings on 11/13/2013 were conducted on the same day, but the first from left to right corresponds to a pre-operational follow up sampling, and the second corresponds to a mid-shift follow-up sampling. c Strains which were persistent in the originally-sampled areas of the facility during the initial longitudinal period are underlined.

202

Texas Tech University, Alex Brandt, May 2014

TABLE 3.11 Initial and follow-up period Listeria prevalence among certain site groups in each facility All Sites Followed During Entire Studya Number Initial Follow-Up Facility Significanced of Sites Prevalence (%)b Prevalence (%)c A 19 46.5 36.8 0.2687 B 34 47.8 27.9 < 0.0001* C 35 49.0 20.0 < 0.0001*

Processing Room Sites Followed During Entire Studye Number Initial Follow-Up Facility Significance of Sites Prevalence (%) Prevalence (%) B 22 42.4 25.0 0.0040* C 9 48.1 0.0 < 0.0001*f a The subset of all sites in a particular facility that were: i) positive at least once during the initial longitudinal study period, and ii) followed throughout the entire study (i.e. were not lost to facility physical changes during follow-up). b The percentage of positive samples from the subset of sites during the initial longitudinal study period c The percentage of positive samples from the subset of sites during the follow-up period after trainings. d Significance changes in prevalence between the initial and follow-up periods are designated with a P- value calculated from a repeated measures regression model in PROC GENMOD of SAS version 9.3. P values less than 0.05 were considered significant, and significance is indicated with a *. e The subset of processing room sites in Facilities B and C, respectively, that were a part of the subset of sites in the upper portion of the table. f A repeated measures regression model based on a Poisson distribution was used in the calculation of this P value as opposed to a Binomial distribution used for other P values in the table.

203

Texas Tech University, Alex Brandt, May 2014

FIG 3.1 Facility A sampling scheme for 2012 follow-up sampling occasions

204

Texas Tech University, Alex Brandt, May 2014

FIG 3.2 Facility B sampling scheme for 2012 follow- up sampling occasions 205

Texas Tech University, Alex Brandt, May 2014

FIG 3.3 Facility C sampling scheme for 2012 follow- up sampling occasions 206

Texas Tech University, Alex Brandt, May 2014

FIG 3.4 Facility A sampling scheme for 2013 follow- up sampling occasions

207

Texas Tech University, Alex Brandt, May 2014

FIG 3.5 Facility B sampling scheme for 2013 follow- up sampling occasions

208

Texas Tech University, Alex Brandt, May 2014

FIG 3.6 Facility C sampling scheme for 2013 follow-up sampling occasions

209

Texas Tech University, Alex Brandt, May 2014

100.00

90.00 80.00 * * * 70.00 * 60.00 Pre-Training

Post-Training 50.00 40.00

Test ScoreTest (%Correct) 30.00

20.00 10.00 0.00 Facility A Facility B Facility1 C Facility D Overall Facility

FIG 3.7 Mean pre- and post-training knowledge assessment scores for all facilities. An asterisk (*) above the column for a particular facility indicates that the post-training knowledge assessment score mean was significantly greater (P < 0.05) than the pre-training knowledge assessment score mean when analyzed using a Paired t-Test in SAS version 9.3. Error bars represent the ± values of the standard error of the mean.

210

Texas Tech University, Alex Brandt, May 2014

Site 3: Grinding Room Drains/Trench Drain/Floor Site 8: Metal Table in Stuffing Room

N N NEG W E NEG W E S S

LW AT 69 LW AT 69 LW AT 69 NEG NEG NEG

NEG NEG

Site 10.1: Stuffing Room Drain North Site 10.2: Stuffing Room Drain South

N N LM DUP- W E LW AT 69 W E 1052A S S

LM DUP- LM DUP- LM DUP- 1052A NEG NEG LW AT 69 1052A 1052A LW AT 69

LM DUP- 1052A NEG LW AT 69

Site 11: Stuffing Room Stress Mats Site 14: Door from Stuffing Room to Cooking

LM DUP- N N 1052A W E NEG W E LW AT 69 S S

LM DUP- LM DUP- LM DUP- LW AT 69 1052A 1052A NEG NEG 1052A LW AT 69 LW AT 69

LM DUP- 1052A NEG LW AT 69

211

Texas Tech University, Alex Brandt, May 2014

Site 16: Tumbling Room Pallets Site 25: Packaging Room Drains

N N LM DUP- W E NEG W E 1052A S S

LM DUP- NEG NEG NEG NEG NEG 1052A

NEG NEG

FIG 3.8 Diagram of Listeria results from directional swabs taken from sites in Facility A. Directional samples were collected in four cardinal directions with respect to the original sample site, which is represented by the middle square; the compass rose next to each sample site provides the orientation of the adjacent samples. A result of “NEG” corresponds to a negative result for that particular sample. Samples positive for Listeria are described by two capital letters , which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM) and Listeria welshimeri (LW). The Listeria monocytogenes molecular subtype (e.g. Eco RI ribotype DUP-1052A) and the Listeria spp. sigB allelic type (e.g. AT 56) follow these species abbreviati ons.

212

Texas Tech University, Alex Brandt, May 2014

Site 1: Processing Room Drain #1 Site 5: Processing Room Drain #2 N N W E W E NEG NEG S S

LM DUP- NEG NEG NEG LI AT 45 1062B LI AT 45 LI AT 45

NEG LI AT 45

Site 10: Smokehouse #2 Bottom and Ramp Site 11: Smokehouse #3 Bottom and Ramp

N N LM DUP- W E W E 1062B S NEG S LI AT 108

LM DUP- 1062B NEG LI AT 108 NEG NEG NEG LI AT 6

LM DUP- NEG 1062B

Site 16: Floor Next to Smokehouses Site 17: Trench Drain #1

N N LM DUP- LM DUP- W E W E 1062B 1062B S S LI AT 45

LM DUP- LM DUP- LM DUP- LM DUP- LM DUP- LM DUP- 1062B 1062B 1062B 1062B 1062B 1062B LI AT LI AT 6 LI AT 108 LI AT 108 LI AT 6 NEW A LM DUP- LM DUP- 1062B 1062B LI AT 108 LI AT 108

213

Texas Tech University, Alex Brandt, May 2014

Site 18: Door from Processing Room to Cooler 2 Site 20: Trench Drain #2 N N LM DUP- W E W E 1062B S NEG S LI AT 45

LM DUP- LM DUP- LM DUP- NEG LI AT 45 NEG 1062B 1062B 1062B

LM DUP- LM DUP- 1062B 1062B LI AT 45 LI AT 45

Site 35: Cooler 2 Floor (Crack) Site 39: Cooler 1 Floor

N N LM DUP- W E LM DUP- W E 1062B S 1062B S LI AT 45 LM DUP- LM DUP- LM DUP- LM DUP- 1062B LM DUP- 1062B 1062B LI AT 45 1062B LI AT 1062B LI AT 6 LI AT 45 LI AT 6 NEW A

LM DUP- LM DUP- 1062B 1062B LI AT 23

Site 42: Cooler 1 Drain Site 45: Kill Floor Drain

N N LM DUP- W E LM DUP- W E 1062B S 1062B S LI AT 6

LM DUP- LM DUP- LM DUP- LM DUP- LM DUP- 1062B 1062B 1062B LI AT 6 1062B 1062B LI AT 6 LI AT 70 LI AT 6 LI AT 6 LI AT 6

LI AT 70 NEG

214

Texas Tech University, Alex Brandt, May 2014

Site 46: Kill Floor Floor

N LW AT W E NEW C S

LW AT NEG LI AT 6 NEW C

NEG

FIG 3.9 Diagram of Listeria results from directional swabs taken from sites in Facility B. Directional samples were collected in four cardinal directions with respect to the original sample site, which is represented by the middle square; the compass rose next to each sample site provides the orientation of the adjacent samples. A result of “NEG” corresponds to a negative result for that particular sample. Samples positive for Listeria are described by two capital letters , which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), and Listeria welshimeri (LW). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1052A) and the Listeria spp. sigB allelic type (e.g. AT 56) follow these species abbreviations.

215

Texas Tech University, Alex Brandt, May 2014

Site 15: Floor Next to Square Drain Site 19: Smoking Area Floor and Floor-Wall Jct.

NEG NEG

NEG NEG NEG LI AT 70 LI AT 70 NEG

LM DUP- NEG 1042B LI AT 70

Site 26: Large Blood Catch Drain Site 27: Kill Floor Floor

LI AT 70 NEG

LI AT 71 NEG LI AT 70 LI AT 70 LI AT 23 LI AT 70

Site 39: Cooler 2 Drain NEG LI AT 71

LM DUP - 1057B

LM DUP- LM DUP - LM DUP- 1057B 1057B 1057B LI AT 70 LI AT 70 LI AT 70

LM DUP - 1042B LI AT 70

FIG 3.10 Diagram of Listeria results from directional swabs taken from sites in Facility C. Directional samples were collected in four cardinal directions with respect to the original sample site, which is represented by the middle square; the compass rose next to each sample site provides the orientation of the adjacent samples. A result of “NEG” corresponds to a negative result for that particular sample. Samples positive for Listeria are described by two capital letters, which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM) and Listeria innocua (LI). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1052A) and the Listeria spp. sigB allelic type (e.g. AT 56) follow these species abbreviations.

216

Texas Tech University, Alex Brandt, May 2014

CHAPTER IV CHARACTERIZING THE COMPOSITION OF BACTERIAL POPULATIONS AT SITES OF RECURRENT LISTERIA CONTAMINATION IN MEAT PROCESSING FACILITIES USING 16S RRNA GENE SEQUENCING

Introduction Characterization of Listeria monocytogenes strain phenotypes such as sanitizer tolerance, surface adherence, and tolerance to stresses encountered in the food production chain has been unable to provide a set of risk factors that are consistently associated with persistence of the pathogen in food processing environments (1, 4, 9, 11, 12, 14, 17-19, 22, 25, 31, 33). This lack of agreement has prompted hypotheses considering other plausible risk factors such as rapid adaptation to facility environments due to prophage recombination (38), favorable growth conditions in harborage sites (7), and the metabolic and secretory actions of other bacteria that co-reside with Listeria in biofilms. In general, the presence of other microorganisms co-residing with L. monocytogenes is thought to decrease the pathogen’s ability to grow due to competition for nutrients or secretion of factors that inhibit growth or attachment to surfaces (21, 23, 24, 30). However, some instances have been documented where other bacterial species have contributed to increased density of L. monocytogenes cells adhered to different surface materials (5, 8, 28, 36). While there is likely a complex set of factors that drive these phenomena, polysaccharide production (36), desiccation tolerance (5), and disinfectant tolerance contributed by neighboring organisms have all been postulated to facilitate the colonization of certain sites with L. monocytogenes. Because knowledge surrounding the interactions of L. monocytogenes with other bacteria commonly encountered in food processing environments is limited (7, 29), the primary objective of this study was to characterize the composition of bacterial populations that are present in sites of recurrent L. monocytogenes contamination. Using 16S rRNA gene sequencing of isolates obtained from sites that were recurrently contaminated with Listeria over the course of a six-month longitudinal sampling period,

217

Texas Tech University, Alex Brandt, May 2014 we were able to identify genera of culturable bacteria that were present in these sites during a mid-shift follow-up sampling occasion.

Materials and Methods Selection of Sampling Sites and Sample Collection Environmental sites selected for investigation in this study were chosen from a pool of sites from one fresh and two ready-to-eat (RTE) meat production facilities that were positive at least once for L. monocytogenes or non-pathogenic Listeria species during a longitudinal sampling study in the facilities. The sites belonging to this pool were also sampled for Listeria pre-operationally on one occasion between seven and nine months after the end of the longitudinal study, and at mid-shift on the day that samples were collected for this study. Two sites from Facility A (RTE), Sites 10 and 11, four sites from Facility B (RTE), Sites 17, 35, 39, and 42, and three sites from Facility C (Fresh), Sites 15, 19, and 39 were selected. A description of the sites, and their Listeria sampling and persistence results are provided in Table 4.1. Selection of the sites was based on their high frequency of contamination with L. monocytogenes and/or non-pathogenic Listeria species during the initial longitudinal study, their status with regard to persistence of specific L. monocytogenes or non-pathogenic Listeria species strains, and their results during the pre-operational follow-up sampling. Sample Collection and Processing Sponge samples were collected from sites as described previously in Chapter II during a mid-shift follow-up sampling occasion between nine and eleven months after the end of initial longitudinal study period. Briefly, a 90 ml aliquot of Brain Heart Infusion (BHI) Broth (Becton, Dickinson and Company, Franklin Lakes, NJ) was added to each sponge sample and samples were mechanically homogenized for 2 min. The resultant solution was then serially diluted in Phosphate Buffered Saline and spread plated onto BHI Agar (Becton, Dickinson, and Company). BHI Agar plates were incubated at 37±2°C for 24±2 h, and were stored at 4°C until viewed for colony morphology. Up to ten different colonies with visually different morphologies were selected from any combination of plates created from the serial dilutions of the original samples. Colonies

218

Texas Tech University, Alex Brandt, May 2014 that were not well-isolated were re-streaked to new BHI Agar plates to obtain pure growth on plates, and were incubated as described above. 16S rRNA PCR and Sequencing All PCR and sequencing preparation steps were conducted based on the methods previously used by Huck et al. (20). Briefly, cells were picked from colonies on BHI Agar plates using sterile toothpicks and transferred to a 96-well PCR microtiter plate. Cells were lysed by microwaving for 4 min. After lysis, PCR master mix was added to the wells, where each 25 µl reaction volume consisted of 12.5 µl GoTaq Colorless Master Mix (Promega, Madison, WI), 10.5 µL nuclease free water, 1 µl of PEU7 forward primer (5’-GCAAACAGGATTAGATACCC-3’) (34) and 1 µl of DG74 reverse primer (5’- AGGAGGTGATCCAACCGCA-3’) (16). The PCR was carried out in an Applied Biosystems 2720 Thermal Cycler (Life Technologies, Thermo Fisher Scientific) with the following conditions: 94°C for 5 min, followed by 30 cycles of 94°C for 1 min, 50°C for 1 min, and 72°C for 1.5 min, then 72°C for 7 min, and a final hold of 4°C until removal. To confirm the presence of PCR products, the resulting solutions were electrophoresed for 1 h at 110 V in a 2.0% Agarose gel stained with SYBR Safe DNA Gel Stain (Invitrogen, Life Techonologies, Thermo Fisher Scientific). Gel images were visualized using an ultraviolet transilluminator, and captured using gel imaging software. Isolates from BHI Agar plates that were negative for the 16s rRNA gene PCR were presumed to be of domains other than Bacteria, and thus were not investigated further. The remaining PCR product (20 μl) was purified with a GenElute PCR Clean-Up Kit (Sigma-Aldrich, St. Louis, MO), checked for purity and concentration on a NanoDrop 2000 UV-Vis Spectrophotometer (Thermo Fisher Scientific) and frozen at -20°C. For sequencing submission, 6 μl of PCR products were mixed with 10 μl of nuclease free water and 2 μl of either PEU7 forward primer or 2 μl of P3SH reverse primer (5'- CTACGGTTACCTTGTTACGACTT-3') (32) both at 4 pmol/μl. Automated Sanger sequencing was performed at the Cornell University Biotechnology Resource Center (Ithaca, NY) using Big Dye Terminator Chemistry and AmpliTaq-FS DNA Polymerase (Life Technologies, Thermo Fisher Scientific) with analysis on an Applied Biosystems Automated 3730xl DNA Analyzer (Life Technologies, Thermo Fisher Scientific).

219

Texas Tech University, Alex Brandt, May 2014

Sequence Assembly, Identification, and Allelic Typing Forward and reverse sequences for each isolate were assembled, trimmed, and proofread using the SeqMan Pro software of the Lasergene Core Suite Version 9 (DNASTAR, Inc., Madison, WI). When necessary, nucleotide ambiguity codes were assigned to sequences based on the guidelines provided by the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology. The entire group of unaligned and untrimmed sequences was saved to a .fas file using the MegAlign in Lasergene Core Suite Version 9, and was uploaded to the Ribosomal Database Project’s (RDP; https://rdp.cme.msu.edu/index.jsp) SeqMatch Function to assign the genus of the isolates. Because multiple matches were assigned to each isolate, and top matches did not consistently correspond to the genus explicitly assigned by RDP, classifications provided here were limited to the assigned genus of the organism. In cases where multiple genera were associated with the isolate, the Family was provided as the identification. For each group of isolates assigned to a particular genus or family, allelic types AT) (a unique combination of sequence polymorphisms in the 16S rRNA gene) were assigned based on an alignment with a 557 bp region that was present for all isolates included in the study. Alignments were made in MegAlign and phylogenetic trees were contstructed to assign ATs to isolates. Arbitrary numbers were assigned to each AT for differentiation. ATs were also used to calculate Simpson’s Index of Diversity (SID) according to the formula 1-D where D = ∑ (n/N)2 and “n” is the number of isolates belonging to a specific AT in the group, and “N” is the total number of all isolates of all ATs in the group. SID was calculated for groups of isolates among a facility and among the specific sites within the facilities.

Results Results from Facility A Sites Of all 20 isolates from sites in Facility A, only three were unable to be characterized; two isolates were negative for 16S PCR and one was due to failed sequencing. Pseudomonas predominated the group of isolates collected from both sites comprising 42.11% of the isolates, followed by indistinguishable Enterobacteriaceae, which comprised 10.53% of isolates (Table 4.2). One isolate each of Acinetobacter, 220

Texas Tech University, Alex Brandt, May 2014

Carnobacterium, Exiguobacterium, Microbacterium, Serratia, Sphingobacterium, and Stenotrophomonas were identified. Each comprised 5.26% of the isolates. Within the Pseudomonas group, two ATs were identified (1 and 5), and the Enterobacteriacae group also contained two ATs (5 and 6). The SID calculated for Facility A was 0.796, which was the lowest of the three facilities (Table 4.3). This low value was likely due to the predominance of the Pseudomonas AT 1 among the isolates. The two sites, Sites 10 and 11, each had very similar SID values of 0.741 and 0.781, respectively (Table 4.3). Results from Facility B Sites Eleven of the 40 isolates picked from Facility B sites were not positive by 16S PCR, presumably because a high amount the isolates were not of prokaryotic origin. Two other isolates had failed sequencing reactions and were unable to be characterized. Among the remaining 27 isolates, Pseudomonas were again the largest group comprising 22.22% of isolates. Staphylococcus comprised the second largest amount of isolates at 18.52% followed by Aeromonas at 14.81%. Buttiauxella, Yersinia, and indistinguishable Enterobacteriaceae all had two isolates apiece, and each comprised 7.41% of the isolates. Single isolates (3.70% each) of Budvicia, Enterobacter, Kluyvera, Lactococcus, Macrococcus, and Psychrobacter were isolated as well. The Pseudomonas group also had the greatest number of ATs, with five distinct ATs identified among its six isolates. Aeromonas had the second highest with three ATs, followed by Staphylococcus, Buttiauxella, and indistinguishable Enterobacteriaceae with two ATs apiece. Isolates of all other groups were comprised of only one distinct AT. The SID calculated for Facility B reflected the high diversity of the group with a value of 0.944 (Table 4.3). Individual site SIDs were similar, and ranged from 0.800 for Site 42 to 0.857 for Site 17 (Table 4.3). Results from Facility C Sites Only two of the 30 isolates from Facility C were unable to be identified, each because of failed sequencing reactions. The remaining 28 isolates were again predominated by Pseudomonas, which comprised 35.71% of isolates, followed by Psychrobacter at 17.86% and indistinguishable Enterobacteriaceae at 14.29%. Acinetobacter and Shewanella were also identified and each comprised 10.71% of the isolates. Carnobacter, Kocuria, and Leucobacter were also isolated, and each made up 3.57% of the isolates. Again Pseudomonas had the most ATs with eight observed, along 221

Texas Tech University, Alex Brandt, May 2014 with four among the Enterobacteriaceae, three among Shewanella, and two among Psychrobacter. The SID for Facility C was very similar to that of Facility B with a value of 0.941 (Table 4.3). Also, individual site values were in a range similar to that of Facility B sites, where values ranged from 0.790 at Site 19 to 0.864 at Site 39 (Table 4.3).

Discussion Pseudomonas is the Predominant and Richest Group When Observing the Composition of the Population Across All Sampling Sites The observation that Pseudomonas comprised the highest number of isolates in each facility, and always either had the most ATs or was tied for the most ATs, was not particularly surprising given that it is has been previously shown in other studies to be a predominant organism in various food processing environments (2, 26). It is interesting to note that seven of the nine sites selected all had at least one isolate of Pseudomonas obtained from them, the exceptions being Facility B Site 35, and Facility C Site 19. While Pseudomonas fragi has been shown to facilitate L. monocytogenes adherence and biofilm formation on glass surfaces (36), thereby potentially contributing to its persistence, other Pseudomonas species, such as Pseudomonas fluorescens have been shown to produce biosurfactants that are inhibitory to L. monocytogenes attachment and biofilm formation (27). Yet in other cases the presence of P. fluorescens in co-culture with L. monocytogenes has led to increased maximum population densities of the latter, though this was found to be dependent on temperature and salt concentration. Lower growth temperatures and salt concentrations led to decreased maximum population densities of the pathogen (6). Thus, further taxonomic classification of Pseudomonas down to the species level may be necessary to determine whether Pseudomonas present at sites of L. monocytogenes recurrence are beneficial or antagonistic for the pathogen’s persistence. Detailed investigations of growth conditions may also be necessary to determine the interactions of the two groups of organisms at certain sites in facilities. Other Genera Observed Are Also Common to Food Processing Operations Staphylococcus, which comprised the second largest amount of isolates in Facility B, though only of two ATs, has been shown to be a regular inhabitant of milk, meat, and pastry processing environments (26), has been previously identified as a suitable 222

Texas Tech University, Alex Brandt, May 2014 indicator for L. monocytogenes contamination in dairy processing environments (15), and certain species of the genus have been shown to increase L. monocytogenes biofilm CFU counts (8). However, evidence exists which suggests that staphylococci may also reduce adhesion and biofilm formation by L. monocytogenes (21), and this may be due to anti- adhesive properties of staphylococci or nutrient competition (23). Yet a more recent study conducted with co-culture of L. monocytogenes with microorganisms isolated from small-scale cheese production facilities on ceramic tiles showed that Staphylococcus along with species of Psychrobacter, Rhodococcus, and Pseudomonas had no effect on L. monocytogenes survival (37). Thus, as noted for Pseudomonas above, different species of the same genera and even different subtypes within a species may contribute to different growth profiles for L. monocytogenes when the two organisms are grown in co-culture. This emphasizes the need to investigate interactions of specific organisms on a minute scale mimicking the specific conditions of a facility site. Buttiauxella, also identified in Facility B with two isolates of two ATs, is a common mesophilic organism present on meat (13). Shewanella, which belonged to three ATs in Facility C, are also common spoilage bacteria of meat and cured meat products (3). Enterobacteriaceae, which were found in all three facilities, are inherent to red meat and poultry processing (35), and are among the most common organisms found in cold- smoked and semi-preserved fish plants (2). The same study also noted that Psychrobacter, which was found in Facility C, was a predominant organism in a caviar processing facility (2). The commonality of these groups further reinforces the rationale for isolating these organisms in such relatively large proportions in this study. Minor Groups of Organisms Found in This Study May Have Different Roles in L. monocytogenes Growth Of particular interest was the presence of some minor groups of organisms identified among these sites of Listeria recurrence. Stenotrophomonas maltophilia has been previously shown to increase L. monocytogenes biofilm CFU counts by 0.69 log10 CFU when compared to pure culture of L. monocytogenes (8), and an isolate belonging to the Stenotrophomonas genus was identified in Facility A Site 11. Likewise, Kocuria varians has also been shown to increase L. monocytogenes cell density in co-culture (though this co-culture also permitted a greater degree of detachment of both organisms 223

Texas Tech University, Alex Brandt, May 2014 from the surface they were colonizing) (28), and a Kocuria spp. isolate was found in Facility C Site 19. This observation is interesting because this site was positive for either L. monocytogenes or non-pathogenic Listeria species at each sampling date, including the pre-operational and mid-shift sampling occasions, and was a site of persistence for both L. monocytogenes DUP-1042B and L. innocua sigB allelic type 70. In contrast, the isolation of Lactococcus from Facility B Site 17 was interesting for the opposite reason, as strains of Lactococcus lactis have recently been shown to be effective at eliminating L. monocytogenes growth in biofilms and flow drains of a ready- to-eat poultry production facility (39). Likewise, species of Serratia (isolated from Facility A Site 10) and Shewanella (isolated from Facility C Sites 19 and 39) have also been shown to reduce the population of L. monocytogenes in dual-species biofilms, though L. monocytogenes survives for longer periods of time than bacteria of either genera when the biofilm is subjected to dessication (10). Carnobacterium strains have also been shown to produce antimicrobials that can decrease L. innocua by 2.7 log10 CFU/g when co-inoculated on the surface of cold-smoked salmon, making it all the more surprising that isolates of the genus were found in two sites in this study: Facility C Site 39 and Facility C Site 10.

Conclusions Results from this study showed that the population of culturable bacteria that co- reside with Listeria in sites of recurrent Listeria contamination are generally comprised of groups that are normally predominant in food processing facilities. However, other minor groups of interest are also present. While this study provides insight into the general composition of these populations, further characterization of co-residing bacteria beyond the genus level may be necessary to fully elucidate the relationships between Listeria and co-residing species. Indeed, different species of a genus, along with different species subtypes may contribute differently to the observed levels of Listeria growth and survival. Additionally, mimicking the conditions present at certain processing facility sites may be necessary to fully elucidate the effect of the presence of other organisms on the relative abundance of Listeria in co-culture. Furthermore, using metagenomic approaches to study the composition of the microbial communities from sites of recurrent 224

Texas Tech University, Alex Brandt, May 2014

Listeria contamination may be a point to consider for future studies in order to evaluate the contribution of non-culturable organisms to Listeria’s success and long-term survival at these sites. Lastly, another hypothesis that should be investigated is whether other organisms present in sites of recurrent contamination actually promote Listeria persistence by providing stressful conditions that cause Listeria cells to enter into persistent long-term survival states as a response to the stress. This would be interesting to investigate, especially since most previous work has been with the notion that other organisms present in these particular sites must improve growth of Listeria, not force it to become persistent due to the stress provided by its neighboring organisms.

2. Bagge-Ravn D, Ng Y, Hjelm M, Christiansen JN, Johansen C, Gram L. 2003. The microbial ecology of processing equipment in different fish industries— analysis of the microflora during processing and following cleaning and disinfection. Int. J. Food Microbiol. 87:239-250.

3. Borch E, Kant-Muermans ML, Blixt Y. 1996. Bacterial spoilage of meat and cured meat products. Int. J. Food Microbiol. 33:103-120.

4. Borucki MK, Peppin JD, White D, Loge F, Call DR. 2003. Variation in biofilm formation among strains of Listeria monocytogenes. Appl. Environ. Microbiol. 69:7336-7342.

5. Bremer PJ, Monk I, Osborne CM. 2001. Survival of Listeria monocytogenes attached to stainless steel surfaces in the presence or absence of Flavobacterium spp. J. Food Prot. 64:1369-1376.

6. Buchanan RL, Bagi LK. 1999. Microbial competition: effect of Pseudomonas fluorescens on the growth of Listeria monocytogenes. Food Microbiol. 16:523- 529.

7. Carpentier B, Cerf O. 2011. Review - persistence of Listeria monocytogenes in food industry equipment and premises. Int. J. Food Microbiol. 145:1-8.

225

Texas Tech University, Alex Brandt, May 2014

8. Carpentier B, Chassaing D. 2004. Interactions in biofilms between Listeria monocytogenes and resident microorganisms from food industry premises. Int. J. Food Microbiol. 97:111-122.

9. Cruz CD, Fletcher GC. 2011. Prevalence and biofilm-forming ability of Listeria monocytogenes in New Zealand mussel (Perna canaliculus) processing plants. Food Microbiol. 28:1387-1393.

10. Daneshvar Alavi HE, Truelstrup Hansen L. 2013. Kinetics of biofilm formation and desiccation survival of Listeria monocytogenes in single and dual species biofilms with Pseudomonas fluorescens, Serratia proteamaculans or Shewanella baltica on food-grade stainless steel surfaces. Biofouling 29:1253- 1268.

11. Djordjevic D, Wiedmann M, McLandsborough LA. 2002. Microtiter plate assay for assessment of Listeria monocytogenes biofilm formation. Appl. Environ. Microbiol. 68:2950-2958.

12. Earnshaw AM, Lawrence LM. 1998. Sensitivity to commercial disinfectants, and the occurrence of plasmids within various Listeria monocytogenes genotypes isolated from poultry products and the poultry processing environment. J. Appl. Microbiol. 84:642-648.

13. Ercolini D, Russo F, Nasi A, Ferranti P, Villani F. 2009. Mesophilic and psychrotrophic bacteria from meat and their spoilage potential in vitro and in beef. Appl. Environ. Microbiol. 75:1990-2001.

14. Fox EM, Leonard N, Jordan K. 2011. Physiological and transcriptional characterization of persistent and nonpersistent Listeria monocytogenes isolates. Appl. Environ. Microbiol. 77:6559-6569.

15. Frank J, Gillett R, Ware G. 1990. Association of Listeria spp. contamination in the dairy processing plant environment with the presence of staphylococci. J. Food Prot. 53:928–932.

16. Greisen K, Loeffelholz M, Purohit A, Leong D. 1994. PCR primers and probes for the 16s ribosomal-RNA gene of most species of pathogenic bacteria, including bacteria found in cerebrospinal-fluid. J. Clin. Microbiol. 32:335-351.

17. Harvey J, Keenan KP, Gilmour A. 2007. Assessing biofilm formation by Listeria monocytogenes strains. Food Microbiol. 24:380-392.

226

Texas Tech University, Alex Brandt, May 2014

18. Heir E, Lindstedt B-A, Røtterud O-J, Vardund T, Kapperud G, Nesbakken T. 2004. Molecular epidemiology and disinfectant susceptibility of Listeria monocytogenes from meat processing plants and human infections. Int. J. Food Microbiol. 96:85-96.

19. Holah JT, Taylor JH, Dawson DJ, Hall KE. 2002. Biocide use in the food industry and the disinfectant resistance of persistent strains of Listeria monocytogenes and Escherichia coli. J. Appl. Microbiol. 92S:111S-120S.

20. Huck JR, Woodcock NH, Ralyea RD, Boor KJ. 2007. Molecular subtyping and characterization of psychrotolerant endospore-forming bacteria in two New York state fluid milk processing systems. J. Food Prot. 70:2354-2364.

21. Jeong DK, Frank JF. 1994. Growth of Listeria monocytogenes at 10oC in biofilms with microorganisms isolated from meat and dairy processing environments. J. Food Prot. 57:576-586.

22. Latorre AA, Van Kessel JA, Karns JS, Zurakowski MJ, Pradhan AK, Boor KJ, Adolph E, Sukhnanand S, Schukken YH. 2011. Increased in vitro adherence and on-farm persistence of predominant and persistent Listeria monocytogenes strains in the milking system. Appl. Environ. Microbiol. 77:3676- 3684.

23. Leriche V, Carpentier B. 2000. Limitation of adhesion and growth of Listeria monocytogenes on stainless steel surfaces by Staphylococcus sciuri biofilms. J. Appl. Microbiol. 88:594-605.

24. Leriche V, Chassaing D, Carpentier B. 1999. Behaviour of L. monocytogenes in an artificially made biofilm of a nisin-producing strain of Lactococcus lactis. Int. J. Food Microbiol. 51:169-182.

25. Lundén J, Autio T, Markkula A, Hellström S, Korkeala H. 2003. Adaptive and cross-adaptive responses of persistent and non-persistent Listeria monocytogenes strains to disinfectants. Int. J. Food Microbiol. 82:265-272.

26. Mettler E, Carpentier B. 1998. Variations over time of microbial load and physicochemical properties of floor materials after cleaning in food industry premises. J. Food Prot. 61:57-65.

27. Meylheuc T, van Oss CJ, Bellon-Fontaine MN. 2001. Adsorption of biosurfactant on solid surfaces and consequences regarding the bioadhesion of Listeria monocytogenes LO28. J. Appl. Microbiol. 91:822-832.

227

Texas Tech University, Alex Brandt, May 2014

28. Midelet G, Kobilinsky A, Carpentier B. 2006. Construction and analysis of fractional multifactorial designs to study attachment strength and transfer of Listeria monocytogenes from pure or mixed biofilms after contact with a solid model food. Appl. Environ. Microbiol. 72:2313-2321.

29. Møretrø T, Langsrud S. 2004. Listeria monocytogenes: biofilm formation and persistence in food-processing environments. Biofilms 1:107-121.

30. Norwood DE, Gilmour A. 2001. The differential adherence capabilities of two Listeria monocytogenes strains in monoculture and multispecies biofilms as a function of temperature. Lett. Appl. Microbiol. 33:320-324.

31. Porsby CH, Vogel BF, Mohr M, Gram L. 2008. Influence of processing steps in cold-smoked salmon production on survival and growth of persistent and presumed non-persistent Listeria monocytogenes. Int. J. Food Microbiol. 122:287-295.

32. Ralyea RD, Wiedmann M, Boor KJ. 1998. Bacterial tracking in a dairy production system using phenotypic and ribotyping methods. J. Food Prot. 61:1336-1340.

33. Ringus DL, Ivy RA, Wiedmann M, Boor KJ. 2012. Salt stress-induced transcription of σB- and CtsR-regulated genes in persistent and non-persistent Listeria monocytogenes strains from food processing plants. Foodborne Pathog. Dis. 9:198-206.

34. Rothman RE, Majmudar MD, Kelen GD, Madico G, Gaydos CA, Walker T, Quinn TC. 2002. Detection of bacteremia in emergency department patients at risk for infective endocarditis using universal 16S rRNA primers in a decontaminated polymerase chain reaction assay. J. Infect. Dis. 186:1677-1681.

35. Sade E, Murros A, Bjorkroth J. 2013. Predominant enterobacteria on modified- atmosphere packaged meat and poultry. Food Microbiol. 34:252-258.

36. Sasahara KC, Zottola EA. 1993. Biofilm formation by Listeria monocytogenes utilizes a primary colonizing microorganism in flowing systems. J. Food Prot. 56:1022-1028.

37. Schirmer BC, Heir E, Moretro T, Skaar I, Langsrud S. 2013. Microbial background flora in small-scale cheese production facilities does not inhibit growth and surface attachment of Listeria monocytogenes. J. Dairy Sci. 96:6161- 6171.

228

Texas Tech University, Alex Brandt, May 2014

38. Verghese B, Lok M, Wen J, Alessandria V, Chen Y, Kathariou S, Knabel S. 2011. comK prophage junction fragments as markers for Listeria monocytogenes genotypes unique to individual meat and poultry processing plants and a model for rapid niche-specific adaptation, biofilm formation, and persistence. Appl. Environ. Microbiol. 77:3279-3292.

39. Zhao T, Podtburg TC, Zhao P, Chen D, Baker DA, Cords B, Doyle MP. 2013. Reduction by competitive bacteria of Listeria monocytogenes in biofilms and Listeria bacteria in floor drains in a ready-to-eat poultry processing plant. J. Food Prot. 76:601-607.

229

Texas Tech University, Alex Brandt, May 2014

TABLE 4.1 Descriptions and Listeria results for sampling sites included in this study Outcome During Initial Persistence Pre-Op Mid-Shift b Site (Description)a Six Months of Sampling Status Follow-Up Follow-Up Initial 1 Initial 2 Initial 3 Initial 4 Initial 5 Initial 6 (Subtype(s))c Outcome d Outcome e YES A10 (Stuffing Room Drains & Floor) LM/LS LM LS LM LM/LS - LS LM (116-239-S-2) A11 (Stuffing Room Stress Mats) LM/LS LM/LS - LM/LS LM/LS LM NO LS LM/LS

B17 (Trench Drain #1) LM/LS LM/LS LS LS LM/LS LS NO LM/LS LM

B35 (Cooler 2 Floor, Crack) LM/LS LM/LS LM/LS LM/LS LS LM NO LM/LS LM/LS

B39 (Cooler 1 Floor) LM/LS LS LM/LS LM/LS LS LS NO LM/LS LM/LS

B42 (Cooler 1 Drain) LM/LS LS LM/LS LS LM/LS LM/LS NO LM/LS LM/LS C15 (Floor Next to Closed YES LM/LS LM/LS LM LM LS - LM - Square Drain) (DUP-1042B) C19 (Smoking Area Floor and Floor- YES (AT 70; LM/LS LM LS LM/LS LS LM/LS LM/LS LS Wall Junction) DUP-1042B) YES C39 (Cooler 2 Drain) LS LM/LS LM/LS LM/LS LM/LS LM/LS - LM/LS (DUP-1042B) a The code for each sampling site is provided along with a description of the site in parentheses; Facility A and B are ready-to-eat meat production facilities, and Facility C is a fresh meat production facility. b Sampling results for each of the different sites during the initial six sampling dates described in Chapter II are provided. LM alone means that L. monocytogenes was only isolated from the site on that particular sampling date. LS alone means that only a non-pathogenic Listeria species isolate was isolated. LM/LS means that both L. monocytogenes and a non-pathogenic Listeria species isolate were isolated from the site on that particular date. c Persistence status with regard to the outcome of a binomial test for subtype frequency applied to the site based on results from an initial longitudinal sampling period. YES signifies that a strain was persistent at the site and NO signifies no persistence at the site. If a site was determined to be a site of persistence, the strain found to be persistent is given in parentheses. For L. monocytogenes the EcoRI ribotype of the strain is given (e.g. DUP-1042B), and for non-pathogenic Listeria species the sigB allelic type (AT) (e.g. AT 70) is provided. d Listeria results for the pre-operational follow-up sampling date. e Listeria results for the mid-shift follow-up sampling date; this was the same date on which the samples for this study were collected.

230

Texas Tech University, Alex Brandt, May 2014

TABLE 4.2 Ribosomal Database Project identifications given to isolates collected Sampling Site (Facility)a Colony Designation Identification (Allelic Type) 10 (A) A Pseudomonas (1) 10 (A) B Pseudomonas (1) 10 (A) C Carnobacterium (2) 10 (A) D Enterobacteriaceae (5) 10 (A) F Pseudomonas (1) 10 (A) G Pseudomonas (5) 10 (A) H Acinetobacter (1) 10 (A) I Serratia (1) 10 (A) J Pseudomonas (1) 11 (A) A Sphingobacterium (1) 11 (A) B Enterobacteriaceae (6) 11 (A) D Pseudomonas (1) 11 (A) E Stenotrophomonas (1) 11 (A) F Pseudomonas (1) 11 (A) G Microbacterium (1) 11 (A) H Pseudomonas (1) 11 (A) J Exiguobacterium (1) 17 (B) A Pseudomonas (3) 17 (B) B Lactococcus (1) 17 (B) F Enterobacteriaceae (7) 17 (B) G Aeromonas (2) 17 (B) H Kluyvera (1) 17 (B) I Pseudomonas (2) 17 (B) J Enterobacter (1) 35 (B) A Aeromonas (1) 35 (B) B Yersinia (1) 35 (B) C Aeromonas (1) 35 (B) D Buttiauxella (1) 35 (B) E Staphylococcus (1) 35 (B) F Staphylococcus (2) 35 (B) H Yersinia (1) 35 (B) I Budvicia (1) 35 (B) J Psychrobacter (2) 39 (B) B Pseudomonas (9) 39 (B) D Staphylococcus (2) 39 (B) F Pseudomonas (1) 39 (B) G Macrococcus (1) 39 (B) I Staphylococcus (1) 39 (B) J Pseudomonas (4) 42 (B) A Pseudomonas (4) 42 (B) C Buttiauxella (2) 42 (B) F Staphylococcus (1) 42 (B) G Aeromonas (3) 42 (B) J Enterobacteriaceae (8) 15 (C) A Enterobacteriaceae (1) 15 (C) B Pseudomonas (8) 15 (C) C Psychrobacter (1) 15 (C) D Pseudomonas (3) 15 (C) E Pseudomonas (10) 15 (C) F Pseudomonas (2) 15 (C) G Pseudomonas (7)

231

Texas Tech University, Alex Brandt, May 2014

TABLE 4.2, Cont. Ribosomal Database Project identifications given to isolates collected Sampling Site (Facility)a Colony Designation Identification (Allelic Type)b 15 (C) H Pseudomonas (3) 15 (C) I Pseudomonas (11) 15 (C) J Pseudomonas (2) 19 (C) A Psychrobacter (1) 19 (C) B Acinetobacter (1) 19 (C) C Acinetobacter (1) 19 (C) D Shewanella (1) 19 (C) F Acinetobacter (1) 19 (C) G Psychrobacter (1) 19 (C) H Shewanella (2) 19 (C) I Enterobacteriaceae (2) 19 (C) J Kocuria (1) 39 (C) A Pseudomonas (6) 39 (C) C Psychrobacter (2) 39 (C) D Enterobacteriaceae (3) 39 (C) E Carnobacterium (1) 39 (C) F Enterobacteriaceae (4) 39 (C) G Leucobacter (1) 39 (C) H Pseudomonas (12) 39 (C) I Psychrobacter (2) 39 (C) J Shewanella (3) a The number for each sampling site is provided along with the facility designation in parentheses; Facility A and B are ready-to-eat meat production facilities, and Facility C is a fresh meat production facility. b Allelic types were assigned for isolates within a certain genus or family based on unique sequence polymorphism combinations in a 557 bp region of the 16S rRNA gene. This 557 bp region was available for all isolates that were sequenced in the study. Allelic types were assigned arbitrary numbers for differentiation and were determined from phylogenetic trees constructed in MegAlign of Lasergene Core Suite Version 9.

232

Texas Tech University, Alex Brandt, May 2014

TABLE 4.3 Simpson’s Index of Diversity values for groups of isolates from facilities and specific sites a b Isolate Group Simpson’s Index of Diversity Facility A Site 10 – Stuffing Room Floor & Drains 0.741 Site 11 – Stuffing Room Stress Mats 0.781 Overall 0.796 Facility B Site 17 – Trench Drain #1 0.857 Site 35 – Cooler 2 Floor, Crack 0.840 Site 39 – Cooler 1 Floor 0.833 Site 42 – Cooler 1 Drain 0.800 Overall 0.944 Facility C Site 15 – Floor Next to Closed Square Drain 0.880 Site 19 – Smoking Area Floor and Floor-Wall Junction 0.790 Site 39 – Cooler 2 Drain 0.864 Overall 0.941

a Isolate groups consisted of those from a specific site (designated with the site code and description) or those that were grouped together from all sites in a facility (overall). b Simpson’s Index of Diversity was calculated using individual 16S rRNA allelic types as the means of isolate subtyping. Values were calculated according to the formula 1-D where D = ∑ (n/N)2 and “n” is the number of isolates belonging to a specific AT in the group, and “N” is the total number of all isolates of all ATs in the group.

233

Texas Tech University, Alex Brandt, May 2014

CHAPTER V IMPLICATIONS OF FINDINGS FROM THIS STUDY AND FUTURE PERSPECTIVES IN FOODBORNE PATHOGEN PERSISTENCE RESEARCH

The persistence of foodborne pathogens in food processing environments is a complex phenomenon that seems to entail the convergence of numerous risk factors. Many knowledge gaps exist regarding persistence, and this study aimed to fill several that are of critical importance. Overall, the outcomes of this study demonstrated i) the utility of combining longitudinal sampling, molecular subtyping, and statistical analysis to identify foodborne pathogen persistence within fresh and ready-to-eat (RTE) meat processing facilities, ii) the low prevalence and limited extent of persistence of Salmonella enterica and Escherichia coli O157:H7 within meat processing facilities, iii) the instigating effect of in-plant trainings with regard to improved personnel knowledge, behavioral changes, and physical facility improvements, iv) the positive impact of these changes on Listeria prevalence, v) the resilience of certain persistent Listeria strains despite implementation of controls, and vi) variation and similarities in the composition of bacterial populations at sites of recurrent Listeria contamination. The information gained from this study is applicable to the entire meat industry, but is especially useful for small and very small meat processors. These processors often have limited resources to combat persistence, and need concise sets of highly-effective best practices to employ. The findings of this study should help drive the development of these best practices, and should ultimately help these establishments manage the risks associated with strain persistence and product contamination. The observance of Listeria strain persistence in this study is not a novel aspect of the research. However, using statistical analysis to define pathogen persistence is a novel aspect, and to our knowledge this study is the first to apply the methodology of Malley et al. (7) to investigate foodborne pathogen persistence in a group of meat processing establishments. Though both statistical tests developed by Malley et al. have limitations, which became evident in this study, they provide substantial progress toward a standardized definition of persistence. Developing modified metrics that utilize the outcomes of both tests, or contain other variables of interest with regard to persistence 234

Texas Tech University, Alex Brandt, May 2014

(i.e. subtype diversity), should be of interest for future studies. If a single metric can simultaneously account for a strain’s overrepresentation and recurrent isolation, both of which seem to be consensus characteristics of a persistent strain, the metric may be more relevant than either of the Malley et al. test outcomes alone. Regardless of the approach, robust measures of persistence are critically needed if research in the form of quantitative microbial risk assessments and meta-analyses is to be conducted. Both types of studies depend on standardized measures to compare results between a large set of studies, and more of these meta-studies are needed to gain a holistic understanding of persistence. Another key aspect of this research is that S. enterica and E. coli O157:H7, when studied in parallel with Listeria, did not exhibit the same extent of environmental persistence. To our knowledge this is the first study to simultaneously document the ecology of all three pathogens within the meat processing environment, and thus is the first to provide some indication of whether S. enterica and E. coli O157:H7 commonly contaminate the same sites as Listeria. Data on S. enterica and E. coli O157:H7 persistence in other facility types (e.g. seafood, produce, nuts, and spices) are certainly needed, and although this study only followed a small group of facilities that solely process meat, the data herein nevertheless provide a point of comparison for future research efforts. Also, this study focused on those sites that are generally associated with Listeria persistence (i.e. drains, cracks, wet and cool sites, etc.). However, if S. enterica and E. coli O157:H7 persist, the sites where they do so may have entirely different characteristics. Indeed, the S. Agona strain linked to the 1998 and 2008 cereal outbreaks seemed to persist in a dry area of the facility. Therefore, studying pathogen persistence in sites with unconventional traits (e.g. those that are dry and dusty) is something to consider for further research. In all, more data is needed for S. enterica and E. coli O157:H7 to better understand their behaviors regarding persistence and risks involved. The persistence of other organisms, besides the three pathogens mentioned above, is also worthy of investigation. Persistence of spoilage organisms, like Pseudomonas in the dairy industry (8, 13), while not as crucial from a safety standpoint, can lead to major quality issues for food producers. Also, different non-O157 Shiga-toxin producing E. coli may be more environmentally hardy than their E. coli O157:H7 relatives. Thus, future research should also investigate the extent to which these non-O157 groups persist in

235

Texas Tech University, Alex Brandt, May 2014 environmental sites of food processing facilities, and aim to determine the extent to which their environmental persistence contributes to food product contamination. Also of interest is determining pathogen persistence at points in the food chain downstream of the processing facility. L. monocytogenes persistence at the retail level, especially in delis (2, 10, 14), is a topic that has been called to attention in recent years. Persistence of strains at these locations is also problematic because of the high degree of potential transfer to products, which are handled extensively in such establishments. Thus, identifying patterns of contamination at retail, as well as means to control it (10) are also key points to cover in future research studies. Yet another key finding of this research is that training programs, which instigated employee knowledge changes, behavioral changes, and physical facility changes, likely played some part in reducing prevalence of Listeria in meat processing facilities. With the exception of unpublished data from Williams et al. (16), the impact of such measures had only been previously documented for RTE seafood operations (6, 7). Thus, this evidence for a positive impact of trainings on Listeria prevalence reductions in meat processing facilities is noteworthy, and demonstrates the importance of trainings as a vital component of a Listeria Control Program. The observation whereby persistent strains remained present after implementation of control measures is a finding shared with previous studies (6, 7, 12). This is part of an emerging theme that suggests that complete elimination of persistent strains is unlikely to occur. This is especially pertinent with regard to the findings for Facility B where persistent strains from older areas of the establishment took up residence in completely new areas of the facility. In essence, it seems that even if control measures are enacted in a facility, the strains may take up harborage in other locations that provide similarly favorable conditions, and continue the cycle of persistence. Therefore, being keenly aware of the risks associated with certain procedures and workflows, and the measures needed to mitigate the risks, may be more beneficial than devoting copious resources to fully eradicate persistent strains. Employee trainings seem to be helpful in this regard, and should be included as key components of comprehensive food safety programs that also include good manufacturing practices, effective cleaning and sanitation procedures, and effective and aggressive environmental monitoring programs.

236

Texas Tech University, Alex Brandt, May 2014

Drains and floors having the greatest prevalence of contamination among all environmental site categories in this study, and in several previous studies (3, 5, 15), stresses the need for facilities to enact effective drain sanitation and floor-based control measures. Daily drain sanitation using designated brushes, low pressure water, and sanitizer is crucial. Using boot baths with sanitizer solution at entry points, floor foamers for high cart traffic areas, and proper treatments for floors in dry areas of the facility (i.e. the use of crystallized or powdered quaternary ammonium compounds or hydrogen peroxide floor treatments) are also essential practices. Though the reductions in Listeria prevalence in Facilities B and C during follow-up cannot be definitively tied to beginning facility-designated footwear and boot bath use, several facts seem to suggest some relationship: i) drain and floor sites had the highest prevalence in each facility during the initial longitudinal period, ii) prevalences of floors and drains only dropped after facility- designated footwear and boot baths were used, and iii) specific areas that were bound by these boot baths (i.e. the processing rooms) had very low levels of contamination thereafter. If this is true, facility designated footwear use and some means of footwear sanitation should be included among those best practices that are considered most critical. Another intriguing aspect of this study was the population composition observed for other microflora in sites of recurrent Listeria contamination. This is especially interesting since species that belong to some of the genera identified can increase L. monocytogenes adherence and cell densities on common food contact surface materials (1, 9). Studies that explore phenotypic and genotypic responses of L. monocytogenes to interactions with these other microflora should provide interesting insight into the role that they play in the organism’s success at these particular sites. Also, the contribution of non-culturable bacteria to Listeria persistence should be evaluated in future studies. Metagenomics tools, which are becoming more readily available, may greatly facilitate these assessments and may reveal interesting interactions between Listeria and other co- residing microflora that are not identified using more conventional methodology. One critical need that remains to be addressed regarding persistence research is developing a paradigm to decipher whether a strain’s recurrence is due to its survival and proliferation in the facility or is due to its constant reintroduction. Unfortunately, most studies (this one included) do not account for many confounding factors that could cause

237

Texas Tech University, Alex Brandt, May 2014 a strain to appear persistent. Thus, some “persistent” strains may not survive and proliferate in the facility at all, and their recurrence may be due to repeated introduction via incoming raw products, employee attire, packaging materials, or other vehicles. Strain persistence among animal herds that are the source of incoming raw materials may also cause the strain to appear persistent in the facility. Also, if a strain is highly prevalent in the respective geographic region, one may inaccurately attribute its recurrence to its persistence in the processing facility, and not to its high probability of isolation. Longitudinal sampling periods (i.e. several months to several years) with samples collected on a monthly or yearly basis are useful in observing long-term strain recurrence. However, these schemes only provide a snapshot of strain contamination patterns and fail to account for fluctuations that occur between sampling occasions. Indeed, a site may be deemed a site of persistence because a strain is isolated there on several consecutive monthly samplings, but one still has no idea of whether the strain was present at the site for the duration of the time interval. Also, because most sampling is conducted at the middle of a processing facility’s shift, in order to observe pathogen dissemination during the course of operations, results may not necessarily identify sites where a pathogen persists, but rather only identify sites where it collects during the course of operations. All this points back to the question of “what is the best paradigm for studying pathogen persistence in a facility?” This is up for discussion, but it seems that a logical first step is actually to conduct a study that establishes long-term strain contamination patterns via longitudinal sampling. This study could then be followed up with a short- term intensified study (i.e. one week) that tracks contamination fluctuations throughout the day and over the course of the week at pre-operational and mid-shift stages. Intense sampling of incoming raw materials, packaging, and employee surfaces would also need to be performed to assess whether strains are being reintroduced throughout the day. If repeated reintroduction of strains would be implicated, then it would be interesting to sample other processing and primary agricultural environments further up in the food production chain to note whether the strain of interest is persisting in those locations. However, one must keep in mind that the paradigm described above is an ideal scenario. In reality, obtaining the desired results from a study of this kind may not be as feasible as initially perceived. Indeed, Hu et al. (4) employed the strategy of intense daily

238

Texas Tech University, Alex Brandt, May 2014 sampling in a cold-smoked salmon processing operation, and had difficulty determining the practices that were responsible for the contamination patterns they observed. They also observed that contamination patterns and prevalences for Listeria were highly variable between days and within a given day. Thus, before undertaking a study of this magnitude, the feasibility of the study should be evaluated, and methods used for sampling and observation may need to be considered to ensure they provide useable data. In addition, another key aspect to address in future research is accounting for the presence of multiple strains of different subtypes at a certain sampling site. Oftentimes only one isolate per sample is selected for subtyping to keep study costs within budget. Thus, some strains that persist at the site may not be accounted for due to methodological limitations. Therefore, careful consideration needs to be given in future studies to how isolates are chosen for subtyping and how to minimize the influence that cultural methods can have on the strains that are isolated from a sample. This would drastically improve confidence in persistence determinations especially in facilities and sites where there are diverse sets of strains present, which may all compete for detection via current methods. Other critically-needed research is that which identifies strain-specific risk factors for persistence. Characterization of gene expression profiles of persistent strains in their native sites of persistence would greatly aid this identification. With technologies currently available (and becoming less expensive) that allow deep RNA sequencing (11), the elucidation of a strain’s transcriptome while present in these persistent sites may provide the researcher the most clear perspective of how the organism responds to its environment under these conditions. This may also help in the development of targeted mitigation strategies that focus on means to block the chain of gene expression and help prevent the enactment of the resulting persistence phenotypes. Determining the contribution of facility-specific risk factors to persistence, such as facility age, production practices, sanitation practices, and inspection types also seems to be of interest. Interestingly, the two facilities with the oldest structural areas both had higher prevalences of L. monocytogenes and non-pathogenic Listeria during the initial longitudinal period. However, these facilities were also custom-exempt processing facilities, and this also could have been a factor. Thus, meta-analyses using larger sets of data from multiple studies, like those described above, are needed to parse out risk factors

239

Texas Tech University, Alex Brandt, May 2014 on the facility level. Such information would likely help the compilation of best practices for food production facilities, and could help them realize how their risks for persistence can fluctuate as the facility ages or production practices change. This study’s findings also stress the importance that environmental monitoring programs have for assessing the control of pathogen contamination in food processing facility premises, and for aiding in seek-and-destroy strategies to eliminate pathogen harborage sites. These strategies are important for facility managers, but also have utility in research studies, as demonstrated herein. In this study, environmental monitoring procedures permitted verification of control measure effectiveness, and without them as key tools in the toolbox of available techniques, this study would not have been possible. In conclusion, progress is being made that is helping us gain a better understanding of foodborne pathogen persistence at multiple levels throughout the food chain, but several critical areas remain to be explored. With more studies conducted on pathogen persistence in different environments, more studies on risk factors for persistence, use of new and standardized methods to define persistence, and development of study paradigms that attribute strain recurrence to persistence and rule out confounding factors, a much clearer perspective of pathogen persistence should emerge. With this knowledge, effective counteraction measures can be developed, and should help mitigate at least some of the food safety threat that is associated with the phenomenon.

References 1. Carpentier B, Chassaing D. 2004. Interactions in biofilms between Listeria monocytogenes and resident microorganisms from food industry premises. Int. J. Food Microbiol. 97:111-122.

2. Hoelzer K, Sauders BD, Sanchez MD, Olsen PT, Pickett MM, Mangione KJ, Rice DH, Corby J, Stich S, Fortes ED, Roof SE, Grohn YT, Wiedmann M, Oliver HF. 2011. Prevalence, distribution, and diversity of Listeria monocytogenes in retail environments, focusing on small establishments and establishments with a history of failed inspections. J. Food Prot. 74:1083-1095.

3. Hoffman AD, Gall KL, Norton DM, Wiedmann M. 2003. Listeria monocytogenes contamination patterns for the smoked fish processing environment and for raw fish. J. Food Prot. 66:52-60.

240

Texas Tech University, Alex Brandt, May 2014

4. Hu YW, Gall K, Ho A, Ivanek R, Grohn YT, Wiedmann M. 2006. Daily variability of Listeria contamination patterns in a cold-smoked salmon processing operation. J. Food Prot. 69:2123-2133.

5. Kabuki DY, Kuaye AY, Wiedmann M, Boor KJ. 2004. Molecular subtyping and tracking of Listeria monocytogenes in latin-style fresh-cheese processing plants. J. Dairy Sci. 87:2803-2812.

6. Lappi VR, Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Longitudinal studies on Listeria in smoked fish plants: impact of intervention strategies on contamination patterns. J. Food Prot. 67:2500-2514.

7. Malley TJ, Stasiewicz MJ, Grohn YT, Roof S, Warchocki S, Nightingale K, Wiedmann M. 2013. Implementation of statistical tools to support identification and management of persistent Listeria monocytogenes contamination in smoked fish processing plants. J. Food Prot. 76:796-811.

8. Martin NH, Murphy SC, Ralyea RD, Wiedmann M, Boor KJ. 2011. When cheese gets the blues: Pseudomonas fluorescens as the causative agent of cheese spoilage. J. Dairy Sci. 94:3176-3183.

9. Midelet G, Kobilinsky A, Carpentier B. 2006. Construction and analysis of fractional multifactorial designs to study attachment strength and transfer of Listeria monocytogenes from pure or mixed biofilms after contact with a solid model food. Appl. Environ. Microbiol. 72:2313-2321.

10. Oliver HF, Ford T. 2013. Listeria monocytogenes in retail delis – prevalence, transmission and control strategies International Association for Food Protection 2013 Annual Meeting, Symposium S26, Charlotte, NC.

11. Oliver HF, Orsi RH, Ponnala L, Keich U, Wang W, Sun Q, Cartinhour SW, Filiatrault MJ, Wiedmann M, Boor KJ. 2009. Deep RNA sequencing of L. monocytogenes reveals overlapping and extensive stationary phase and sigma B- dependent transcriptomes, including multiple highly transcribed noncoding RNAs. BMC Genom. 10:641.

12. Ortiz S, Lopez V, Villatoro D, Lopez P, Davila JC, Martinez-Suarez JV. 2010. A 3-year surveillance of the genetic diversity and persistence of Listeria monocytogenes in an Iberian pig slaughterhouse and processing plant. Foodborne Pathog. Dis. 7:1177-1184.

241

Texas Tech University, Alex Brandt, May 2014

13. Ralyea RD, Wiedmann M, Boor KJ. 1998. Bacterial tracking in a dairy production system using phenotypic and ribotyping methods. J. Food Prot. 61:1336-1340.

14. Sauders BD, Sanchez MD, Rice DH, Corby J, Stich S, Fortes ED, Roof SE, Wiedmann M. 2009. Prevalence and molecular diversity of Listeria monocytogenes in retail establishments. J. Food Prot. 72:2337-2349.

15. Thimothe J, Nightingale KK, Gall K, Scott VN, Wiedmann M. 2004. Tracking of Listeria monocytogenes in smoked fish processing plants. J. Food Prot. 67:328- 341.

16. Williams SK. 2010. Listeria monocytogenes and other Listeria species in small and very small ready-to-eat meat processing plants. M.S. Thesis: Department of Animal Sciences, Colorado State University, 168 p.

242

Texas Tech University, Alex Brandt, May 2014

APPENDIX A MASTER SAMPLE SITE LIST FOR INITIAL SIX-MONTH SAMPLING PERIOD IN FACILITY A

1. Chopper/Grinder 29. Packaging Room Stress Mats 2. Grind Room Sink 30. Pack Table Surface and Scale 3. Grind Room Drains/Trench Drain/Floor 31. Truck Rods Post-Cook 4. Grind Room Scales 32. Truck Framework and Wheels Post-Cook 5. Pre-Cook Employee Apron and Gloves 33. Pack Room Door Mat/Dip Stations 6. Pre-Cook Employee Boots 34. Door Packaging —> Cooler (Both Sides) 7. Door Grinding —> Packaging (Both Sides) 35. Door Receiving —> Packaging (Both Sides) 8. Metal Table in Stuffing Room 36. Post-Cook Employee Apron and Gloves 9. Truck Rods Pre-Cook 37. Post-Cook Employee Boots 10. Stuffing Room Drains/Floor 38. Door Cooler 1—> Hallway 1 (Both Sides) 11. Stuffing Room Stress Mats 39. Cooler 1 Drain/Floor 12. Stuffer/Table 40. Cooler 1 Pallets 13. Metal Combo Bins 41. Cooler 1 Shelves 14. Door Stuffing —> Cooking (Both Sides) 42. Door Hallway 2 —> Packaging (Both Sides) 15. Hoses 43. Hallway Door Mat and Floor 16. Tumbling Room Pallets 44. Raw Product Pallets 17. Cook Room Floor (Water Puddles) 45. Raw Product Floor 18. Cooling Shower Interiors 46. Receiving Floor 19. Cooking Room Drain/Trench Drains 47. Pallet Jack Handle 20. Smokehouse Handles 48. Door Cooler 2—> Hallway 1 (Both Sides) 21. Floor/Gap Between Smokehouses 49. Cooler 2 Shelves/Floor 22. Door Cooking —> Packaging (Both Sides) 50. Door Cooler 3—> Hallway 1 (Both Sides) 23. Cooking Room Sinks 51. Cooler 3 Shelves/Floor 24. Packaging Room Sink 52. Finished Products 25. Packaging Room Drains/Floor 53. Retail Counter Slicer 26. Squeegees—Rubber and Cloth 54. Retail Counter Display Case Interior 27. Red Tubs for Finished Product 55. Retail Counter Scale 28. Packaging Machine—Both Sides

243

Texas Tech University, Alex Brandt, May 2014

APPENDIX B MASTER SAMPLE SITE LIST FOR INITIAL SIX-MONTH SAMPLING PERIOD IN FACILITY B

1. Processing Room Drain #1 29. Scales 2. Truck Wheels and Framework Pre-Cook 30. Mixer #2 3. Mixer #1 31. Grinders #1 and #2 4. Smokehouse #1 Controls and Handle 32. Gray Plastic Tubs 5. Processing Room Drain #2 33. Processing Hand Sink #2 6. Vacuum Packagers 34. Processing Room Drain #4 7. Band Saw and Table 35. Cooler 2 Floor (Crack) 8. Tumbler Exterior 36. Truck Framework and Wheels Post-Cook 9. Truck Rods 37. Tubs With Raw Product in Cooler 2 10. Smokehouse #2 Bottom and Ramp 38. Cooler 2 Cooling Fans 11. Smokehouse #3 Bottom and Ramp 39. Cooler 1 Floor 12. Squeegees 40. Metal Carts in Cooler 1 (Interior) 13. Door Processing -> Break Room (Both Sides) 41. Metal Shelving in Cooler 1 14. Injector Exterior and Belt 42. Cooler 1 Drain 15. White Plastic Tubs 43. Processing Employee Apron 16. Floor Next to Smokehouses 44. Processing Employee Boots 17. Trench Drain #1 45. Kill Floor Drain 18. Door Processing -> Cooler 2 (Both Sides) 46. Kill Floor Floor 19. Carts for Gray Tubs (Wheels and Framework) 47. Kill Floor Employee Apron 20. Trench Drain #2 48. Kill Floor Employee Boots 21. Hoses 49. Kill Floor Hose 22. Processing Hand Sink #1 50. Kill Floor Small Handsaw 23. Processing Room Drain #3 51. Holding Pens Metal 24. Ice Machine 52. Retail Finished Product 25. Door Processing -> Freezer (Both Sides) 53. Retail Large Display Case 26. Wooden Table (Surface and Legs) 54. Retail Floor Display Case 27. Stuffing Machine and Hopper 55. Retail Sliding Door Display Case 28. Floor Under Tables

244

Texas Tech University, Alex Brandt, May 2014

APPENDIX C MASTER SAMPLE SITE LIST FOR INITIAL SIX-MONTH SAMPLING PERIOD IN FACILITY C

1. Packaging Tables 29. Cradle 2. Band Saw and Table 30. Kill Floor Hose 3. Slicer 31. Kill Floor Employee Aprons 4. Grinder With Hopper 32. Stunning Chute Pipework 5. Processing Area Drain 33. Shackling Control Button 6. Processing Room Sinks 34. Maintenance Area Floor 7. Processing Room Hose 35. Door Maintenance -> Cooler 1 (Both Sides) 8. Fabrication Tables and Hopper 36. Cooler 1 Draining Trap Door 9. Blue Tubs for Holding Product 37. Cooler 1 Walls 10. Fabrication Room Stress Mats 38. Cooler 1 Floor 11. Truck Wheels and Framework 39. Cooler 2 Drain 12. Hooks 40. Cooler 2 Floor 13. Door Processing -> Smoking (Both Sides) 41. Cooler 2 Walls 14. Door Processing -> Outside Slab (Both Sides) 42. Control Button 15. Floor Next To Square Drain (Drain Is Closed) 43. Cooler 3 Floor 16. Brooms/Mops/Squeegees in Kill Floor Area 44. Cooler 3 Walls 17. Smoking Area Drain 45. Blood and Offal Collection Barrels 18. Smokehouse Door and Handle 46. Cattle Pens and Chutes Outside 19. Smoking Area Floor and Floor-Wall Junction 47. Restroom -> Smoking Area Door (Both Sides) 20. Buckets in Smoking Area and Coolers 48. Processing Employee Boots 21. Smoking Area Sink 49. Processing Employee Apron 22. Kill Floor Employee Boots 50. Slab Area Outside 23. Large Circular Saw 51. Freezer Door (Both Sides) 24. Small Hand Saw 52. Crates Holding Finished Product (Fabrication) 25. Door Kill Floor -> Outside Slab (Both Sides) 53. Crates Holding Finished Product (Freezer) 26. Large Blood Catch Drain 54. Cross Bar With Hooks 27. Kill Floor Floor 55. Beef Carcass Samples (3 Carcass Halves Each) 28. Door Kill Floor -> Cooler 2 (Both Sides)

245

Texas Tech University, Alex Brandt, May 2014

APPENDIX D MASTER SAMPLE SITE LIST FOR INITIAL SIX-MONTH SAMPLING PERIOD IN FACILITY D

1. Vacuum Packager Granite Surface 29. Processing Worker Shoes 2. Processing Area Scale #1 30. Carcass Cooler Drain 3. Gray Plastic Tubs 31. Carcass Cooler Hose 4. Grinder #1 32. Carcass Cooler Floor 5. Grinder #2 33. Carcass Cooler Wall 6. Table Next to Processing Area Scale #1 34. Drip Area Floor 7. Packaging Tables 35. Door Drip Area -> Kill Floor (Both Sides) 8. Band Saw and Table 36. Stunning Chute Area 9. Table Between Band Saw and Grinder #2 37. Stunning Area Drain 10. Basin Area Next to Band Saw 38. Bleeding Area Floor 11. Cutting Tables 39. Door Bleeding -> Exterior (Inside Only) 12. Metal Carts (Interior) 40. Kill Floor Area Hand Sinks 13. Moveable Table/Pedestal 41. Bleeding Area Walls 14. Processing Area Washing Sinks 42. Kill Floor Drain #1 15. Processing Area Hand Sinks 43. Cradle 16. Door Freezer -> Processing (Both Sides) 44. Bleeding Area Yellow Hose 17. Processing Area Drain #1 45. Offal Truck #1 18. Processing Area Drain #2 46. Kill Floor Area White Hose 19. Carts for Gray Tubs (Wheels and Framework) 47. Kill Floor Floor 20. Door Office -> Processing (Both Sides) 48. Kill Floor Drain #2 21. Processing Area Scale #2 49. Offal Truck #2 22. Metal Equipment Racks 50. Plastic Shovels 23. Processing Area Floor 51. Platform 24. Freezer Floor 52. Large Circular Saw and Small Handsaw 25. Door Processing -> Cooler (Both Sides) 53. Kill Floor Employee Apron 26. Door Processing -> Break Room (Both Sides) 54. Kill Floor Employee Boots 27. Door Restroom -> Break Room (Both Sides) 55. Beef Carcass Samples (3 Carcass Halves Each) 28. Processing Worker Apron

246

Texas Tech University, Alex Brandt, May 2014

APPENDIX E INITIAL SIX-MONTH SAMPLING PERIOD LISTERIA CONTAMINATION PATTERNS IN FACILITY A

Sample Site 05/27/2011 06/24/2011 08/16/2011 09/16/2011 10/14/2011 11/11/2011

Drain and Floor Sites LM 116-239-S-2 LM 116-239-S-2 LM 116-239-S-2 - - LW AT 27 Site 3 LW AT 69 LW AT 69 LM 116-239-S-2 LM 116-239-S-2 LM DUP-1052A LI AT 56 LM 116-239-S-2 - Site 10 LW AT 69 LI AT 56 LM DUP-1052A LM DUP-1052A LM DUP-1052A LM 116-239-S-2 Site 11 - LM 116-239-S-2 LW AT 69 LI AT 56 LI AT 56 LI AT 56 Site 17 - - - LI AT 56 - -

Site 19 - LM DUP-1048A LM DUP- 1052A LI AT 56 - LI AT 56 Site 21 - LI AT 56 LM DUP- 1052A - - - Site 25 - - - LW AT 69 LW AT 69 LW AT 69

Site 39 ------

Site 43 ------Site 45 - - LW AT 32 - - - Site 46 ------

Site 49 ------Site 51 ------Non-Food Contact Environmental Sites Site 2 ------

Site 7 ------LM DUP-1052A LM DUP-1052A - - LM DUP-1052A - Site 14 LI AT 56 Site 15 ------Site 16 - - - LI AT 56 LM DUP-1052A -

Site 18 ------Site 20 ------Site 22 ------Site 23 ------

Site 24 ------Site 26 ------Site 32 ------Site 33 ------

Site 34 ------Site 35 ------

------Site 38 247

Texas Tech University, Alex Brandt, May 2014

Sample Site 05/27/2011 06/24/2011 08/16/2011 09/16/2011 10/14/2011 11/11/2011 Non-Food Contact Environmental Sites Site 40 ------Site 41 ------Site 42 ------Site 44 - - - - - LW AT 69

Site 48 ------Site 50 ------Site 54 ------Employee Surfaces and Employee Contact Sites LM DUP-1052A LM DUP-1052A LM 116-239-S-2 - LM DUP-1052A - Site 5 LI AT 37 LI AT 56 LM DUP-1052A LM DUP-1052A - - LM DUP-1052A LW AT 27 Site 6 LI AT 56 Site 29 ------Site 36 ------

Site 37 - - - - LM DUP-1052A -

Site 47 ------Food Contact Sites and Food Samples Site 1 ------

Site 4 LI AT 37 - - - LM 116-239-S-2 - LM DUP-1052A LM DUP-1052A - - LM 116-239-S-2 - Site 8 LI AT 56 Site 9 ------LM DUP-1052A Site 12 LM DUP-1052A - LI AT 56 - - LI AT 56 Site 13 LM 116-239-S-2 - LW AT 19 LM DUP-1052A LM DUP-1052A - Site 27 ------Site 28 ------Site 30 ------LM DUP-1052A - - - - - Site 31 LI AT 56 Site 52 ------Site 53 ------Site 55 ------

† The symbol “-” corresponds to a negative result for the particular site on the respective sampling date. Samples positive for Listeria are described by two capital letters which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), Listeria welshimeri (LW). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1052A) and the Listeria spp. sigB allelic type (e.g. AT 56) follow these species abbreviations.

248

Texas Tech University, Alex Brandt, May 2014

APPENDIX F INITIAL SIX-MONTH SAMPLING PERIOD LISTERIA CONTAMINATION PATTERNS IN FACILITY B

Sample Site 06/10/2011 07/12/2011 09/09/2011 09/30/2011 10/28/2011 12/02/2011

Drain and Floor Sites

Site 1 LM DUP-1062B LM DUP-1062E - - - - LM DUP-1062B LM DUP-1062E - - - LI AT 45 Site 5 LI AT 23 LI AT 23 LM DUP-1062E LI AT 38 LI AT 31 LI AT 38 LI AT 6 LI AT 6 Site 16 LI AT 44 LM DUP-1062B LM DUP-1062B LM DUP-1053E LI AT 6 LI AT 11 LI AT 38 Site 17 LI AT 94 LI AT 23 LI AT 17 LM DUP-1062B LM DUP-1062B - - - LM DUP-1053E Site 20 LI AT 23 Site 23 - LM DUP-1062B - - - LI AT 11 LM DUP-1062E - - - - - Site 28 LI AT 6 LM DUP-1062B - LI AT 6 - - - Site 34 LI AT 38 LM DUP-1062E LM DUP-1062E LM DUP-1053E LM DUP-1062B LI AT 45 LM DUP-1062B Site 35 LI AT 23 LI AT 23 LI AT 6 LI AT 11 LM DUP-1053E LM DUP-1042C LM DUP-1062B LI AT 38 LI AT 45 LI AT 6 Site 39 LI AT 6 LI AT 11 LI AT 11 LM DUP-1062E LM DUP-1062B LM DUP-1053E LM DUP-1062E LI AT 6 LI AT 11 Site 42 LI AT 38 LI AT 45 LI AT 45 LI AT 11 LM DUP-1062B LI AT 26 LI AT 11 LM DUP-1062E LM DUP-1053E - Site 45 LI AT 6 LM DUP-1062B LM DUP-1053E LM DUP-1062B - LI AT 6 LM DUP-1062E Site 46 LI AT 11 LI AT 11 LI AT 6 Non-Food Contact Environmental Sites Site 2 LI AT 6 LM DUP-1053E - - - -

Site 4 ------Site 8 ------LM DUP-1062E LI AT 6 - LI AT 6 - - Site 10 LI AT 6 Site 11 - LI AT 23 LI AT 6 LI AT 6 LI AT 6 - LM DUP-1062B LM DUP-1062B LI AT 23 LI AT 23 LM DUP-1053E - Site 12 LI AT 23 Site 13 - - - - - Site 18 LM DUP-1062E - - - - - LM DUP-1062E LM DUP-1062E - LI AT 11 - - Site 19 LI AT 109 Site 21 ------Site 22 ------Site 24 - - - - - LI AT 30 Site 25 ------Site 33 - - LI AT 11 - - - Site 36 - - NS - - NS Site 38 ------249

Texas Tech University, Alex Brandt, May 2014

Sample Site 06/10/2011 07/12/2011 09/09/2011 09/30/2011 10/28/2011 12/02/2011 Non-Food Contact Environmental Sites Site 41 - - - NS NS - Site 49 LI AT 23 - LI AT 11 - - - Site 51 - LI AT 37 - - - -

------Site 53 Site 54 ------

Site 55 ------Employee Surfaces and Employee Contact Sites

------Site 6 Site 43 LI AT 23 - LI AT 11 LI AT 11 - - LM DUP-1062E LM DUP-1062E LM DUP-1062E LI AT 11 - - Site 44 LI AT 6 LI AT 31 Site 47 - LI AT 11 - - - - LM DUP-1062E Site 48 LI AT 23 NS - - - LI AT 23 Food Contact Sites and Food Samples

Site 3 ------Site 7 - - LI AT 11 - - LI AT 37 Site 9 ------

Site 14 ------

Site 15 ------Site 26 LM DUP-1062E LI AT 23 - - - - Site 27 - - - LI AT 11 - -

Site 29 - LI AT 6 LI AT 11 - LI AT 11 -

Site 30 ------Site 31 - - - LI AT 11 - LI AT 37 Site 32 - - - LI AT 11 - -

Site 37 - - LI AT 11 LM DUP-1053E - -

Site 40 - - LM DUP -1062B LI AT 11 - - Site 50 ------Site 52 ------

† The symbol “-” corresponds to a negative result for the particular site on the respective sampling date. “NS” denotes that the sample site was not sampled on that particular sampling date. Samples positive for Listeria are described by two capital letters which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), and Listeria innocua (LI). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1062B) and the Listeria spp. sigB allelic type (e.g. AT 6) follow these species abbreviations.

250

Texas Tech University, Alex Brandt, May 2014

APPENDIX G INITIAL SIX-MONTH SAMPLING PERIOD LISTERIA CONTAMINATION PATTERNS IN FACILITY C

Sample Site 06/03/2011 07/05/2011 08/24/2011 09/23/2011 10/21/2011 11/16/2011 Drain and Floor Sites LM DUP-1042B LI AT 70 LI AT 71 LI AT 6 LI AT 6 LM DUP-1042B Site 5 LI AT 6 Site 10 - LM DUP-1042B - - LI AT 70 - LM DUP-1042B LM DUP-1042B LM DUP- 1042B LM DUP-1042B LI AT 6 - Site 15 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1057B LI AT 71 LI AT 71 Site 17 LI AT 70 LI AT 71 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 70 LI AT 26 Site 19 LI AT 70 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 53 LI AT 70 Site 26 LI AT 71 LI AT 70 LI AT 70 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 70 LM DUP-1042B Site 27 LI AT 37 LI AT 26 LI AT 70 LI AT 70 Site 34 - - LM DUP -1042B LM DUP-1042B LI AT 70 LI AT 70 Site 38 - - - LI AT 109 - LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 109 Site 39 LI AT 71 LI AT 23 LI AT 70 LI AT 70 LI AT 71 LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 109 LI AT 71 LI AT 70 Site 40 LI AT 70 LI AT 26 LI AT 70 LM DUP-1042B LM DUP-1042B LM DUP-1042B LI AT 109 - LI AT 70 Site 43 LI AT 70 HLI AT 141 LI AT 70 Non-Food Contact Environmental Sites Site 6 ------

Site 7 - - LI AT 70 - - - LM DUP-1042B - - - LI AT 70 - Site 11 LI AT 70 LM DUP-1042B LM DUP-1042B - - LM DUP-1042B LI AT 70 Site 13 LI AT 70 LI AT 71 Site 14 ------LM DUP-1042B LI AT 70 - - LI AT 70 LI AT 70 Site 16 LI AT 23 LM DUP-1042B LM DUP-1042B - - LI AT 70 LI AT 70 Site 18 LI AT 70

- - LM DUP-1042B - - - Site 20 Site 21 LM DUP-1057B - - - - -

Site 25 ------Site 28 - - - LI AT 70 - - Site 29 - LI AT 30 - - - - LM DUP- 1042B - - - - - Site 30 LI AT 70 Site 32 LI AT 11 - - LI AT 71 - - Site 35 ------Site 36 - LM DUP-1042B LM DUP-1042B - - -

Site 37 ------251

Texas Tech University, Alex Brandt, May 2014

Sample Site 06/03/2011 07/05/2011 08/24/2011 09/23/2011 10/21/2011 11/16/2011 Non-Food Contact Environmental Sites Site 41 ------Site 44 - - LI AT 70 - - - Site 45 ------Site 46 ------LM DUP-1042B LI AT 70 - LM DUP-1042B - - Site 47 LI AT 71 Site 50 LI AT 31 - - LI AT 70 - LI AT 70

------Site 51 Site 52 ------

------Site 53 Site 54 - - LI AT 6 - - -

Employee Surfaces and Employee Contact Sites LM DUP-1042B Site 22 - LI AT 71 LI AT 26 LI AT 70 - LI AT 70 Site 31 - - - LW AT NEW - -

Site 33 - - LI AT 71 - - - Site 42 LI AT 30 - - - - - LM DUP-1042B Site 48 LM DUP-1042B LI AT 70 LM DUP-1042B - LM DUP-1042B LI AT 70 Site 49 ------Food Contact Sites and Food Samples Site 1 - - LI AT 23 - - - Site 2 - - LI AT 23 - - -

Site 3 ------Site 4 ------Site 8 - - LI AT 23 - - - Site 9 ------

Site 12 ------Site 23 ------Site 24 - - LI AT 26 - - LI AT 71 Site 55 - - LI AT 30 - LI AT 37 -

† The symbol “-” corresponds to a negative result for the particular site on the respective sampling date. Samples positive for Listeria are described by two or three capital letters which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), hemolytic Listeria innocua (HLI), and Listeria welshimeri (LW). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP-1042B) and the Listeria spp. sigB allelic type (e.g. AT 70) follow these species abbreviations.

252

Texas Tech University, Alex Brandt, May 2014

APPENDIX H INITIAL SIX-MONTH SAMPLING PERIOD LISTERIA CONTAMINATION PATTERNS IN FACILITY D

Sample Site 06/17/2011 07/19/2011 09/12/2011 10/07/2011 11/04/2011 12/09/2011 Drain and Floor Sites Site 17 ------Site 18 - - - LM DUP-1030B - -

Site 23 ------

Site 24 ------Site 30 - - - - - LW AT 129 Site 32 - - - - NS -

Site 34 ------LM DUP-1030B LI AT 6 - LW AT NEW - - Site 37 LI AT 109 LM DUP-1030A LS AT 121 - LW AT 129 - - Site 38 LI AT 6 LM DUP-1030B - - - - - Site 42 LI AT 6

Site 47 LI AT 6 - - - - LI AT 6 Site 48 ------Non-Food Contact Environmental Sites Site 10 ------

Site 13 ------Site 14 ------

------Site 15 Site 16 ------

Site 19 ------Site 20 ------

------Site 22 Site 25 - - LW AT 129 LW AT 129 LW AT 129 -

Site 26 ------Site 27 ------Site 31 - - - - - LW AT 129

Site 33 ------

Site 35 ------Site 36 - LI AT 6 - - - LI AT 31 Site 39 ------

Site 40 ------253

Texas Tech University, Alex Brandt, May 2014

Sample Site 06/17/2011 07/19/2011 09/12/2011 10/07/2011 11/04/2011 12/09/2011 Non-Food Contact Environmental Sites Site 41 ------

Site 43 ------Site 44 LI AT 124 - - - LI AT 53 -

------Site 45 Site 46 ------

Site 49 ------

------Site 50 LI AT 6 - - - - - Site 51 Employee Surfaces and Employee Contact Sites Site 28 ------Site 29 ------Site 53 ------

Site 54 - - LI AT 53 - - - Food Contact Sites and Food Samples Site 1 ------Site 2 ------

Site 3 ------

Site 4 ------Site 5 ------Site 6 ------

Site 7 ------

Site 8 ------Site 9 ------Site 11 ------

Site 12 ------

Site 21 ------Site 52 ------Site 55 ------

† The symbol “-” corresponds to a negative result for the particular site on the respective sampling date. “NS” denotes that the sample site was not sampled on that particular sampling date. Samples positive for Listeria are described by two capital letters which correspond to the abbreviation for the species of the organism(s) isolated: Listeria monocytogenes (LM), Listeria innocua (LI), Listeria seeligeri (LS), and Listeria welshimeri (LW). The Listeria monocytogenes molecular subtype (e.g. EcoRI ribotype DUP- 1030B) and the Listeria spp. sigB allelic type (e.g. AT 129) follow these species abbreviations.

254

Texas Tech University, Alex Brandt, May 2014

APPENDIX I FACILITY A ISOLATION PATTERNS FOR LISTERIA MONOCYTOGENES RIBOTYPE DUP-1052A

255

Texas Tech University, Alex Brandt, May 2014

APPENDIX J FACILITY A ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 56

256

Texas Tech University, Alex Brandt, May 2014

APPENDIX K FACILITY A ISOLATION PATTERNS FOR LISTERIA WELSHIMERI SIGB ALLELIC TYPE 69

257

Texas Tech University, Alex Brandt, May 2014

APPENDIX L FACILITY B ISOLATION PATTERNS FOR LISTERIA MONOCYTOGENES RIBOTYPE DUP-1053E

258

Texas Tech University, Alex Brandt, May 2014

APPENDIX M FACILITY B ISOLATION PATTERNS FOR LISTERIA MONOCYTOGENES RIBOTYPE DUP-1062B

259

Texas Tech University, Alex Brandt, May 2014

APPENDIX N FACILITY B ISOLATION PATTERNS FOR LISTERIA MONOCYTOGENES RIBOTYPE DUP-1062E

260

Texas Tech University, Alex Brandt, May 2014

APPENDIX O FACILITY B ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 6

261

Texas Tech University, Alex Brandt, May 2014

APPENDIX P FACILITY B ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 11

262

Texas Tech University, Alex Brandt, May 2014

APPENDIX Q FACILITY B ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 23

263

Texas Tech University, Alex Brandt, May 2014

APPENDIX R FACILITY B ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 38

264

Texas Tech University, Alex Brandt, May 2014

APPENDIX S FACILITY B ISOLATION PATTERNS FOR SALMONELLA ENTERICA SEROTYPE-PULSOTYPE HEIDELBERG

265

Texas Tech University, Alex Brandt, May 2014

APPENDIX T FACILITY C ISOLATION PATTERNS FOR LISTERIA MONCYTOGENES RIBOTYPE DUP-1042B

266

Texas Tech University, Alex Brandt, May 2014

APPENDIX U FACILITY C ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 70

267

Texas Tech University, Alex Brandt, May 2014

APPENDIX V FACILITY C ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 71

268

Texas Tech University, Alex Brandt, May 2014

APPENDIX W FACILITY C ISOLATION PATTERNS FOR SALMONELLA ENTERICA SEROTYPE-PULSOTYPE III 50:R:Z

269

Texas Tech University, Alex Brandt, May 2014

APPENDIX X FACILITY C ISOLATION PATTERNS FOR SALMONELLA ENTERICA SEROTYPE-PULSOTYPE INFANTIS

270

Texas Tech University, Alex Brandt, May 2014

APPENDIX Y FACILITY C ISOLATION PATTERNS FOR SALMONELLA ENTERICA SEROTYPE-PULSOTYPE UGANDA

271

Texas Tech University, Alex Brandt, May 2014

APPENDIX Z FACILITY D ISOLATION PATTERNS FOR LISTERIA MONOCYTOGENES RIBOTYPE DUP-1030B

272

Texas Tech University, Alex Brandt, May 2014

APPENDIX AA FACILITY D ISOLATION PATTERNS FOR LISTERIA INNOCUA SIGB ALLELIC TYPE 6

273

Texas Tech University, Alex Brandt, May 2014

APPENDIX AB FACILITY D ISOLATION PATTERNS FOR LISTERIA WELSHIMERI SIGB ALLELIC TYPE 129

274

Texas Tech University, Alex Brandt, May 2014

APPENDIX AC FACILITY D ISOLATION PATTERNS FOR SALMONELLA ENTERICA SEROTYPE-PULSOTYPE TYPHIMURIUM VAR 5-

275

Texas Tech University, Alex Brandt, May 2014

APPENDIX AD ENROLLMENT LETTER MAILED TO POTENTIAL FACILITIES AT STUDY OUTSET

276

Texas Tech University, Alex Brandt, May 2014

APPENDIX AE ENROLLMENT FORM MAILED TO POTENTIAL FACILITIES AT STUDY OUTSET

277

Texas Tech University, Alex Brandt, May 2014

APPENDIX AF KNOWLEDGE ASSESSMENT USED AT PRE- AND POST- TRAINING TIME POINTS

278

Texas Tech University, Alex Brandt, May 2014

279

Texas Tech University, Alex Brandt, May 2014

280

Texas Tech University, Alex Brandt, May 2014

281

Texas Tech University, Alex Brandt, May 2014

282

Texas Tech University, Alex Brandt, May 2014

APPENDIX AG FACILITY B 2012 BEHAVIORAL ASSESSMENT QUESTIONNAIRE

283

Texas Tech University, Alex Brandt, May 2014

284

Texas Tech University, Alex Brandt, May 2014

APPENDIX AH FACILITY C 2012 BEHAVIORAL ASSESSMENT QUESTIONNAIRE

285

Texas Tech University, Alex Brandt, May 2014

286

Texas Tech University, Alex Brandt, May 2014

APPENDIX AI FACILITY B 2012 DIRECT OBSERVATION FORM

287

Texas Tech University, Alex Brandt, May 2014

288

Texas Tech University, Alex Brandt, May 2014

APPENDIX AJ FACILITY C 2012 DIRECT OBSERVATION FORM

289

Texas Tech University, Alex Brandt, May 2014

290

Texas Tech University, Alex Brandt, May 2014

APPENDIX AK FACILITY B 2013 BEHAVIORAL ASSESSMENT QUESTIONNAIRE

291

Texas Tech University, Alex Brandt, May 2014

292

Texas Tech University, Alex Brandt, May 2014

APPENDIX AL FACILITY C 2013 BEHAVIORAL ASSESSMENT QUESTIONNAIRE

293

Texas Tech University, Alex Brandt, May 2014

294

Texas Tech University, Alex Brandt, May 2014

APPENDIX AM FACILITY B 2013 DIRECT OBSERVATION FORM

295

Texas Tech University, Alex Brandt, May 2014

296

Texas Tech University, Alex Brandt, May 2014

297

Texas Tech University, Alex Brandt, May 2014

APPENDIX AN FACILITY C 2013 DIRECT OBSERVATION FORM

298

Texas Tech University, Alex Brandt, May 2014

299

Texas Tech University, Alex Brandt, May 2014

300