This item is only available as the following downloads:
1 THE ROLE OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE I N THE INTERACTIONS OF Salmonella Typhimurium O N THE TOMATO SURFACE By MARIANNE KATHRYN FATICA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVE RSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Marianne Kathryn Fatica
3 To my brother Michael, a person who inspires an d encourages those around him to continually strive for more
4 ACKNOWLEDGMENTS I would like to express my gratitude to my major advisor, Dr. Keit h Schneider for his support and patience as well as all of the opportunities he has provided me during my graduate studies. Dr. Schneider not only helped introduce me to the field of food science, he also provided me with the freedom and challenge of designing my own research proje ct. The independence I was granted with my research has shaped me into a s tron ger scientist. I would like to thank Dr. Schneider for conti nually stressing the importance of the applied nature o f my work as a food scientist. I also appreciate his open door policy and constant willingness to discuss my ideas and thoughts. I would al so like to thank my committee members, Dr. Max Teplitski, Dr. Michelle Danyluk, and Dr. Anita Wrig ht for their time and input during my graduate research. I am particularly grateful to Dr. Max Teplitski for allowing me the privilege to complete my graduat e research in his lab. I ap preciate the perspective he has added to my research, as well as the time and support he has provided in discussing my project. I would also like to express my appreciation to Dr. Teplitski and his graduate students for teachin g me a number of molecular techniques that not only improved my graduate research, b ut also my scope as a research scientist I could not have completed my graduate work without the help and support of the members of the Schneider lab and the Teplitski l ab. I would like to thank Alina Balaguero and Federi co Caro in particular for their endless trips to pick up tomatoes. I also thank William Zaragoza, Clayton Cox, and Cory Krediet for their advice and support. Thanks are also extended to Dr. Max Teplits ki, Pacific Tomato Growers, DiMare Fresh, West Coast Tomato and Gadsden Tomato Company for providing tomatoes for my research.
5 I cannot go without thanking my family for their support throughout my education. I appreciate all the time and emotional sup port my family has provided me, as well as all the sacrifices they have made. I would also like to express my gratitude to the people who have become family to me, including Andrea Burkhart, Tiffiney Yen, and Lau r a Strawn. I could not have made it to gra duate school without the help and support of three amazing people, Stephen Rice, Dustin Edwards, and Rebekah Nau lt ; I thank you for being there for me Finally I would like to thank those who have been supportive of me along the way, including Lizzie Evan s, Melissa Jones, Rebecca Creasy, Courtney Buckley, Christine Booth, Rick Swain, and Gabe Shook.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURE S ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 2 LITERATURE REVIEW ................................ ................................ .......................... 18 Produce Associated Foodborne Illness ................................ ................................ ... 18 Salmonella in the Produce Environment ................................ ................................ 20 AgfD Regulated Attachment Mechanisms of Salmonella ................................ ........ 24 Research Aim and Objectives ................................ ................................ ................. 28 3 GENETIC MUTANT CONSTRUCTION AND CHARACTERIZATION .................... 32 Background ................................ ................................ ................................ ............. 32 Construction o f Salmonella Deletion Mutants ................................ ......................... 33 DNA Amplification and Visualization ................................ ................................ 33 Phage Mediated Transduction ................................ ................................ ......... 35 Construction of Genetic Complements and Empty Vector Controls ........................ 37 Cloning ................................ ................................ ................................ ............. 37 Electroporations ................................ ................................ ............................... 41 Phenotypic Characterization of Salmonella Strains ................................ ................ 42 Growth Curves of Salmonella Strains ................................ ................................ ..... 45 4 THE ROLES OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE IN THE INITIAL ATTACHMENT AND PERSISTENCE OF Salmonella Typhimurium ON THE TOMATO SURFACE ................................ ........ 55 Bac kground ................................ ................................ ................................ ............. 55 Materials and Methods ................................ ................................ ............................ 58 Culture Maintenance ................................ ................................ ........................ 58 Acquisit ion of Tomato Samples ................................ ................................ ........ 58 Evaluation of Hydrophobicity of Salmonella Strains ................................ ......... 59 Desiccation of Salmonella on the Tomato Surface ................................ ........... 60 Assessment of Tomato Inoculation Procedures ................................ ............... 61
7 Initial Attachment of S. Typhimurium 14028, MKF1, MKF2, and WJZ2 to the Tomato Surf ace ................................ ................................ ............................. 62 Competitive Co infections of Salmonella Strains on the Tomato Surface ........ 62 Persistence of Salmonella Strains in Buffered Solu tion ................................ .... 64 Persistence of Salmonella Strains on the Tomato Surface .............................. 64 Results ................................ ................................ ................................ .................... 66 Initial Attachment of Salmonella to the Tomato Surface ................................ ... 66 Persistence of Salmonella on the Tomato Surface ................................ ........... 70 Discussion ................................ ................................ ................................ .............. 73 5 THE ROLES OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE IN THE BIOFILM FORMATION OF Salmonella Typhimurium ON POLYSTYRENE, THE TOMATO SURFACE, AND TOMATO SEGMENTS AND ROLE OF THE RDAR MORPHOLOGY IN ENVIRONMENTAL FITNESS ............. 87 Background ................................ ................................ ................................ ............. 87 Materials and Methods ................................ ................................ ............................ 89 Biofilm Formation of S. Typhimurium on Polystyrene ................................ ....... 89 Biofilm Formation of S. Typhimurium MKF5, MKF1, MKF2, and WJZ2 on the Tomato Surface ................................ ................................ ....................... 90 Biofilm Formation of S Typhimurium MKF5, MKF1, MKF2, and WJZ2 on Tomato Segments ................................ ................................ ......................... 91 Characterization of Salmonella Outbreak Strains ................................ ............. 92 Results ................................ ................................ ................................ .................... 93 Discussion ................................ ................................ ................................ .............. 97 6 CONCLUSIONS AND FUTURE STUDIES ................................ ........................... 114 APPENDIX A MEDIA COMPOSITION ................................ ................................ ........................ 117 B ATTACHMENT AND BIOFILM FORMATION OF Salmonella STRAINS AFTER INCUBATION UNDER STATIC OR SHAKING CONDITIONS ............................. 120 LIST OF REFERENCES ................................ ................................ ............................. 124 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 133
8 LIST OF TABLES Table page 3 1 PCR primers ................................ ................................ ................................ ....... 47 3 2 Bacterial strains ................................ ................................ ................................ .. 48 3 3 List of plasmids ................................ ................................ ................................ ... 50 4 1 Statistical analysis of competitive index values of Salmonella recovered from the green tomato surface ................................ ................................ .................... 78 4 2 Statistical analysis of competitive index values of Salmonella recovered from the red tomato surface ................................ ................................ ........................ 78 4 3 Statistical analysis comparing competitive index values of Salmonella recovered from green versus red tomato surface ................................ ............... 78 4 4 Statistical analysis of recovered Salmonella concentrations from the green tomato surface over 6 days ................................ ................................ ................ 79 4 5 Statistical analysis of recovered concentrations of Salmonella from the red tomato surface over 6 days ................................ ................................ ................ 79 5 1 Salmonella enterica cultures obtained from produce associated outbreaks ..... 103
9 LIST OF FIGURE S Figure page 2 1 The AgfD regulation system ................................ ................................ ............... 31 3 1 Depictions of Salmonella Typhimurium strains on Co ngo red media .................. 51 3 2 Qualitative assessment of cellulose production of Salmonella Typhimurium strains ................................ ................................ ................................ ................. 52 3 3 Quantitative assessme nt of cellulose production of Salmonella Typhimurium strains ................................ ................................ ................................ ................. 53 3 4 Growth curves of Salmonella Typhimurium 14028 and mutant strains for 12 hr at 30C ................................ ................................ ................................ ........... 54 4 1 Desiccation of Salmonella Typhimurium 14028 and MKF1 on the tomato surface over time ................................ ................................ ................................ 80 4 2 Recoverable Salmonella Typhimurium from the tomato surface through spot o r dip inoculation ................................ ................................ ................................ 81 4 3 The initial attachment of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 on the tomato surface ................................ ................................ ............... 82 4 4 Competitive co infections of Salmonella strains on the tomato surface .............. 83 4 5 Persistence of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 in 0.1 M potassium phosphate buffer ................................ ................................ ..... 84 4 6 Persistence of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 on the green and red tomato surface ................................ ................................ ....... 85 4 7 Persistence of Salmonella c omplement and vector control strains on the surface of green and red tomatoes ................................ ................................ ..... 86 5 1 Biofilm formation of Salmonella on polystyrene ................................ ................ 104 5 2 Biofilm formation of Salmonella Typhimurium MKF5, MKF1, MKF2, and WJZ2 on the surface of intact green and red tomatoes ................................ .... 105 5 3 Biofilm formation of Salmonella Typhimurium complement a nd vector control strains on the surface of intact green and red tomatoes ................................ ... 106 5 4 Biofilm formation of Salmonella Typhimurium MKF5, MKF1, MKF2, and WJZ2 on green and red tomato segments ................................ ....................... 107
10 5 5 Biofilm formation of Salmonella complement and vector control strains on green and red tomato segments ................................ ................................ ....... 108 5 6 Biofilm formation of Salmonella strains on green and red tomato segments in DI water supplemented with MnCl 2 and MgSO 4 ................................ ................ 109 5 7 Depictions of Salmonella outbreak strains on Congo red media ...................... 110 5 8 Growth curves of produce related Salmonella outbreak strains for 12 hr at 30C ................................ ................................ ................................ ................. 111 5 9 Cellulose quantification of Salmonella outbreak strains ................................ .... 112 5 10 Biofilm formation of Salmonella outbreak strains on green tomato segments .. 113 B 1 Attachment of Salmonella on the surface o f Campari tomatoes after static or shaking overnight growth ................................ ................................ .................. 122 B 2 Salmonella biofilm formation in polystyrene plates after static or shaking overnight growth ................................ ................................ ............................... 123
11 LIST OF ABBREVIATIONS Amp Ampicillin ANOVA Analysis of variance BATH Bacterial adherence to hydrocarbons b dar Brown, dry and rough bp Base pair BPW Buffered peptone water c di GMP Cyclic diguanylate CDC Centers for Disease Control and Prevention CI C ompetitive index CIAP Calf intestinal alkaline phosphatase CR Congo red CW Calcofluor white DI Deionized dNTPs Deoxynucleotide triphosphates FDA Food and Drug Administration FSMA Food Safety Modernization Act GAPs Good Agricultural Practices IFAS Institut e of Food and Agricultural Sciences EBU Evans Blue Uranine EDTA Ethylene d iamine tetraacetic a cid EGTA Ethylene glycol tetraacetic a cid Gent Gentamicin Kan Kanamycin LB Luria Bertani
1 2 OD 400 Optical density at 400 nm PBS Phosphate buffered saline PCR Polymer ase chain reaction pdar Pink, dry, and rough PFGE Pulsed field gel electrophoresis rdar Red, dry, and rough RH Relative humidity RPM Revolutions per minute SAT Salt agglutination test saw Smooth and white TAE Tris Acetate EDTA T BMPs Tomato Best Management Practices Tet Tetracycline T GAPs Tomato Good Agricultural Practices TSB Trypt ic soy broth US United States USDA United States Department of Agriculture UV Ultraviolet x gal 5 bromo 4 chloro 3 indolyl D galactoside XLD Xylose lysine d esoxycholate
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philos ophy THE ROLE OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE I N THE INTERACTIONS OF Salmonella Typhimurium ON THE TOMATO SURFACE By Marianne Kathryn Fatica May 2013 Chair: Keith Schneider Major: Food Science and Human Nutrition The rise o f produce linked salmonellosis outbreaks has directed attention towards the environmental persistence of Salmonella spp. A complex phenotype of Salmonella known as rdar (red dry and rough), has been linked to increased resistance to environmental desic cation and stress Mutations in genes encoding thin aggregative fimbriae and/or cellulose biosynthesis result in a loss of the rdar phenotype. Thin aggregative fimbriae and cellulose are regulated by AgfD in Salmonella either directly or indirectly. T he t ranscription of agfD is greatest in conditions that would be encountered outside of the animal host, including 28 C, low osmolarity, and nutrient starvation during statio nary growth Both thin aggregative fimbriae and cellulose have also been implicated as important factors in the attachment and persistence on abiotic, animal cell, and plant surfaces, as well as important structural components of biofilm formation The ro le s of the AgfD regulator and AgfD regulated thin aggregative fimbriae and /or cellu lose in the interactions of Salmonella Typhimurium on the tomato surface were assessed through deletion mutants. The fitness of the mutants was compared to the wild type S. Typhimurium in the initial attachment to the tomato surface, persistence
14 on tomato es over 6 days, and biofilm formation on polystyrene, intact tomatoes, and tomato segments. The AgfD, thin aggregative fimbriae, and/ or cellulose deficiencies in Salmonella produced no biological effect in the initial attachment to tomatoes with in 5 min o f contact, but significantly impaired th e formation of biofilm on polystyrene, intact tomato surfaces, and tomato segments. Only the strains deficient in cellulose showed diminished persistence capabilities on the surface of both green and red tomatoes. In persistence the cellulose deficient mutants were more sustainable on the surface of re d, ripened tomatoes than green. These variations between tomato maturities indicate that both bacterial beha vior and produce ripeness may be factors in Salmonella su rvival on the tomato surface. The importance of cellulose production in the persistence and biofilm formation on the tomato surface may also be useful in providing a p o ssible target for preharvest treatments to reduce Salmonella contamination on produce.
15 CHAPTER 1 INTRODUCTION The estimated accounts of food borne disease in the US include 38.4 million illnesses, 71,878 hospitalizations, and 1,686 deaths annually N on typhoidal Salmonella is responsible for an estimated 1 million cases of illness es 20,00 0 hospitalizations, and over 350 deaths annually ( 68 69 ) Previous to 1990, most cases of salmonellosis were attributed to contaminated poultry and poultry products R ecently there have been multiple Salmonella outbreaks associated with a variety of produce, including tomatoes, seed sprouts, cantaloupes, jalapeo and Serrano peppers, and unpasteurized fruit juic es ( 12 13 15 17 33 38 53 74 ) T he most recent data estimat es that 46% of US foodborne ou tbreaks were associated with contaminated produce ( 57 ) The multi stat e tomato associated salmonellosis outbreak of S. Javiana in 1990 led to the reco gnition that raw tomatoes were possible vehicles for foodborne illness ( 15 38 ) Numerous Salmonella based outbreaks linked to tomatoes have been reported since ( 12 13 15 17 33 38 ) Tomatoes are suscep tible to contamination with Salmonella spp. during various production stages including cultivation, harvesting, packing, distribution, or preparation ( 19 ) Ther e have even been cases of multiple outbreaks being traced back to the same contamination source. The S. Newport tomato outbreaks of 2002, 2005, and 2006 were grown in the eastern shores of Virginia, where an irrigation pond was suspected as the common con tamination source ( 15 33 ) A packing facility was implicated for the source of contamination in both the 1990 S. Javiana and 1993 S. Montevideo outbreaks ( 15 33 38 )
16 The co ntamina tion of produce in the preharvest environment is challe nging to eliminate since the commodity is often co nsumed as a raw, fresh food The best method of minimizing the risk of foodborne illness is the prevention of contamination since there is no e ffective kill step for bacterial contaminants on raw, whole produce Prevention methods are detailed within the Good Agricultu ral Practices (GAPs), which were established in 1998 with the US Food and Drug Administration (FDA) Guide to Minimize Microbial F ood Safety Hazards for Fresh Fruits and Vegetables In Florida, Tomato GAPs (T GAPs) and Tomato Best Management Practices (T BMPs) have been mandatory since 2008 ( 27 ) GAPs are not yet implemented throughout the fresh produce industry, but the FDA Food Safety Modernization Act (FSMA) requires the implementation of prevention based food safety program s of which G APs are usually a prerequisite ( 59 ) Establishing protective programs like (T GAPS) is helpful in reducing the risk of pathogen introduction into produce fields ( 27 ) bu t there are too many gaps in know ledge to establish preharvest treatments targeted against enteric pathogens. There are numerous routes for Salm onella contamination of produce in the field environment including contaminated irrigation water, soil, manure, or through animal reservoirs including birds, insects, reptiles, and amphibians ( 42 67 79 ) T he field contamination of produce infers the ability of Salmonella spp. to survive on plants In order for th e pathoge n to be transmitted through produce the bacterium must first attach and persist on the plant surface. T he genetic mechani sms employed by Salmonella in the survival on pro duce are not well known, but research has indicated the attachment of Salmo nella on the tomato surface is rapid ( 44 ) Extracellular components of Salmonella including thin aggregative fimbriae and cellulose, have been
17 implicated a s i mportant in the attachment and persistence on abiotic materials animal cell s and plant surfaces including sprout seedlings and parsley leaves ( 3 5 45 49 ) Thin aggregative fimbriae and cellulose are also important structural components of biofilm formation ( 80 ) It has been hypothesized that the environmental persistence of pathogens is achieved through biofilm formation and that 30 80% of bacterial populations on plant surface are present in biofilms ( 51 80 ) Thin aggregative fimbriae and cellulose are proposed to be regulated by AgfD in Salmonella with direct regulation of thin aggregative fimbriae and indirect regulation of cellulose through adrA Transcription of agfD is greatest in conditions that would be encountered outside of the animal host, including 28 C, low osmolarity, and nutrient starvation during stationary growth ( 63 ) Furthermore, the components regulated by AgfD contribute to the production of an extracellular matrix char acterized by a rough, aggregative morphology in Salmonella grown in 25 30 C. The patterned morphology referred to as rdar (red, dry, and rough), is proposed to be more resistant to desiccation and environmental stress, which are important in environmenta l survival of Salmonella ( 100 ) An understanding of the mechanisms Salmonella employ in the initial attachment, persistence, and biofilm formation on p lant surfaces may allow for the development of intervention procedures and strategies to reduce the risk of preharvest tomato contamination.
18 CHAPTER 2 LITERATURE REVIEW Produce Associated Foodborne Illness The Centers for Disease Control and Prevention (CDC) estimate that 95% of Salmonella based infections originate from foodbor ne sources ( 29 ) Several Salmonella enterica serovars have been associated with both animal and produce based outbreaks, but two serovars, S. Enteri tidis and S. Typhimurium, are associated with over half of the reported salmonellosis cases in the US ( 29 ) The primary r outes of Salmonella infection in humans are through fecal oral transmission or the ingestion of contaminated products. Salmonellosis typically manifests as vomiting, diarrhea, cramps, and fever; and while the illness is typically self limiting, there are high risk populations of young children, the elderly, and the immunocompromised where the illness may be fatal ( 60 ) The infectious dose is determined by physiological characteristics of the Salmonella strain and the health status of the host. The dose may also be affected by the composition of the contaminated food product ( 92 ) The typical infectious dose may range from 10 6 to 10 8 CFU, whereas the min imum infectious dose for high risk members of the population has been estimated to be less than 100 CFU ( 21 ) The infection of humans by non typhoidal Salmonella spp. is based on the ability of the bacterium to survive the adverse conditions of the stomach and reach the site of colonization. Salmonella spp. survives a broad range of pH conditions from 4.05 to 9.5, with an optimum of pH 6.5 7.5. The optimum growth temperature is 37C, but Salmonella spp. are able to proliferate between 7 48C ( 21 ) The wide adaptability of Salmonella spp. allows the bacterium to persist outside of the animal host. Due to the
19 adaptability of Salmonella spp., salmonellosis is a leading foodbo rne illness in the US ( 18 69 ) It is estimated that foodb orne diseases cause 38.4 million illnesses, 71,878 hospitalizations, and 1,686 deaths annually in the US ( 68 69 ) The current trends indicate an increase in produce based outbreaks. foodborne outbreaks were determined to b e were linked to produce ( 75 ) T he most recent data estimated t hat 46 % of US foodborne ou tbreaks were attributed to contaminated produce ( 57 ) The number of produce based outbreaks associated with Salmonella spp. has also increased. Previous to 1990, most cases of salmonellosis were attributed to contaminated poultry and poultry products ( 83 ) In 2002 2003, 31 Salmonella spp. outbreaks of produce origin were reported in comparison to the 29 reported poultry related outbreaks ( 16 ) Salmonella spp. have been associated with a variety of produce, with large salmonellosis outbreaks being linked to tomato e s, seed sprouts, cantaloupes, jalapeo and Serrano peppers, and unpasteurized fruit juices ( 12 13 15 17 33 38 53 74 ) In the US, there have been numerous Salmonella spp. based outbreaks li nked to particular produce commodities including tomatoes ( 12 13 15 17 33 38 ) A significant, multi state tomato associated salmonellosis outbreak occurred in 1990. The outbreak investigation led to the recognition that raw tomatoes were a possible vehicle for foodborne illness ( 15 38 ) In the 1990 outbreak, consumption of contaminated whole round tomatoes resulted in 176 cases of S Javiana infection The outbreak was traced to a repacking facility ( 38 ) In 1993, another multi state outbre ak occurred with S Montevideo and tomatoes The source of contaminati o n was traced back to the
20 same packing facility implicated in the 1990 outbreak ( 38 ) In 1999, another multi state outbreak of salmonellosis was traced back to t he consumption of raw, round tomatoes. S. Baildon was implicated as the etiological agent in the outbreak ( 22 ) S. Newport was linked to tomato a ssociated outbreaks in 2002, 2005 and 2006 The implicated tom atoes of the outbreaks were grown and packed in the eastern shores of Virginia. The S. Newport strain was isolated from an irrigation pond near the tom ato fields after the 2005 outbreak occurred The S. Newport strain of the 2006 outbreak was not directly isolated from the irrigation pond, but the tomatoes were grown in the same area and the strain had an identical pulsed field gel electrophoresis (PFGE ) pattern as the strain in the previous outbreaks ( 15 33 ) Contaminated irrigation water was also suspected to be the source of the 2008 S. Saintpaul outbreak associated with jalapeo and Serrano peppers ( 17 ) This outbreak is significant because there were not only over 1,400 illnesses in 42 states, but tomatoes were or iginally implicated as the suspected source of the outbreak. The early association of the outbreak with tomatoes resulted in a nationwide loss of 300 million dollars and a loss of 100 million dollars to Florida tomato growers alone ( 17 36 ) Salmonella in the Produce Environment In the past, it was perceived that foodborne diseases were primarily linked to animal based products rather than produce ( 75 83 ) The current trends indicate an increase in produce based outbreaks with an estimated 46% of US foodborne illnesses from 1998 2008 associated with produce ( 57 ) Produce is often consumed as a raw, fresh commodity with little microbial reduction, pr ocessing, or packaging, thus increasing consumer risk from contaminants. The potential contamination of produce is possible during cultivation, handling, processing, and preparation of the products as
21 illustrated in previous produce associated outbreaks ( 19 75 85 ) Once Salmonella has attached to the p roduce, the pathogen may persist throughout the shelf life of the product. Salmonella have been isolated from both green and red tomatoes as well as sprouts, mangoes, peppers, and cantaloupes among other commodities ( 14 17 44 89 ) The initial attachment of S Montevideo on the surface of tomatoes has been documented to occur within only 30 sec of contact ( 44 ) Therefore, preharvest or postharvest contact between t he produce and Salmonella spp. may result in contamination of the product. The correlation between human pathogens and produce is not currently understood. Salmonella spp. are characterized as bacteria with wide pH and temperature tolerances, which aid in survival within various food matrixes as well as in the animal host ( 21 92 ) The adaptability of the bacterium c ould be sufficient to allow persistence in the secondary niche of the produce environment ( 15 38 ) The pathogen may utilize plants as a vehicle for transfer between animal hosts. The initial introduction of the bacterium into the animal host is often through a food or water source, where the animal further sheds the bacterium back into the environment. Once the ba cterium has re entered the environment, Salmonella spp. may p ersist on the plants until another animal ingests the plant. Thus, enteric pathogens may exhibit mechanisms to survive on plants to use them as a transfer medium between the environment and thei r animal hosts. The phyllosphere is characterized as the aboveground surface of the plant, which is a potentially hostile environment for an enteric pathogen. The phyllosphere exposes bacteria to high doses of UV light, poor nutrients, an aerobic environm ent, and variable
22 temperature conditions ( 37 96 ) These conditions are in contrast to the intestinal environme nt tha t is shielded from UV light, is nutrient rich, anaerobic, and has little temperature variability. A single study inoculated six enteric bacteria and viruses onto cantaloupe, lettuce, and bell pepper crops under controlled tempe rature and humidity co nditions The pathogens were still detectable in the plant environments after 14 days, demonstrating the survival of enterics within the phyllosphere ( 81 ) This s tudy is important in demonstrating that enteric pathogens, including Salmonella spp., can survive in the phyllosphere despite the harsh and variable conditions. While laboratory studies are important in illustrating the survival of enterics, it remains u nclear how the bacteria adapt and survive in the plant environment. Furthermore, in vivo and laboratory studies cannot simulate the true produce environment. There are multiple factors involved in the interaction between Salmonella spp. and produce inclu ding the specific commodity, cultivar, physiological state of the plant and the extrinsic environmental conditions ( 3 43 73 95 ) There also appears to be Salmonella serovar depend factors facilitating plant colonization. Serovars of Salmonella typically associat ed with poultry, including S. Enteritidis and S. Dublin, appear less apt to grow and persist in tomatoes than S. Montevideo and S. Newport serovars ( 73 ) While current research methods cannot address all of the factors involved in enteric and plant interactions, the studies highlight the influences of each factor to better understand the overall interactions in the plant s ystem. The bacteria may persist in the plant environment through the formation of biofilm s on the plants S. Thompson was visualized on the surface and within cuts of lettuce leaves through episcopic differential interference contrast microscopy coupled
23 w ith epifluorescence to find that the aggregated cells appeared slimy, suggesting the formation of a biofilm on the lettuce ( 94 ) Protection and enhanced survival of human pathogens on th e plant surface may also result from the association of the enterics within the biofilms of the plant microflora ( 51 ) While biofilm formation on the surface of p roduce has been observed, the formation of Salmonella biofilms within plant tissue remains to be shown ( 80 ) The role of plant microflora in the survival of human pathogens on plants also remains unclear. A survey of supermarket produce revealed that Salmonella incidence was twic e as likely in the presence of soft rot caused by plant pathogen s implicatin g a role of plant pathogens in the incidence of Salmonella ( 95 ) Another study showed that the microflora of carrots, green peppers, lettuce, green and purple cabbage, celery, and g reen and yellow onions were found to be inhibitory to human pathogens, including S. Montevideo ( 72 ) In the co infection of damaged cilantro leaves, Salmonella populations were greater when co inoculated with Erwinia chrysanthemi than wi th Pseudomonas viridiflava ( 9 ) Another area of concern is that Salmonella spp. is reaching contamination levels on produce that are capable of causing illness. The typical infectious dose of Salmonella spp. r anges from 10 6 to 10 8 CFU ( 21 ) These are tremendous concentration s to be reached on the plant surface with limited water and nutrient availability T he infectious nature of the Salmonella spp. may be de rived from the adaptations needed to survive o n produce rather than reaching the infe ctious dose concentrations. One of the first barrier s in human defenses against enteric pathogens is considered to be the acidity of the stomach. Salmonella spp. have an acid response
24 system which may enable survival in the acidic conditions of the stomach ( 28 ) The increased survival of Salmonella spp. in acidic condition s was observed after inoculation into fresh produce ( 8 ) Salmonella serovars adapted in apple, orange, and tomato juices also display ed enhanced survival in simulated gastric fluid in comparison to the non adapted controls ( 103 ) The tolerance and adaptation to low pH environments within acidic produce or acidic wounds from plant pathogens may increase the fitness of Salmonella spp. in the low pH of the stomach and decrease the infectious dose. AgfD Regulated Attachment M echanism s of Salmonella In understanding the survival of Salmonella spp. on the surface of plants and produce commodities, the first aspect to be studied is the mechanism of attachment. The aerial surface of plants are protected from desiccation and bacterial inf iltration by the plant cuticle wax es, and polysaccharides ( 77 ) Bacterial attachment to surfaces is characterized as a two step process including a reversib le phase and irreversible attachment. Reversible binding is hypothesized to be directed by physiochemical properties of the bacterium and surface, involving weak attractions including Van der Waals forces and hydrogen bonding. Flagella, pili, and fimbria e are suspected to influence this first step of attachment. The stronger, irreversible attachment is hypothesized to involve both chemical and physical properties with extracellular components, such as cellulose, contributing to attachment ( 44 84 ) Current research implicates that Salmonella spp. may attach to the plant surface in a similar manner. Thin aggregative fimbriae have been established to facilitate adhesion to solid surfaces in Escherichia coli ( 91 ) The thin aggregative fimbriae are filamentous proteinaceous non flagella appendages expres sed on the external surface of the bacterial cell ( 20 ) which are also referred to as Tafi in Salmonella ( 98 ) There are up to
25 15 putative fimbrial operons identified in Salmonella that can facilitate adhesion, such as agf (thin aggregative fimbriae), along with fim (type 1 fimbriae), pef (plasmid encoded fimbriae), and lpf (long polar fimbriae) ( 40 86 ) T he production of cellulose in Salmonella is generally used for protection ( 76 ) Thin aggregative fimbriae and c ellulose also have structural role s in biofilm formation, with cellulose providing strength to biofilms ( 31 80 ) Salmonella biofilm formation on Aspergillus niger has also been shown to include cellulose chitin interactions, with insufficient attachment to the fungi by cellulose deficient mutants. Thin aggregative fimbriae and cellulose also contribute to the a ggregative multicellular behavior of the bacteria ( 4 5 63 100 ) Thin aggregative fimbriae, cellulose formation, and the O antigen capsule were found to influence colonization of alfalfa seedlings ( 4 5 ) In Salmonella, there is a regulator known as AgfD that directly regulates thin aggregative fimbriae and indirectly regulates cellulose biosynthesis through adrA gene activa tion ( Figure 2 1) ( 100 ) The AgfD regulated thin aggregative fimbriae and cellulose of Salmonella spp. may be significant in the attachment of the bacteria to plants ( 2 4 5 63 ) The AgfD transcriptional regulator is composed of a C terminus DNA binding domain homologous to those of the LuxR family of regulators and a N terminus putative receiver domain ( 35 64 ) AgfD is proposed to directly regu late thin aggregative fimbriae and indirectly regulate cellulose in Salmonella ( 100 ) The thin aggregative fimbriae and cellulose contribute to the production of an extracellular matrix, which produces the distinct rdar (red, dry, and rough) morphology of Salmonella The rda r morphology is expressed as patterned, aggregative colonies of Salmonella when grown on Congo Red media at 25 30 C ( 31 61 ) It has been hypothesized that Salmonella
26 exhibiting the rdar morphology are m ore resistant to desiccation and envir onmental stress, which is vital in environmental survival ( 100 ) It has also been pr oposed that non rdar morphology promotes fitness of S. Typhimurium within an infected tomato ( 104 ) The agfD promoter is regulated by d ifferent environmental factors through global regulators and the secondary messenger molecule cyclic diguanylate (c di GMP). The transcriptional regu l ators of agfD include OmpR, RpoS, MlrA, CxpR, H NS, a nd IHF ( Figure 2 1). The regulators are able to activate agfD expression in response to environmental conditions including temperature, oxygen tension, nutrient stress, osmolarity, and ethanol presence ( 11 47 80 100 ) AgfD is produced from the expression of agfD The AgfD regulator stimulates the expression of agfBAC, which encodes for the production of the thin ag gregative fimbriae ( 62 100 ) RpoS and CxpR can also directly regulate the production of thin aggregative fimbriae without activation by AgfD ( 100 ) AgfD also indirectly regulates cellulose produc tion through the expression of adrA. AdrA contains a GGDEF domain that results in the production of c di GMP. The transcription of the bsc ABZC bcsEFG operon, which drives cellulose biosynthesis is activated by c di GMP AgfD activates expression of adr A ; AdrA then activates the biosynthesis of cellulose through the production of c di GMP The AgfD dependent activation of bcsABZC through AdrA has been demonstrated in Salmonella at 28C in Luria Bertani media. There are AdrA independent pathways of c di GMP production in Salmonella as well as AgfD independent pathways of cellulose production in Salmonella at 37C ( 23 32 100 ) The c di GMP molec ule is also known to promote sessile behavior in Salmonella, increasing biofilm formation ( 100 )
27 Thin aggregative fimbriae promote autoaggregation of cells and have a role in the attachment of S. enterica to plant seedlings and animal cells. The fim briae are utilized by both S. enterica and E. coli during the attachment to alfalfa sprout seedlings and enhance the attachment of S. enterica to intestinal epithelial cells ( 4 46 82 ) Cellulose, along with thin aggregative fimbriae, contributes to both the aggregative multicellular behavior of Salmonella spp. and adherence to animal cells. The actions of cellulose and thin aggregative fimbriae also confer resistance against chlorination, acid, and possibly desiccation ( 32 70 100 ) The AgfD regulated thin aggregative fimbriae and cellulose may aid in the survival of Salmonella spp. in the harsh plant environment. T ranscription of agfD is maximum during stationary growth at 28 C and low osmolarity, which are conditions that would be encountered outside of the host ( 62 ) While oxygen and nutrient availabilities interplay to dictate the activity of th e transcription of agfD there is maximum expression in both microaerophilic conditions in rich media and in aerobic conditions of minimal media. Nutrient starvation and ethanol presence also increase the expression of agfD, with the depletion of nitrogen and phosphate acting as a signal for activation during stationary growth ( 30 ) The thin aggregative fimbriae act to regulate aggregative behavior and short range cell to cell interactions, while the adrA gene regulates long range cell to cell interactions through activation of cellulose biosynthesis ( 63 ) Previous studies have characterized the attachment of S. enterica to alfalfa seedlings and parsley leaves but the AgfD regulated gene system of Salmonella spp. has not been characterized on other produce ( 4 5 49 ) In vitro studies have shown
28 agfD mutant of S. Typhimurium ATCC 14028 completely lacked multicellular behavior and was unable to adhere to glass, form a pellicule in liquid media, or produce an extracellular matri agfD mutant, there was no significant influence on the virulence of the S. Typhimurium in a mouse model, indicating that the regulator is utilized outside of the animal host ( 63 ) On alfalfa seedlings rpoS, agfD, and bcsA were identified as required genes for the attachment of S. Newport. The results also showed that no individual gene was responsible for attachment indicating a cooperative mechanism ( 4 5 ) The studies also noted the downregulation of AgfD during the attachmen t and colonization of the alfalfa sprouts, which means that genetically based assays targeting upregulated genes during attachment may not correctly characterize the role of the AgfD regulator ( 4 5 ) The genetic characterization of the attachment of Salmonella spp. to tomatoes and the identification of critical extracellular components will lead to a better understandi ng of the mechanisms used for colonization of plants. The understanding of the microbial behavior of Salmonella spp. and its asso ciation with tomatoes may provide specific targets for preharvest treatments for reducing the initial attachment of Salmonella spp. to tomatoes in the field environment Research Aim and Objectives Understanding the initial interactions between Salmonella spp. and produce is crucial in establishing targeted preharvest measures to reduce produce contamination. Many preharvest t echniques currently used with tomatoes aim to reduce the contact of pathogens and produce, such as the use of uncontaminated water sources an d animal exclusion ( 27 ) Unfortunately, there are numerous sources of contamination and endless unforeseen circumstances that can introduce a patho gen into a produce field
29 including events that ar e out of human control such a s storm s and flooding Estab lishing protective programs such as the Florida Tomato Good Ag ricultural Practices (T GAPS) are helpful in reducing the risk of pathogen introduction i nto produce fields ( 27 ) but there are too many gaps in the know ledge base to establish targeted preharvest treatme nts The main aim of this research is to determine the role of the ext racellular components of Salmonella spp., specifically AgfD regulated thin aggregative fimbriae and cellulose, in the initial interactions between the bacterium and the tomato surface of intact, unprocessed tomatoes. The significance of this aim is that p roduce contamination events cannot occur in the field without the attachment and persistence of the bacterial pathogen. There are numerous factors that contribute to the attachment and persistence of pathogens on produce, as previously stated, and not all of these components can be incorporated into this research. The aspect of Salmonella spp. and produce intera ctions being studied within the scope of this project are the roles of the AgfD regulated thin aggregative fim briae and cellulose production i n th e initial attachment, persistence, and biofilm formation of S. Typhimurium on the tomato surface The focus was directed towards the attachment and persistence mechanism s of Salmonella spp. on the intact tomato surface because T GAPS currently mandates th e culling, sorting, and removal of injured fruit to minimize microbial contamination ( 27 ) Therefore, the rese arch focuses on the attachment and persistence of S. Typhimurium on the produce that reaches consumers. Furthermore, unprocessed tomatoes that received no sanitation or wax treatments were utiliz ed in the studies to maintain the
30 condition of the tomatoes in the field at time of harvest The current information about the active environmental conditions of the AgfD system and the data demonstrating the decreased attachment and biofilm formation of agfD, agfB, and bcs deletion mutants of Salmonella supports the hypothesis that AgfD regulated thin aggregative fimbriae and cellulose are critical to the initial attachment, persistence, and biofilm formation of Salmonella spp. on the tomato surface. The overall goal is to determine the extracellular components o f Salmonella spp. that contribute to attachment and persistence mechanisms. This knowledge may be used in the future to ascertain how Salmonella are capable of evading the current washing the sanitation processes for tomatoes and possibly establish prehar vest prevention measures that target the actions of these extracellul ar components to inhibit attachment and contamination of produce The main objectives of this research are as follows: 1. Determine the roles of AgfD regulated thin aggregative fimbriae an d cellulose in the initial attachment of Salmonella enterica serovar Typhimurium ATCC 14028 to the surface o f intact, unprocessed green and red round tomatoes. 2. Characterize the roles of AgfD regulated thin aggregative fimbriae and cellulose in the persiste nce of Salmonella enterica serovar Typhimurium ATCC 14028 on the surface of unprocessed green and red round tomatoes 3. Assess the roles of AgfD regulated thin aggregative fimbriae and cellulose in the biofilm formation of Salmonella enterica se rovar Typhimu rium ATCC 14028 on intact unprocessed green and red round tomatoes and tomato segments
31 Figure 2 1. The AgfD regulation system. The components regulating agfD transcription are indicated in the boxes below agfD and agf BAC. HN S and MrlA are also regu lators not included in the diagram. As AgfD is expressed, the expression of agf BAC and adrA are activated. AdrA generates cyclic digu anylate (c di GMP), which further activates cellulose biosynthesis through bcsABZC
32 CHAPTER 3 GENETIC MUTANT CONSTRU CTION AND CHARACTERIZATION Background The main aim of this research wa s to determine the role of the extracellular components of Salmonella spp., specifically AgfD regulated thin aggregative fimbriae and cellulose as well as the AgfD regulator in the ini tial attachment, persistence, and biofilm formation of Salmonella on the tomato surface. Salmonella mutants of S. Typhimurium ATCC 14028 deficient in AgfD production, thin aggregative fimbriae and /or cellulose production were designed to assess the role of these extracellular components in the interactions between Salmonella spp. and tomatoes. The advantages of using Salmonella Typhimurium ATCC 14028 as the parent Salmonella strain rather than an outbreak strain include the access to the full genome sequ ence, susceptibility to phage P22, and lac k of antibiotic resistances ( 1 84 ) S. Typhimurium 14 028 has also been used to research Salmonella with lettuce ( 71 ) alfalfa ( 25 ) and tomatoes ( 56 73 ) S. Typhimurium 14028 was compared with tomato associated outbreak strains on the surf ace and within tomatoes to find the strains displayed similar behaviors ( 56 ) This section de scribes the methodologies us ed in the construction and characterization of the del etion mutants as well as th e genetic complements and vector control strains. The AgfD regulated thin aggregative fimbriae and cel lulose of Salmonella are components that co ntribute to the production of the extracellular mat rix, characterized as rdar (red dry, and rough) colony morphology The rdar morphology is expressed as the patterned, aggregative colonies of Salmonella when grown on Congo r ed (CR) media at 25 30 C ( 31 61 ) It has been hypothesized that the rdar morphology of
33 Salmonella may be part of an environmental persiste nce strategy, since the morphology has not been shown to increase virulence capabilities ( 99 ) The disruption of thin aggregative fimbriae and/or cellulose production in Salmonella spp. can be observed through alternative aggregative patterns of the bacterial colonies on CR media. The strand structure of the thin aggreg ative fimbriae subunits and the glucose linkages of cellulose bind to the hydrophobic dye of the CR media, contributing to the rdar appearance. The lack of intact thin aggregative fimbriae result s in pdar (pink, dry, and rough) colony morphology ( 62 ) The lack of cellulose biosynthesis results in bdar (bro wn, dry, and rough) colony mor phology, and the lack of both thin aggregative fimbriae and cellulose results in saw (smooth and white) morphology ( 61 62 ) The Salmonella mutan ts utilized in these studies were phenotypic ally characterized by their morphology on CR media, as well as on Calcofluor white (CW) agar for cel lulose visualization Construction of Salmonella Deletion Mutants DNA Amplification and Vi sualization In t he construction of the Salmonella enterica serovar Typhimurium ATCC 14028 deletion mutants polymerase chain reaction (PCR) was performed for confirmation and amplification of spe cific genes using DNA of individual colonies or plasmid DNA The recipe routinely used for a single PCR reaction of 25 l is as follows, 20.5 l sterile DNA grade water (Fisher Scientific, Pittsburgh, PA) 2.5 l of 10X Standard Taq buffer with MgCl 2 (New England Biolabs, Ipswich, MA) 1 l of 2.5 mM deoxynucleotid e triphosphates ( dNTP s), 0.5 l of 50 M l of 50 M reverse primer, (New England Biolabs, Ipswich, MA) To perform colony PCR, an individual colony was lightly touched on an agar plate with
34 appropriate antibiotics with a sterile pipe tte tip and was added to the PCR reaction. PCR of plasmid DNA was performed by adding 1 l of DNA to the react ion mix. The DNA was amplified in either the TC412 (Techne, Minneapolis, MN) or MJ mini (Bio Rad, Hercules CA) thermocycler on a cycle running f or 10 min at 95 C, 35 cycles of 1 min at 95 C, 1 min at 53 C, 2 min at 72 C, and a final extension at 72 C for 10 min unless otherwise specified. The primers utilized in these studies are listed in Table 3 1. The DNA amplified from the PCR reaction was visualized through agarose gel electrophoresis. Tris Acetate Ethylene d iamine tetraacetic a cid ( EDTA ) buffer (1X) referred to as TAE buffer, was used to make 0.9% (w/v) agarose gels containing 1 3% ethidium bromide. The gels were poured into 7 cm x 10 cm trays in a mini sub cell GT electrophoresis system (Bio Rad, Hercules CA). The gel was loaded with 4 l of exACTGene 1kb Plus DNA Ladder (Fisher Scientific, Pittsburgh, PA) and 8 15 l of sample The FB300 power supply (Fisher Scientific, Pittsburgh, PA) was used to complete the electrophoresis. The ge ls were visualized with the Molecular Imager Gel Doc XR+ System running version 3.0 of the Image Lab software (Bio Rad, Hercules CA). The desired Salmonella deletion mutants included the following targe ted genes, agfD the gen e producing the AgfD regulator, agfB the gene encoding the fimbrin protein, AgfB, which is the proposed anchor site of the fimbrial assembly, and bcs the promoter region and cellulose synthase gene, bcsA ( 63 97 106 ) The three genes were targeted and amplified in S. Typhimurium 14028 through the PCR methods described in the previous section to ensure major mutations were not present. Due to the large size of the bcs segment (3.7kb) the following thermocycler conditions were
35 used, 10 min at 95 C, 35 cycles of 2 min at 95 C, 2 min at 5 3 C, 5 min at 72 C, and a final extension at 72 C for 10 min The primers used to target each gene are listed in T able 3 1. The PCR products resulted in the predicted gene size s indicating no significant mutations or defects in the genes of the parent s train PC R was also repeated in the Salmonella strains containing the desired mutations. Phage Mediated Transduction Transduction was used to transfer mutations between Salmonella strains utilizing bacteriophage, P22. After confirmation of the desired mu tations, cell lysates were made from the mutant donor strains. The lysates were made by growing the individual donor strains overnight at 37 C, shaking at 200 revolutions per minute ( rpm ) in 5 ml of L uria Bertani (LB) broth (Fisher Scientific, Atlanta, GA ) with the appropriate antibiotics (media com position listed in Appendix A ) The overnight cultu res were washed three times in p hosphate buffered s aline (PBS) (Fisher Scientific, Pittsburgh, PA) For each donor strain, a series of eight cultures tubes we r e set up, each containing 5 ml of LB without antibiotics. The overnight culture w as used to inoculate each of the 8 tube s with 100 l of culture per tube. One hundred l of the P22 phage was added to the first tube and was serially d iluted (1/10) to th e following six tubes by trans ferring 500 ml to each The remaining tube was the bacterial control. The culture tubes were incubated for 2 h r at 37C, shaking at 200 rpm. After incubation, the eight culture tub es were observed and the culture tube exhib iting the greatest amount of clearing was centrifuged with the Eppendorf Centrifuge 5415 D at 12,000 x g for 15 min. The supernatant was removed, 100 l of chloroform was added to the supernatant, and the solution was centrifuged at 12,000 x g for 15 min. The final supernatant was pla ced in a sterile glass tube with an additional 100 l of chloroform. E ach donor lysate was stored at 4C. The Salmonella
36 cultures utilized as donor strains for lysate construction include S. Typhimurium LT2 agfD11::FRT K an FRT mutant from Santiviago et al. ( 66 ) an d S. Typhimurium JSG1748 ( 58 ) agfD11::FRT Kan FRT and P22/ bcs::FRT Kan FRT respectively The full descriptions o f the strains are listed in T able 3 2. The transductions were performed by growing the wild type recipient culture ( S. Typhimurium 14028) overnight in 5 ml of LB at 37 C, shaking at 200 r pm. The overnight culture was washed three t imes in PBS. A set of eight culture tubes were set up, each containing 100 l of LB without antibiotics. One hundred l of culture was added to seven of the eight tubes, with the uninoculat ed tube being set aside. The first tube was then mixed with 10 l of P22 Phage and was serially diluted (1/10) in the next five tubes by subsequently transferring 2 0 l to eac h tube The inoculated tube without the P22 phage was the bacterial control. The tube that was previously set aside was supplemented with 10 l of P22 phage and was the phage control. After incubation of the culture tubes at 37 C for 25 min, the tubes were supplemented with 1 ml of LB with 10 mM ethylene glycol tetraacetic a cid (EGTA). The cultures wer e incubated at 37 C for 1 h r and were then centrifuged at 12,000 x g for 1 min. The culture was resuspended in 100 l of the supernatant and each soluti on was plated onto LB agar supplemented with k anamycin (50 g/ml) and EGTA (10 mM) The EGTA was added to the transduction reactions to prevent further phage absorption. The EGTA strongly chelates calcium (Ca ++ ) that is necessary for the P22 phage attach ment ( 39 ) The plates were incubated at 37 C overnight and resulting colonies were purified on LB agar supplemented with kanamycin and EGTA three times before being tested on Evans Blue Uranine (EBU) agar (Appen dix A ) by being cross streaked
37 against P22 phage. The EBU agar test confirm s that the culture has been cu red of all phage while maintaining phage sensitivity The antibiotic utilized in the phag e med iated transduction studies was k anamycin because the deletion mutations of all of the donor strains c ontained k anamycin resistance cassettes. The Salmonella deletion mutants were constructed to result in strains deficient in AgfD regulation, thin aggr egative fimbriae and/or cellulose synthesis. The two Salmonella strains utilized to make the P22 lysates resulted in agfD 11 :: FRT K an FRT and P22/ bcs::FRT Kan FRT agfD11 :: FRT Kan FRT was transduced with the wild type, S. Typhimurium 14028 to result in the strain, M KF1 (Table 3 2) The second lysate, P22/ bcs :: FRT Kan FRT was constructed from a cellulose deficient strain JSG1748, provided from the previous work of Prouty and Gunn ( 58 ) The lysate was transduced with the wild type, S. Typhimurium 14028 to result in MKF2 (Table 3 2). The transduction was d one to ensure the deletion mutant was in a 14028 strain. The P22/ bcs::FRT Kan FRT was also transduced with S. Typhimurium TIM2265 a strain deficient in thin aggregative fimbriae production. The final construct, WJZ2, lacks both bcs and agfB genes, makin g it deficient in both cellulose and thin aggregative fimbriae synthesis (Table 3 2). Construction of Genetic Complements and Empty Vector Controls Cloning To support the results obtained from the use of MKF1 (14028 agfD::FRT K an FRT ), MKF2 (14028 bcs::FR T K an FRT) and WJZ2 (14028 bcs::FRT K an FRT ) on the tomato surface control strains of the deletion mutants including complementation vectors expressing the deleted genes were necessary The genetic complements we re used to restore the origina l phenotypes of the strains when the
38 deleted genes were expressed on extrachromosal plasmids. Restoration of the phenotype confirm s that the results observed from the deletion mutants were due to the disruption of the specific gene The primers listed in T able 3 1 were utilized t o amplify the three target areas of agfD, the bcsA gene and upstream promoter and agfB in S. Typhimurium 14028. The respective PCR products were isolated through gel electrophoresis, as previously described, and were purified u tilizing the Illustra TM DNA and gel band purification kit (GE Healthcare, Buckinghamshire, UK). The PCR products were each cloned into pCR2.1 TOPO (Invitrogen, Carlsbad, CA) to make pMKF1, pMKF2, and pMKF3, respectively. The pCR2.1 plasmids containing t he gene inserts were individually transformed into chemically competent E. coli (Table s 3 2 and 3 3 ). The E. coli was prepared as described in Inoue et al. ( 41 ) The transformation was performed by adding the ligation reaction to a fro zen aliquot of E. coli The mixture was incubated on ice for 30 min, when the cells were heat shocked for 30 sec at 42 C and then recovered on ice for 2 min. The cells underwent incubation in 1 ml of NZY+ for 1 hr at 37 C (NZY broth from Fisher Scientific, Pittsburgh, PA) (medi a composition in Appendix A ) The transformed cells were centrifuged and plated onto LB agar supplemented with kanamycin (50 g/ml) and 5 bromo 4 chloro 3 indolyl D galactoside (X gal) (40 g/ml) for b lue white screening analysis. Primers specific to e ach insert and M13 primers were used for PCR confirmation (Table 3 1) Construction of the complementation vectors proceeded with the subcloning of the pMKF1 (pCR2.1 agfD ), pMKF2 (pCR2.1 bcs ), and pMKF3 (pCR2.1 agfB ) into their respective recipient plas mids. The vector and plasmid containing the gene insert of interest were purified using the QIAprep spin mini prep kit (Qiagen, Santa Clarita, CA)
39 and were prepared for subcloning by restriction digest with the appropriate restriction enzymes (New England BioLabs, Ipswich, MA). The vectors were digested in parallel overnight at 37 C in NEB 10x Buffer (New England BioLabs, Ipswich, MA). In the case w h ere only one restriction enzyme was used, the vector sample was treated after incubation with 1 l of calf intestinal alkaline phosphatase (CIAP) in NEB Buffer 3 for 1 hr at 37 C to prevent self After separation through gel electrophoresis, the desired fragments of the linear plasmid and the DNA insert were excised from the gel and purified using the Illustra TM DNA and gel band purification kit. The restriction fragments were quantified using a Nanodrop 1000 spectrophotometer with software version 3.6.0 (Thermo Scientific, Wilmington, DE) Ligation reactio ns contained 4 l of the recipient vector, 4 l of the DNA insert, 1 l of the 10x T4 ligase buffer, and 0.5 l of the T4 ligase (New England BioLabs, Ipswich, MA). Two control reactions were also performed, one reaction lacking the insert to assess for s elf ligation and another reaction lacking the T4 ligase to assess for undigested vector. After overnight incubation, the ligations were transformed into chemically competent E. coli Subcloning was confirmed through blue white screening on LB agar s upplemented with X gal (40 g/ml) and the appropriate antibiotic and PCR analysis. The agfD complementation vector was constructed using pMKF1 and the plasmid, pWSK29. Since AgfD is a regulator of Salmonella, it was desirable to use a lo w copy number pl asmid for complementation EcoRI and SalI were used for the restriction digest of pMKF1 and pWSK29 in NEB 10x Buffer 3 to result in pMKF4 Construction of pMKF4 was confirmed through blue white screening on LB agar
40 supplemented with X gal (40 g/ml) and ampicillin and PCR analysis with MKF1, MKF2, and M13 primers (Table 3 1). Several attempts were made to subclone the bcs insert from pMKF2 into pWSK29 using EcoRI, but s ubcloning remained unsuccessful Therefore, the original pCR2.1 vector containing the bcs insert, pMKF2, was used as the complementation vector. The agfB complementation vector was constructed using MKF3 and the plasmid, pBBR1 MSC5 Gent. The pMKF3 and pBBR1 MSC5 vectors were digested with ClaI in NEB 10x Buffer 4 and were treated with CI AP to prevent self ligation to result in pMKF5. Confirmation of pMKF5 was achieved through blue white screening on LB agar supplemented with X gal (40 g/ml) and gentami cin ( 50 g/ml ) and PCR analysis with MKF11, MKF13, and M13 primers (Table 3 1). The fi rst step of confirming pMKF4 (pWSK29 carrying agfD ), pMKF2 (pCR2.1 carrying bcs ), and pMKF5 (pBBR1 M S c 5 carrying agfB ) was PCR screening white colonies from the LB agar with X gal with M13 primers (Table 3 1) The primers bind to Plac up and downstream f rom the insertion site s of pWSK29, pCR2.1, and pBBR1 M S C 5. A successful g ene insertion was indicated by PCR product s of approximately 1560 bp for pMKF4, 3900 bp for pMKF2, and 1100 bp for pMKF5. Without a gene insert, the M13 primers result in a product of 200 bp, so the size of the desired PCR product was determined by adding the 200 bp to the size of each gene insert. Once a positive PCR sample was obtained, PCR was repeated with primers M13F and MKF2 for pMKF4, M13F and MKF13 for pMKF2, and M13 and MK F6 for pMKF5. M13F is a forward primer of Plac and MKF2, MKF13, and MKF6 are the reverse primers specific to each gene. The resulting PCR products from these primer pairs indicated that the gene was inserted in the same orientation as Plac. An internal reverse primer was utilized for
41 pMKF5 to allow use of the standard thermocycler conditions and result in a smaller PCR product of 1400 bp. After PCR confirmation, a diagnostic digest was performed with each of the plasmids with EcoRI and SalI for pMKF4, E coRI for pMKF2, and ClaI for pMKF5 as previously described. Gel electrophoresis was used to visualize the size of the gene insert. The original gene inserts were recovered with products of ~1360 bp for agfD from pMKF4, 3700 bp for bcs from pMKF2, and 910 bp for agfB from pMKF5. Electroporations Salmonella cannot be made chemically competent. Therefore, transformations by electroporation were used to shuttle plasmids from E. coli into Salmonella and between Salmonella strains. Electrocompetent transform ations were utilized to move plasmids between the bacterial strains in the construction of the Salmonella complement and empty vector con trol strains. Electrocompet ent recipient strains were prepared from overnight cultures grown in LB with the appropriat e antibiotics. The overnight cultures were incubated on ice for 15 min, followed by a washing step to remove all nutrient and salts. The cultures were each washed 4 times with ice cold DNA grade water, w ith a final resuspension in 200 l of cold DNA wate r. The washed cells were then incubated on ice for an additional 10 minutes. With the cells remaining on ice, the electroporations were performed with 50 l of the culture and 3 l (15 ng/l) of the appropriate plasmid in chilled 2 mm gap electroporation cuvettes (Eppendorf, lectro porator (Bio Rad, Hercules, CA). The cells were recovered in 1 ml of NZY+ and incubated at 37 C for 1 hr. After incubation, the cells were centrifuged at 12,000 x g for 1 min and resuspended in 100 l of supernatant. The cells were then plated onto LB with appropriate antibiotics.
42 The Salmonella mutant complementation and empty vector controls were constructed b y first purifying the necessary plasmids, pMKF4 (pWSK29 agfD ), pMKF2 (pCR2.1 bcs ), and pMKF5 (pBBR1 MS C 5 agfB ), from E. coli spin mini prep kit. To transfer the plasmids from an E. coli strain to Salmonella, the plasmids were individually were passed through an intermediate host, the restrict ion minus modification plus Salmonella strain J S 198 which was derived from S. Typhimurium 14028 prior to transformation into the final S. Typhimurium host strains ( 26 ) From JS198 the plasmids were again purified utilizing the QIAprep spin mini prep kit and were transformed into their final host strains. The pMKF4 plasmid was transformed with MKF1 to construct the agfD complement strain, MKF1 pMKF4. The pMKF2 plasmid was transfo rmed with both MKF2 and WJZ2. The final bcs complement strain was MKF2 pMKF2 (Tables 3 2 and 3 3) The double mutant complement was completed by transforming pMKF5 into WJZ2 pMKF2 to complement both of the deleted genes of WJZ2 pBBR1 MS C 5 was used for th e vector of pMKF5 because of the encoded gentamicin resistance so both complementation plasmids of WJZ2 could be selected for with the use of different antibiotics (Tables 3 2 and 3 3). As a negative control, the vectors without gene inserts were also tra nsformed into the appropriate strains to result in MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR.21 pBBR1 (Table 3 2). Phenotypic Characterization of Salmonella S trains Phenotypic assays were performed on S. Typhimurium 14028 and the mutants to assess the morphol ogy on Congo red ( CR ) media and the cellulose production of the strains. The extracellular components, including thin aggregative fimbriae and cellulose, bind the hydrophobic dyes of the CR media, contributing to the resulting color
43 of the colonies. The colonies may also form complex surface patterns that can be completely removed from the agar plates. The presence of the extracellular components of Salmonella and the rough, aggregative pattern results in the rdar (red, dry, and rough) morphology on CR a gar. The lack of intact thin aggregative fimbriae and/or cellulose result in alternative morphologies including pdar (pink, dry, and rough) from a lack of intact thin aggregative fimbriae, bdar (brown, dry, and rough) from deficient cellulose production, and s aw (smooth and white) from a lack of both extracellular components ( 61 62 ) CR r ed dye ( MCB, Norwood, OH) and 20 g/ml of Brilliant Blue R 250 ( Acros Organics, Geel, Belgium ) in salt less LB agar (as described in Appendix A ) ( 23 62 ) Cultures were started from frozen glycerol stoc ks and used to inoculate 5 ml of LB broth with appropriate antibiotics Inoculated cultures tubes were incubated overni ght at 30C, sh aking at 200 rpm and then 6 l of each culture were spotted onto CR agar The plates were incubated at 30C for 48 hr. Photographs of cultures on CR media are displayed in F igure 3 1 The S. Typhimurium 14028 has the rdar morphology, with the modified 14028 strains, MKF5 and JS246 having similar rdar phenotypes. The MKF1 strai n expressed a s pdar morphology, MKF2 appeared as bdar, and the double mutant, WJZ2, appeared as a light pink to saw phenotype. The phenotypes of the mutant colonie s on CR media confirm the successful deletion of the targeted genes. The phenotype of MKF1 pMKF4 showed exaggerated rdar morphology lik ely due to multiple expression of the AgfD regulator on the pMKF1 plasmids The MKF2 pMKF2 and WJZ2 pMKF2 pMKF5 comple ment strains indicate only partial restoration of the rdar phenotype with an increase in red color, but
44 little patterned aggregative appearance. The three strains carrying the empty vector s controls do not exhibit phenotype restoration indicating that t he phenotype restoration of the complements was a result of the expression of the vector gene insert. Cellulose production of the Salmonella strains was assessed through both qualitative and quantitative assays. Salt less LB agar supplemented with C alcofl uor white with E vans Blue dye (Sigma Aldrich, St. Louis, MO) to a final concentration of 40 resulted in Calcofl uor w hite (CW) media (Appendix A ). The c alcoflu o r detects the 1,4 glucose link ages of the cellulose and fluorescence under a 366 nm ultr aviolet ( UV ) light source ( 106 ) Glycerol stocks of Salmonella strains were used to inoculate 5 ml of LB broth with appropriate antibiotics Cultures were incubated overnight at 30 C, shaking at 200 rpm. The cultures were streaked onto the CW a gar plates and incubated without light exposure at room temperature for two days. Afte r 48 hr, the plates were placed on the UV trans illumina tor and were photographed ( Figure 3 2) The plates were then placed b ack into the dark for 30 min After the da rk treatment, the cultures were suspended in sterile deionized (DI) water and samples were loaded into a 96 well polystyrene plate along with a control of DI water from an uninoculated CW agar plate. The optical density and fluorescence readings of the cu ltures were measured and recorded as p erformed in Da Re and Ghigo ( 23 ) Final cellulose production was quantified using the following equation, ((fluorescence of sample) (fluorescence of control) / (OD 600 of sample OD 600 of control) ) Photographs of the CW plates are displayed in Figure 3 2, with cellulose quantification in Figure 3 3. The greatest fluorescence was observed from the 14028, MKF5, JS246, and MKF1 pMKF4 strains ( Figure s 3 2 and 3 3 ) It was expected to see high cellulose
45 production from the 14028 strain and the modified 14 028 strains, MKF5 and JS246. A reduction in fluorescence was recorded for the MKF1, MKF2, and WJZ2 strains. A greater amount of fluorescence wa s observed in the S. Typhimurium MKF1 than in MKF2 and WJZ2 because cellulose production through adrA can be ex pressed in the absence of AgfD Cellulose production was directly affected in the MKF2 and WJZ2 strains with the deletion of a portion of the bcs operon. Full to partial restoration of c ellulose production restoration was observed in the S. Typhimurium M KF1 pMKF4, MKF2 pMKF2, and WJZ2 pMKF2 pMKF5 strains by an increase in recorded fluorescence. The three strains carrying the empty vector s ( controls ) d id not exhibit phenotype restoration, indicating that the phenotype restoration of the complements was no t the result of the transformation of the plasmid(s) alone. Growth Curves of Salmonella S trains The growth rates and patterns of S. Typhimurium 14028 and the strains derived from it were assessed before the use of the strains in studying the roles of AgfD, thin aggregative fimbriae, and/or cellulose production in the attachment, persistence, and biofilm formation of Salmonella on the tomato surface. Growth curves were used to establish if the genetic mutations affect ed the growth rate of the cultures. The use of cultures with slowed or altered growth patterns could result in inaccurate data on the fitness of th e cultures on the tomato surface Glycerol stocks of the Salmonella strain s were used to inoculate 5 ml of LB broth with appropriate antibiotics a nd incubated overnight at 30C, at 200 rpm Overnight cultures were diluted in sterile PBS to a concentration of 10 6 CFU/ml. A volume of 1 ml of the diluted culture was added to 99 ml of sterile LB, resulting in a starting culture of approximately 10 4 CF U/ml. Cultures were incubated at 30C for 12 h r while shaking at
46 200 rpm. Cultures were serially diluted each hour, including time 0 (10 minutes after inoculation) to the dilutions of the countable range (25 250 CFU/ml) and plated onto LB agar. The or iginal cultures were also plated. Plates were incubated overnight at 37 C and CFU were counted. The average l og 10 CFU/ml population was calculated for each time point. A nalysis of variance (ANOVA) was performed using JMP 9.0 Pro statistical software (SA S Institute Inc., Cary, NC) to determine sign ificant differences among resulting concentration s at each hour. With a significance of p<0.05, no significant differences were found between the S. Typhimurium 14028 and isogenic mutant strains depicte d in F ig ure 3 4
47 Table 3 1. PCR primers Name Primer Sequence Primer Target MKF1 MKF2 MKF5 MKF6 MKF11 MKF12 MKF13 M13F M13R GTCGAC CAGTAACTCTGCTGCTACAATCCAG GTTACGATGAAGAGTATGTCCGTG CATTAACCTGGACAGCACAAAGAC AGCACTGCCAGAAACTACGATT ATCGAT ATACCTCACAGCAGCTTGA TCTACTGCC ATCGAT CAGCATGTTATAGGTCAGATCCAG AGGTCTTCGTTATAGGTCGGTACA GTAAAACGACGGCCAG CAGGAAACAGCTATGAC F orward prime r of agfD in S. Typhimurium, SalI site R everse primer of agfD in S. Typhimurium F orward primer of agfB in S. Typhimurium R ever se primer of agfB in S. Typhimurium Fo rward primer of bcs in S. Typhimurium ClaI site R everse primer of bcs in S. Typhimurium ClaI site In ternal primer of bcs in S. Typhimurium Forward primer of lacZY (Invitrogen) Reverse primer of lacZY (Invitrogen)
48 Table 3 2. Bacterial strains Isolate or strain Relevant characteristics S ource, construction, or reference Salmonella Typhimurium Isolates 14028 Wild type American Type Culture Collection MKF5 14028 Kan R S pontaneous Kan R of S. Typhimurium 14028 JS246 14028 yjeP :: res1 tetR res1 Tet R ( 55 ) JS198 LT2 metE551 metA22 ilv452 trpB2 hisC527(am) galE496 xyl 404 rpsL120 flaA66 hsdL6 hsdSA29 zjg8103::pir+recA1 ( 26 ) MKF1 agfD11::FRT Kan FRT Kan R This study, agfD11::frt kan frt mut ation from Santiviago et al. ( 66 ) transduced into S. Typhimurium 14028 M KF1 pMKF4 agfD11::FRT Kan FRT pWSK29 carrying agfD Kan R Amp R This study MKF1 pWSK29 agfD11::FRT Kan FRT pWSK29 Kan R Amp R This study JSG1748 bcs::FRT Kan FRT Kan R ( 58 ) MKF2 14028 bcs::FRT Kan FRT Kan R This study, bcs::FRT Kan FRT from JS1748 transduced into S. Typhimurium 14028 MKF2 pMKF2 14028 bcs::FRT Kan FRT pCR2.1 carrying bcs Kan R Amp R This study MKF2 pCR2.1 14028 bcs::FRT Kan FRT pCR2.1, Kan R Amp R This study WJZ2 14028 bcs::FRT Kan :FRT Kan R Gent R This study WJZ2 pMKF2 pMKF 5 14028 bcs::FRT Kan pCR2.1 bcs pBBR1 agfB Kan R Gent R This study WJZ2 pCR2.1 pBBR1 14028 bcs::FRT Kan pCR2.1 pBBR1 MS C 5, Kan R Gent R This study
49 Table 3 2. Continued Isolate or strain Relevant characteristics* Source, construction, or reference JTN176 Spontaneous non rdar mutant of S. Typhimurium 14028 ( 104 ) JTN212 Spontaneous non rdar mutant of S. Typhimurium 14028 ( 104 ) Escherichia coli isolates F M15 (lacZYA argF) U169 recA1 endA1 hsdR17 (rk thi 1 gyrA96 relA1 Invitrogen ( 55 ) *Tet R designates tetracycline resi stance, Kan R kanamycin resistance; Amp R ampicillin resistance; Gent R gentamicin resistance
50 Table 3 3. List of plasmids Plasmid name Relevant characteristics* Source or reference pCR2.1 TOPO Cloning vector, Kan R Amp R Invitrogen pWSK29 Low copy numbe r cloning vector, Amp R ( 93 ) pBBR1 MS C 5 Cloning vector, G ent R ( 48 ) pMKF1 pCR2. 1 containing agfD amplified by MKF1 and MKF2 from 14028 with expression from the native promoter, Kan R Amp R This study pMKF2 pCR2.1 containing bcs amplified by MKF11 and MKF12 from 14028 with expression from the native promoter region and cellulose synt hase gene, bcsA, that are in the same orientation as Plac on pCR2.1 Kan R Amp R This study pMKF3 pCR2.1 containing agfB amplified by MKF5 and MKF6 from 14028 with expression from the native promoter Kan R Amp R This study pMKF4 pWSK29 containing agfD sub cloned from pMKF1 from 14028 with expression from the native promoter and in the same orientation as Plac on pWSK29 Amp R This study pMKF5 pBBR1 MS C 5 containing agfB subcloned from pMKF3 from 14028 with expression from the native promoter and in the same orientation as Plac on pBBR1 M S C 5, Gent R This study *Kan R designates kanamycin resistance; Amp R ampicillin resistance; Gent R gentamicin resistance
51 Figure 3 1. Depictions of Salmonella Typhimurium strains on Congo r ed media. A) S. Typhimurium 14 028, B) MKF5, C) JS246, D) MKF1, E) MKF2, F) WJZ2, G) MKF1 pMKF4, H) MKF1 pWSK29, I) MKF2 pMKF2, J) MKF2 pCR2.1, K) WJZ2 pMKF2 pMKF5, and L) WJZ2 pCR2.1 pBBR1. Salmonella cultures spotted onto Congo r ed media and incubated at 30C for 48 hr. An Olympus M VX10 confocal microscope and Olympus MVX TV1XC camera was used to photog raph the strains The S. Typhimurium 14028 has the rdar morphology, with MKF5, the spontaneous 14028 kanamycin resistant strain and JS246 have similar rdar phenotype s The MKF1 stra i n appears as the pdar morphology, MKF2 displays bdar, and the double mutant, WJZ2, has a light pink to saw phenotype. The phenotype of MKF1 pMKF4 is an exaggerated rdar morphology The bcs and double mutant complement strains indicate only partial resto ration of the rdar phenotype with an increas e in red color, but little patterned aggregative appearance. The three empty vector controls do not indicate phenotype restoration from the plasmid presence alone.
52 Figure 3 2. Qualitative assessment of cellu lose production of Salmonella Typhimurium strains. A) S. Typhimurium 14028, B) MKF5, C) JS246, D) MKF1, E) MKF2, F) WJZ2, G) MKF1 pMKF4, H) MKF1 pWSK29, I) MKF2 pMKF2, J) MKF2 pCR2.1, K) WJZ2 pMKF2 pMKF5, and L) WJZ2 pCR2.1 pBBR1. Above are the S. Typhim urium strains under dark conditions on Calcofluor White (CW) a gar. Photographs were taken with a digital camera on the UV trans illuminator after the plates were incubated for 48 hr at room temperature
53 0 100 200 300 400 500 600 Relative Fluorescence Units Salmonella strain Figure 3 3. Quantitative asse ssment of cellulose production of Salmonella Typhimurium strains. The graph represents the quantified relative fluorescence units of cellulose produced by each strain. Cells were resuspended on the CW agar in DI water and the fluorescence was recorded in 96 well plates at an excitation of 360 40 nm and emission of 460 40 nm. Optical density of the cultures was measured at 600 nm. Relative flu orescence was then calculated by fluorescence of sample minus the fluorescence of control divided by the OD 60 0 of sample minus the OD 600 of control and is expressed as fluorescence per unit of OD 600 standard deviation from three replicate studies.
54 0 2 4 6 8 10 0 1 2 3 4 5 6 7 8 9 10 11 12 Average Log CFU/ml Time (Hours) MKF5 JS246 MKF1 pMKF4 MKF2 pMKF2 WJZ2 pMKF2 pMKF5 MKF1 pWSK29 MKF2 pCR2.1 WJZ2 pCR2.1 pBBR1 Figure 3 4 Growth c urves of S almonella Typhimurium 14028 and mutant strains for 12 hr at 30 C A) Illustrates the growth curves of the S. Typhimurium 14028 strain along with the three deletion mutants de ficient in AgfD, cellulose, and cellulose and thin aggregative fimbriae production. B) Illustrates the growth curves of S. Typhimuriu m MKF5 and JS246, two modified strains of 1 4028, as well as the genetic complements and empty vector control strains of the three deletion mutants. Overnight cultures of each strain were diluted to the approximate concentration of 10 4 CFU/ml and incubated at 30C, shaking at 200 rpm for up to 12 h r Samples were diluted and plated once per hour, including hour 0. The final growth curves were the average of three replicates per strain with error bars indicating standard deviation 0 2 4 6 8 10 0 1 2 3 4 5 6 7 8 9 10 11 12 Average Log CFU/ml Time (Hours) 14028 MKF1 MKF2 WJZ2 B
55 CHAPTER 4 THE ROLES OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE IN THE INITIAL ATTACHMENT AND PERSISTENCE OF Salmonella Typhimurium ON THE TOMATO SURFACE Background There are numerous routes of contact between produce and Salmonella spp. in the field environment including contaminated irrigation water, soil, manure, fecal m atter, or through animal reservoirs ( 42 67 79 ) The microbi ological quality of irrigation water is crucial for maintaining the safety of produce. Salmonella spp. also survive in water sediments, making flooding of fields, either seasonally or from stor m events, a potential source of both produce and soil contamination ( 6 79 ) The introduction of Salmonella spp in produce fields can also occur through the use of improperly composed manure. The use of compost is beneficial in the maintenance of soil fertility, but improperly composed manure may contain gastrointestinal pathogens from the animal source(s), inclu ding Salmonella spp. S. Typhimurium was found to survive for more than 200 days in soil amended with contaminated compost ( 42 ) Field contamination from feces may result from domestic or wild animals roaming in the fields as well. Salmonella spp. are able to colonize both cold blooded and warm blooded animals. The colonization of the animals does not necessarily infer illness for the host, resulting in an imal reservoirs or carriers. S. Typhimurium and S. Enteritidis are considered to have a broad host range since they cause disease in a wide range of animals. S. Typhi and S. Gallinarum are host specific serovars, with the capability of causing disease in only humans and fowl, respectively. The initial introduction of the bacterium into the animal host is often through a food or water source, where the animals further shed the bacterium back into the environment ( 101 ) The shedding
56 events may occur whether the Salmonella is virulent or non virulent in the host. Birds and flies are among many animal vect ors capable of widespread dissemination of Salmonella spp. in the environment ( 24 ) Once the bacterium has re entered the environment, plants may be utilized as a vehicle for p athogen transfer. Thus, enteric bacteria may exhibit mechanisms to survive on plants to use them as a transfer medium between animal hosts ( 7 ) The genetic mechanisms employed in the interactions between Salmonella spp. and prod uce are not yet well understood Thr ough any of the previously mentioned routes of contamination, the Salmonella spp. must attach and persist in the plant environment in order to cause an outbreak event. In previous research on the attachment and survival of Salmonella spp. on produce, Ukuku and Sapers reported S. Stanely to be firmly attached to the surfa ce of cant aloupes within 4 hr ( 87 ) and Iturriaga et al. indicated that 0.3 0.7% of Salmonella inoculum (10 5 10 8 CFU/ml) attached to the tomato surface within 30 sec of contac t ( 44 ) It has also been demonstrated that S Montevideo persisted on tomato surfaces at 20 30C and survived at 10 30C at 45 to 60% relative humidity (RH) fo r up to 18 d ays ( 105 ) S. Montevideo persistence on tomatoes was also associated with humidity conditions, with increased pe rsistence at higher relative humidity (60 97% RH) ( 43 ) The aboveground surface s of plant s are protected from desiccation and bacterial infiltration by the plant cuticle, waxes, and polysaccharides ( 77 ) Bacterial attachment to surfaces is charac terized as a two step process including reversible phase and an irreversible attachment Reversible binding is hypothesized to be directed by physiochemical properties of the bacterium and surface with f lagella, pili, and fimbriae
57 suspected to influence t his first step of attachment. The stronger, irreversible attachment is hypothesized to involve both chemical and physical properties with extracellular components, such as cellulose, contributing to attachment. There is also evidence that there are locat ions of favored attachment on plant surfaces including the cuticle, stomata, and/or trichomes ( 44 84 ) In Salmonella, AgfD directly regulates thin aggregative fimbriae and indirectly regulates cellulose biosynthesis through adrA ( 100 ) agfD mutant of S. Typhimurium 14028 demonstrated reduced adherence to glass, pellicule formation in liquid media, and extracellular matrix production at 30 C ( 63 ) Salmonella strains lacking thin aggregati ve fimbriae production have demonstrate d significantly reduced attachment to polystyrene and glass surfaces ( 45 ) T he production of cellulose in Salmonella is generally used for protection and plays a structural role in the strength of biofilms ( 31 76 ) Salmonella mutants deficient in cellulose f ailed to attach or form biofilm to chitin beads or fungi ( 10 ) The attachment of Salmonella mutants deficient in b oth thin aggregative fimbriae and cellulose to parsley were significantly reduced ( 49 ) P revious research also indicated the imp ortance of thin aggregative fimbriae and cellulose in the attachment of S. enterica to alfalfa sprout seedlings ( 4 5 ) In this study, the aim was to characterize the roles of the AgfD regulator and AgfD regulated thin aggregative fimbriae and cellulose in the initial attachment and persistence of Salmonella spp. on the surface of green and red tomatoes. Previous rese arch indicates that Salmonella spp. is capable of rapidly attaching to the surface of tomatoes. Furthermore, the AgfD regulated thin aggregative fimbriae and cellulose production of Salmonella spp. are propos ed to have important role s in the attachment to
58 both abioti c and plant surfaces. This information was used to develop the hypothesis that S. Typhimurium deficient in AgfD regulated thin aggregative fimbriae and cellulose production would demonstrate reduced initial attachment and persistence on the to mato surface. The role of the AgfD regulator was also assessed to maintain if AgfD independent thin aggregative fimbriae and cellulo se production could be achieved in the strains. Materials and Methods Culture Maintenance All Salmonella strains were maint ained as frozen glycerol stocks in LB broth with 35% glycerol. The strains were grown through sub culturing from stocks into 5 ml of LB broth with appropriate antibiotics Antibiotics were utilized at the following concentrations: ampicillin (Amp), 10 0 g/ml; gentamicin (Gent), 50 g/ml ; kanamycin (Kan), 50 g/ml, and tetracycline (Tet), 10 g/ml. Cultures were prepared through overnight incubation at 30 C, shaking at 200 rpm prior to all experiments unless specified. The determination that incubation while shaking did not alter the attachment and biofilm formation abilities of the strains is detailed in Appendix B. Acquisition of Tomato Samples All produce utilized in the study was unwaxed. Red, vine ripe, round tomatoes cultivar Campari were obtained from the local supermarket (Publix Grocery). According to the packaging label, the tomatoes originated from Mexico, Canada, or New Jersey, USA. The green and red unwaxed, unprocessed tomatoes cultivar s Florida 47 and Quincy were obtained from various lo cations throughout Florida, including the U niversity of F lorida I nstitute of F ood and A gricultural S ciences (IFAS) North Florida Research and Education Center in Marianna, FL ; DiMare Fr esh in Tampa, FL; Pacif ic T omato Growers
59 in Palmetto, FL; W est Coast To mato in Palmetto, FL ; and Gadsden Tomato Company in Quincy, FL Green tomatoes refers to ripeness stage 1 and red tomatoes refers to ripen ess stage 6 of the US Department of Agriculture (USDA) color classification in the US Standards for Grades of Fresh T omatoes ( 88 ) Evaluation of H ydr ophobicity of Salmonella S trains Evaluation of hydrophobicity was performed using the bacterial adherence to hydrocarbons (BATH) method first described by Rosenberg et al. ( 65 ) and a modified salt agglutination test (SAT) ( 52 102 ) In preparation of the BATH assay, o vernight cultures of S. Typhimurium 10428, MKF1, MKF2, WJZ2, and JS246 were prepared and washed twice in PUM buffer (Appendix A) and diluted to an optical density (OD) of 0.8 when measured at 400 nm ( OD 400 ) Round bottom test tubes were filled with 1.2 ml of n hexadecane ( Acros Organics, Geel, Belgium) w ere added to the washed cells Following a prein cubation at room temperature (22 C ) for 10 min, the solutions were vortexed for 120 sec and incubated at room temperature for 15 min to allow for hydrocarbon separation. The aqueous phase was carefully rem oved with a Pasteur Pipette and transferred to a 1 ml cuvette where the optical density was read at 400 nm by the BioS p ec mini spectrophotometer (Shimadzu, Kyoto, Japan) The results were recorded as the percentage absorbance of the aqueous phase after tr eatment relative to the initial absorbance of the bacterial suspension The equation is diagramed as follows [(OD 400 original bacterial suspension OD 400 aqueous phase sample)/(OD 400 original bacterial suspension)] x 100. The assay was performed with t hree biological replicates
60 For the SAT assay, S. Typhimurium 14028, JS246, MKF1, MKF2, and WJZ2 were grown on LB agar at 37 C overnight. The cells were suspended in 0.002 M sodium phosphate buffer (Appendix A) washed twice and diluted to 10 8 CFU/ml base d on OD 600 and plate counts Once diluted, with an equal volume of varying concentrations of ammonium sulfate ranging from 0.2 to 4.2 M in 0.5 M increments in 0.002 M sodium phosphate buffer. The bacterial an d ammonium sulfate solutions were mixed for two minutes and incubated at room temperature for 30 min. The lowest molar concentration of ammonium sulfate producing visible aggregation was scored as the numerical value for the bacterial surface hydrophobici ty, the SAT value. Desiccation of Salmonella on the Tomato Surfac e Overnight cultures of Salmonella Typhimurium 14028 and MKF1 were washed three times in sterile DI water and diluted to a starting concentration of 10 7 CFU/ml. The initial inocula were dilu ted and plated onto XLD. Vine ripened, red Campari tomatoes were then spot inoculated with 10 spots in volumes of 10 l on each tomato. The tomatoes were stored at approximately 22 C, 60 80% RH. After 5 min, two tomatoes of each inoculum were placed into individual stomacher bags filled with 100 ml of PBS. Each bag was shaken for 30 sec and the rinsates were diluted and plated onto xylose lysine desoxycholate (XLD) agar (Becton Dickson, Franklin Lakes, NJ). This procedure was repeated at 15, 30, 60, and 120 min. The plates were incubated overnight at 37C to obtain plate counts of the surface concentrations of recovered Salmonella Statistical analysis was completed th rough JMP 9.0 Pro with paired t tests of the recovered concentrations at each t ime po int with a significance set to p<0.05.
61 Assessment of Tomato Inoculation Procedures The study was performed to compare the concentrations of recoverable Salmonella from the surface of vine ripened, red Camp ari tomatoes after either spot or dip inoculation. Overnight cultures of S. Typhimurium 14028, MKF2, and WJZ2 were w ashed three times in sterile DI water The strains were diluted with 1 ml of each strain added to 9 m l of sterile DI water to result in three separate inocul a The initial inocul a were dil uted and plated onto XLD. The blossom end of the Campari tomatoes wer e dip inoculated for 2 min in S. Typhimurium 14028, MKF2 or WJZ2 inocul a For the spot inoculation studies, the washed cultures were directly used to inoculate 10 spo ts of 10 l each o nto the blossom end of the Campari tomatoes. The initial inocul a were diluted and plated onto XLD agar. The studies were performed in triplicate, with three biological replicates. Tomatoes rested at room temperature for 5 min and were then rinsed in 10 m l of sterile DI. The rinsates were diluted and plated onto XLD agar. The tomatoes were stored in a two level environmental chamber maintained at 22 C, 60 80% relative humidity (RH). After 2 hr the blossom end s of the tomatoes were rinsed in 10 ml of st erile DI and the rinsate s w ere plated onto XLD. Tomatoes were then sterile l y moved into individual stomacher bags (Whirl Pak, Fort Atkinson, WI) containing a membrane filter and 100 ml of PBS. The tomatoes we re rubbed by hand for 60 sec with the blossom scar facing the membrane filter. Rinsate from th e filtered side of the bag was collected and plated onto XLD. Plates were incubated overnight at 37 C to obtain plate counts of recovered bacterial concentrations from the tomato surface Preliminary tests were also performed to show the membrane filter was not resulting in decreased plate
62 counts due to bacterial binding to the filter. The final results were averaged from three biological replicates performed in triplicate for a sample size of 9 tomatoes p er strain Initial Attachment of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 to the Tomato Surface Overnight cultures of S. Typhimurium 14028, MKF1, MKF2 and WJZ2 were washed three times in sterile DI water and diluted to a concentration of 1.0 x 1 0 8 CFU/ml. The initial inocul a were diluted and plated onto XLD agar. The washed cultures were used to inoculate the blossom end of unwaxed, unprocessed green Florida 47 tomatoes 10 times in 10 l volumes (for a total of 100 l per tomato) The inoculat ed tomatoes remained at room temperature for 5 min. The blossom end of the tomatoes underwent an initial rinse in individual containers of 10 ml of sterile DI water The rinsates were diluted and spread on XLD. Tomatoes were stored in a two level enviro nmental chamber at 22 C, 60 80% RH for 2 h r After 2 hr the blossom end of the tomatoes were rinsed in 10 ml of PBS and were placed into Whirl Pak stomacher bags containing 100 ml of PBS. Each tomato was rubbed by hand in the stomacher bags for 60 sec. The rinsates from the secondary rinse and from the stomacher bag were diluted and plated onto XLD agar. The plates were incubated overnight at 37C to ob tain plate counts of the recovered Salmonella concentrations. Average results were obtained from thre e biological replicate s, performed in triplicate Analysis of variance (ANOVA) comparisons were performed for each the initial rinse, s econdary rinse, and tomato rinse results with JMP 9.0 Pro statistical software. Competitive Co infections of Salmonell a S trains on the Tomato Surface Overnight cultures of S. Typhimurium 14028 JS246, MKF1 MKF2, and WJZ2 were washed three times in sterile DI water The washed cells were measured at 600
63 nm in a spectrophotometer and a 1:1 concentration ratio inoculum was generated for 14028:JS246, 14028:MKF1, 14028:MKF2, and 14028:WJZ2. The initial inocul a were diluted and plated onto XLD. The blossom end of unwaxed, unprocessed, green and red Florida 47 tomatoes were each dip inoculated in to a co inoculum for 2 min. T he to matoes rested, blossom end up, at room temperature for 5 min. The tomatoes then underwent an initial rinse in a container of 10 ml of sterile DI. The rinsate was diluted and plated o nto XLD. The tomatoes were stored in an environmental chamber at 2 2 C, 60 80% RH for 2 hr. After 2 hr the blossom end s o f the tomatoes were again rinsed in 10 ml of sterile DI. A sample of the rinsate was directly plated onto XLD. The blossom end of the tomato was then swabbed and spread on a section o f an XLD plate and the sample was streaked for isolation with a sterile, metal loop. The plates were incubated at 37C overnight. One hundred c olonies from each of the initial co inoculum, initial rinse, and tomato swab plates were patched onto either LB agar with tet racycline or kanamycin. Samples of the 14028:JS246 co inoculum were patched onto LB agar with tetracycline to determine the number of tetracycline resistant JS246 colonies. The remaining samples were patched onto LB agar with kanamycin to determine the n umber of kanamycin resistant colonies. The numbers of antibiotic resistant colonies from the plates were used to determine the wild type to mutant ratios. Changes in the wild type to mutant ratio between the initial co inoculum and the samples recovered from the tomato surface were used to calculate the competitive index with the following equation, c ompetitive index ( CI ) = [Mutant out /14028 out ]/ [Mutant in /14028 in ]. The sample size was 15 green or red tomatoes for each strain. The resulting CI values wer e log transformed
64 and statistical analysis was performed in JMP 9.0 Pro by assessing pairwise t t ests of the CI results for the 14028:mutant studies versus the 14028:JS246 control at a significance of p<0.05 Persistence of Salmonella Strains in Buffered Solution O vernight cultures of S. Typhimurium 14028, MKF1, MKF2 and WJZ2 were washed three times in sterile PBS with a final wash and suspension in 0.1 M p otassium p hosphate b uffer (Appendix A) A set of five test tubes were filled with 5 ml of sterile 0.1 M p otassium p hosphate b uffer and 5 l of each washed culture was added to result in a starting culture of 10 6 CFU/ml. The remaining uninoculated tube served as a negative control. After vortexting, a 100 l sample was removed from each culture tube. The sample was diluted and plated onto LB agar. The plates were incubated at 37 C overnight to obtain plate counts. The culture tubes were left at 22 C for 20 days and were sampled on days 0, 1, 2, 4, 6, 8, 10, 12, 14, 16 18, and 20 The experiment wa s performed with three biological replicates. Persistence of Salmonella Strains on the Tomato Surface Overnight cultures of S. Typhimurium 14028, MKF1, MKF2 and WJZ2 wer e washed three times and serially diluted twice in 0.1 M potassium phosphate buffer to result in an approximate concentration of 10 7 CFU/ml. The initial inocul a were diluted and plated onto XLD. The blossom end of five green and five red Florida 47 tomatoes were spot inoculated 10 times in 10 l volumes with each inoculum After 10 min o f rest at room temperature one tomato of each inoculum was rinsed in a container 10 ml of sterile PBS The rinsate was diluted and was plated onto XLD. The plates were incubated overnight at 37 C The tomatoes were also sampled on days 1, 2, 4, and 6. The tomatoes were stored in an environmental chamber at 22C, 60 80% RH until
65 culture recovery. The average recovery data was obtained from six biological replicates. Statistical analysis of the persistence data was performed using JMP 9.0 Pro. ANOVA a nd post hoc pairwise t tests were comp leted with a significance of p<0.05. A control assay was also performed alongside the previously described recovery process. After the tomatoes were sampled and the rinsates were plated, the tomatoes were weighed. A volume of PBS solution equal to that of the tomato weight was added to the s tomacher bag to make a 1:1 volume to weight ratio. The tomatoes were stomached in a Stomacher 4000 Circulator (Seward, West Sussex, UK) at 250 rpm for 30 sec and dilutions were pl ated onto XLD. The XLD plates were incubated at 37C overnight to obtain plate counts. Con trol tomatoes of each maturity were also inoculated on day 0 with each inoculum with 10 spots of 10 l each and were sampled without rinsing after 10 min of rest at room temperature. The control tomatoes were used to determine if the original inoculum could be recovered from the stomaching recovery process. The control assay was performed to determine if a decrease in persistence over time was due to the loss of vi able culture on the tomato surface or due to attached cells that could not be removed from the tomato surface though the rinsing process alone. Persistence studies of the Salmonella complementation and vector controls were also performed on unwaxed, unpr ocessed green and red Florida 47 tomatoes. The assay was completed with S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, WJZ2 pMKF2 pMKF5 and the empty vector controls, MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 There were a total of four biological re plicates for each strain at
66 each maturity level and to ensure plasmid retention while on the tomato surface, the recovered colonies were patched onto LB agar supplemented with the appropriate antibiotics. The plates were incubated at 37 C overnight. Resu lts Initial Attachment of Salmonella to the Tomato Surface The investigation of the role of the AgfD regulator and AgfD regulated thin aggregative fimbriae and cellul ose i n the surface attachment of the pathogen to the surface of tomatoes was performed wit h three deletion mutants derived from S. Typhimurium 14028. Since it was propos ed that the initial, reversible attachment of bacteria m ight be mediated by physiochemical proper ties, the overall surface properties of the mutant strains were assessed using the BATH and SAT assay s There were no significant differences between the percent absorbance for the strains under all conditions, indicating no overall changed in adherence properties to hydrocarbons was displayed by the mutant strains (data not shown). S. Typhimurium 14028, JS246, MKF1, MKF2, and WJZ2 all had minimal absorption with a maximum of only 7.6% indicating very hydrophilic properties. The SAT assay resulted with values of 4.2 for the surface hydrophobicity of all the cultures, also indicati ng hyd rophilic properties ( 50 54 ) After the characterization of the adherence properties of the Salmonella strains initial studies were performed to determine the optimal conditions to investigate Salmonella attachment The contact time between the S Typhimurium and tomatoes in the studies was minimalized to maintain conditions that may occur in the field. A study was performed to determine if MKF1 ( agfD::F RT K an FRT) was more susceptible to desiccation on the tomato surface over time in comparison to S. Typhimurium 14028. The results indicate d that desiccation does occur over time, but there was no significant
67 difference in recovery of 14 028 or MK F1 at any time point ( Figure 4 1 ). Furthermore, at 5 min there was a modest reduction of ~0.5 log 10 CFU/ml in recovery from the initial inoculum concentration for both S. Typhimurium 14028 and MKF1 The modest reduction of recovery could be attributed to the attachment of the Salmonella to the tomato surface since it has been documented that attachment of S. Montevideo occur s within 30 sec on the tomato surface ( 4 4 ) The results of Iturriaga et al. ( 44 ) along with the desiccation study indicated that a contact time of 5 min of Salmonella with the tomato surface was su fficient to study initial attachment. The ideal inoculation procedures for the attachment assays were also assessed through comparison of spot versus dip inoc ulation. The results show a 10 fold loss of recovery with the dip inoculation compared to spot i noculation procedures ( Figure 4 2 ). There is also uncertainty of the bacterial concentrations that actually contact the tomato surface and the surface area of the tomato that is inoculated with the dip inoculation procedures. The results indicate that sp ot inoculation should be used in any assay resulting in plate counts as the measure of attachment to the tomato surface. The investigation of the initial attachment of S. Typhimurium 14028, MKF1, MKF2, and WJZ2 on the tomato surface was completed on unwaxe d, unprocessed, green field grown Florida 47 tomatoes. The assay was designed to compare overall attachment of the strains side by si de The strains were recovered at three time points The f irst recovery was the initial rinse, which was performed to rem ove all cells that did not initially attach to the tomato within 5 min of contact. The initial rinse step utilized DI water rather than PBS to not alter the condition or add any buffering capacity to the cells remaining attached to the tomato surface. Af ter 2 hr at 22 C, 60 80% RH, the strains
68 were recovered through a secondary rinse which was performed to remove cells that had initially, reversibly attached but did not remain attached. The final tomato rinse, which involved the physical removal of the c ells by rubbing the tomato, was completed to obtain the final counts of attached cells. Therefore, in terms of this investigation, attachment is defined as the Salmonella remaining attached after two surface rinses but can be subsequently removed through physical abrasion including hand rubbing and swabbing. The results of the surface attachment studies indicated there were no significant difference s in the initial attachment of S. Typhimuri um 14028, MKF1, MKF2, and WJZ2 to the tomato surface ( Figure 4 3 ) To more directly determine the effect of the mutation s in the initial attachment of Salmonella a competitive index (CI) assa y was employed. A competitive index assay uses ratio s that indicate the fitness of a genetic mutant in comparison to the wild ty pe within a specific environment. In this case, the fit ness of the deletion mutants on the tomato surface was compared to that of S. Typhimurium 14028 on the tomato surface. The tomatoes were inoculated with a 1:1 ratio of mutant and wild type culture. The fitness of the mutant is determined by comparing the initial 1:1 ratio to the ratio of the cells recovered from the tomato surface in the initial rinse, secondary rinse, and tomato swab. The data from the secondary rinse are not shown since the inform ation from the initial rinse and tomato swab were most informative in determining initial attachment Dip inoculation was utilized in the CI studies s ince bacterial concentrations were not determined from plate counts The results of F igure 4 4 and T able s 4 1, 4 2, and 4 3 show that the gene deficiencies do alter the fitness of the Salmonella despite not changing t he overall attachment in the side by side studies ( F igure 4 3 )
69 In F igure 4 4 graphs A and B show the CI values of the initial rinses of the green and red tomatoes, respectively. A CI greater than zero indicates an increase of fitness for the mutant strain in comparison to the wild type, and a CI value less than zero indicates a decreased f itness for the mutant strain. For the initial rinses, a ll strains except for S. Typhi murium MKF1 on red tomatoes resulted in a CI greater than zero ( Figure 4 4 ) This dictates that the ratios obtained from the initial rinses contained a higher ratio of mutant cells than wild type. The cells recovered from the initial rinse of the tomatoes represent the cells that did not initially attach, indicating that strains deficient in thin aggregative fimbriae and/or cellulose production did not attach to the tomato surface as efficiently as the wild type strain. St atistical significance was determined for the initial rinse data for WJZ2 within the green tomato samples and MKF1 within the red tomato data (Table s 4 1 and 4 2) Statistical analysis was completed through comparative t t ests at a significance of p<0.05 with the modified S. Typhimurium 14028 strain, JS246, acting as a control. The CI results for the 14028:JS246 co inoculum are minimal since the res tet res insertion into JS 246 should not alter attachment; the 14028:JS246 results were used to represent th e variability present within the assay. Graphs C and D of F igure 4 4 show the CI val u es of the Salmonella recovered from the tomato swabs of the green and red t omatoes, respectively. T he three deletion mutants all show a similar change in fitness, with a ll CI values being less than zero in the tomato swab results ( Figure 4 4 ). The cells recovered from the tomato swab s represent the attached cells on the tomato surface, so the negative CI values indicate that more wild type cells had been attached to the tomato surface than the mutant
70 strains. The results of both the initial rinse and tomato swab recovery steps show a decrease in fitness of the MKF1, MKF2, and WJZ2 mutant strains in the attachment to the tomato surface. For the Salmonella attachment to g reen tomatoes, s tatistical significance was determined for WJZ2 recovered from the tomato swab (Table 4 1). Furthermore, the deficiency of the AgfD regulator in MKF1 had a greater effect on red tomatoes than green, while cellulose deficiencies of MKF2 and WJZ2 reduced the fitness of the strains on green tomatoes more than on red tomatoes ( F igure 4 4 ). In comparing the CI values of each strain from green and red tomato es the CI values of WJZ2 recovered from the tomato swab were significantly different in green versus red tomatoes (Table 4 3) Persistence of Salmonella on the Tomato Surface In the persistence of Salmonella on the tomato surface, several suspension medias were compared. The suspension media utilized in the initial attachment studies was ste rile DI water, which was suitable due to the short term nature of the study (2 hr). In testing the persistence of S. Typhimurium on the tomato surface 5 days, the concentrations of S. Typhimurium 14028, MKF1, MKF2, and WJZ2 in the D I suspension fell below the level of detection after day 1 (data not shown) Two other solutions were used to assess the impact o f the carrier medium 0.1 M potassium phosphate buffer and 0.1% buffered peptone water ( BPW ) (Becton Dickson, Franklin Lakes, NJ) In both media, th e Salmonella was recoverable from the tomato surface over 5 days. The concentration of the Salmonella control, strain 14028 in 0.1% BPW modestly increased by 1 log 10 CFU/ml from day 2 to day 5 This indicated that the Salmonella was utilizing the nutrie nts provided in the BPW to sustain on the tomato surface. This suggests that Salmonella may be nutrient limited on the tomato surface and the reduction of viable
71 cell numbers (when applied in DI or PBS) may be due to a combined effect of nutrient limitati on and desiccation The S. Typhimurium 14028 suspended in the 0.1 M potassium phosphate buffer is buffered from osmotic stress, but is not provided with sufficient growth nutrients. In the buffer, the concentration of recovered Salmonella on the tomato s urface wa s also within the recoverable range throughout the 5 days and did not increase over time These results conclude d that the persistence assays would be performed with culture suspended in 0.1 M phosphate buffer to minimize osmotic stress while det ermining persistence of Salmonella on the tomato without an artificial nutrient source The results of the persistence assay indicate an advantage for cultures producing cellulose on the surface of tomatoes ( Figure 4 6 ). The resulting co lony forming units of the cellulose deficient mutants diminished 5 fold on green tomatoes and 2.5 to 3 .0 fold on red tomatoes over 6 days. A comparison of the recovered concentrations on green and red tomat oes resulted in a significant difference on day 6 for both MKF2 an d WJZ2 strains. The resulting average concentrations for MKF2 and WJZ2 on day 6 of the green tomato samples f e ll below the level of detection because three of the replicate studies resulted in recovered concentrations just above the lev el of detection, th e remaining three replicates had concentrations below detection limits of the assay. Only a modest reduction was observed in the wild type culture, indicating the cellulose deficiency was responsible for the reduced persistence as opposed to desiccation s tress. The S. Typhimurium MKF1 strain sustained on the tomato surface similar to the wild type ( Figure 4 6 ). ANOVA assays of the green tomato data indicated significant differences in the recovered cultures on days 4 and 6, with only day 6 recovered
72 con c entrations being significant within the r ed tomato data. The post hoc t test results are documented in Tables 4 4 and 4 5. In comparing the concentrations recovered from the gre en and red tomatoes, significance was determined for MKF2 and WJZ2 on day 6, as previous stated. The control studies of the stomached tomatoes indicated that the S. Typhimurium was being effectively removed from the tom ato surface through the rinsing steps (data not shown) In the control studies, there was a slight loss in the r ecovery of the initial inoculum, but in testing the tomatoes that had undergone the rinsing and sampling for the persistence assay, the resulting concentrations were below the level of detection for all samples. The persistence of the S. Typhimurium comp lement and empty vector strains were al so investigated ( Figure 4 7 ). Graphs A and C illustrate the persistence of the complement ed mutants S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, and WJZ2 pMKF 2 pMKF5, on green and red tomatoes, respectively. The strains maintain in initial population level s until day 4 on both green and red tomatoes. On day 4, S. Typhimurium MKF1 pMKF4 sustained while the recovered concentrations of S. Typhimurium MKF2 pMKF2 and WJZ2 pMKF2 pMKF5 decreased by 1 log 10 CFU/ml on both green and red tomatoes Recovery of all thre e strains decreased on day 6 on both green and red tomatoes, but the recovered colony forming unit s for the MKF2 pMKF2 and WJZ2 pMKF2 pMKF5 strains were 1.5 to 2 .0 fold great er than those of the respective deletion mutants for days 2 6, indicating a partial restoration of the phenotype. The empty vector control strains tested on green and red tomatoes did not restore the phenotype to the level of the complemented strains, as shown in graphs B
73 and D of Figure 4 7 T he patch plating of the complement and vector control strains on LB agar supplemented with antibiotics selecting for the plasmids resulte d in 100% growth for all samples This indicated that the plasmids were maintained under the test conditions and that plasmid loss was not a confounding factor in the experiments despite the lack of antibiotic pressure. Discussion The first stage of produce contamination is the initial attachment of the bacteria to the produce surface. The objective s of the studies were to characterize the role s of AgfD regulated thin aggregative fimbriae and cellulose production as well as the AgfD regulator in the initial attachment and persistence of Salmonella on the surface of tomatoes. The scope of the research foc used on the fie ld environment and surface contamination of the produce surface in the field. Therefore, the tomatoes used in the experiments were unwaxed tomatoes, either vine ripened Campari tomatoes or field tomatoes from various areas in Florida. It was important to use unwaxed tomatoes, since the waxed surface may alter the attachment and persistence of the bacterium. The use of waxed, processed tomatoes would also not effectively simulate field contamination. The vine ripened red, Campari tomatoes were utilized f or initial preliminary studies due to their availability. The attachment and persistence studies were completed with the green field tomatoes since this is the stage of maturity round tomatoes are harvested. Red field tomatoes were further tested to disc over if there was a significant difference in the attachment and persistence of Salmonella to green (unripened) versus red (ripened) tomatoes. Lastly, the attachment studies used sterile DI water as the suspension and initial rinse medium to remove extern al factors such as nutrient accessibility or pH buffering that would be provided by nutrient rich or buffered
74 solutions. The buffering was necessary to maintain Salmonella on the tomato surface for more than 2 days, so 0.1 M potassium phosphate buffer was utilized for the persistence assays. The media supplied the necessary buffering capacity but did not support growth by providing nutrients. The suspension solutions simulated extreme conditions with a lack o f nutrients and organic debris which is not l ikely to o ccur in the field, but implies that the results seen in the lab may be attainable in the field environment where the contamination conditions may be more favorable to the bacteria. It was hypothesized that the S. Typhimurium strains deficient i n AgfD regulated thin aggregative fimbriae and cellulose production would demonstrate reduced initial attachment and persistence on the tomato surface. Complete loss of attachment was not expected since the strains were only single or double deletion muta nts. Attachment is a cooperative mechanism where no individual gene is responsible for initial attachment ( 4 5 ) T he BATH and SAT assays indicated that any change in initial attachment was not due to modifications of the overall surface properties of the strains; but the side by side attachment studies showed that there was no overall change in the initial attachment of Salmonella strains ( Figure 4 3) The recovery process was not sensitive enough to determine the effect of the genetic deficiencies of the S. Typhimurium strains. The question being addressed wa s the role of thin aggregative fimbriae and cellulose in th e initial attachment to the tomato surface. The variability between tomatoes, tomato microflora, and dilution and plating procedures of the assay might have been greater than the effect caused by the deletion mutations. The original question of the role of thin aggregative fimbriae and cellulose deficiencies in initial attachment was reassessed through competitive index assays. The CI assays indicated
75 that impairing agfD, agfB, and/or bcs in S. Typhimurium 14028 reduced the fitness and attachment capabil ities of the Salmonella on the surface of both green and red tomatoes. While there were some statistically significant results from the CI assay, the results do not indicate the differences would be biologically significant. The attachment studies indica te that while these extracellular components are involved in the initial attachment of Salmonella to the tomato surface, they do not predominantly dictate attachment to tomatoes. The initial attachment assays were performe d over a t ime frame of 2 hr wit h initial rinses occurring after only 5 min of contact. Under these time limitations, cellulose may not have been adequately produced To assess the role of AgfD regulated thin aggregative fimbriae and cellulose over a long time period on the surface of tomatoes, persisten ce assays were completed over 6 days. The extrinsic, environmental conditions are very important in the persistence of Salmonella on produce ( 43 105 ) so conditions were maintained at 22 C, 60 80% RH throughout the persistence studies. The Salmonella strains were assessed in the suspension media, 0.1 M phosphate buffer, over 20 days to ens ure any loss of persistence on tomatoes was not due to a deficiency of t he strain. All four strains persisted in the media over the 20 days, with no growth in the negative control, assuring the l ong term viability of the strain s ( Figure 4 5). The studies on the tomato surface indicate d that cellulose production has a role in the persistence of Salmonella The cellulose deficient strain s resulted in a 5 fold decrease in concentration on green tomatoes and 2.5 to 3 .0 fold loss on red tomatoes ( Figure 4 6) The partial restoration of persistence in the complement strains, and lack of phenotype restoration in the empty vector controls, also indicates that the decrease in
76 persistence of MKF2 and WJZ2 was due to the gene deletions rather than artifacts of the mutation process ( Figure 4 7) As covered in the introduction, the thin aggregative fimbriae and cellulos e of the AgfD regulated system have alternative regulators beyond AgfD. It is hypothesized that the a lternative regulators activate d thin aggregati ve fimbriae and cellulose production under the conditions of the tomato surface since the MKF1 strain sustained at the same concentrations as the wild type, 14028 The influence of tomato ripeness did not affect persistence of S Typhimurium 14028 or MKF1 so the differential persistence of MKF2 and WJZ2 on suggests that cellulose deficient Salmonella ar e more sustainable on red, ripe tomatoes versus green tomatoes On both green and red tomatoes, the 14028 and MKF1 strains experience a decrease of 1 log 1 0 CFU/ml over the 6 day trial In all of the analyses, t here was no statistical difference in the recovered concen trations of MKF2 and WJZ2, indicating that thin aggregative fimbriae had no significant role in the persistence of S. Typhimurium under the t ested conditions. A study of rdar and non rdar Salmonella strains obtained from human or poultry samples found that thin aggregative fimbriae deficient (bdar) strain s exhibited an attachment deficiency, but had greater attachment and persistence to polyst yrene and glass surfaces than the saw strain deficient in thin aggregative fimbriae and cellulose production ( 45 ) The difference in results betw een the two studies could be explained by the undetermined mutat ions of the bdar strain in their study. The S. Typhimurium strain was obtained from human samples and was characterized as bdar on CR and CW agar, but the mutation causing this morphology may not have be en in the agf operon, resulting in different attachment results.
77 It is hypothesized that either the recovery conditions were not precise enough to distinguish slight differences in persistence be tween the MKF1 and WJZ2 strains or the alterna tive fimbriae of the WJZ2 strain were utilized in initial attachment, making the gene deletion inconsequential in the attachment and persistence process. In the initial genetic analysis of fimbrial genes of Salmonella up to 15 putative fimbrial operons w ere identified ( 40 86 ) Selective p ressures from animal hosts may contribute to the multiplicity of fimbriae in Salmonella Fimbriae are externally expressed features that can trigger antigenic effects in a host. The partially redundant functionality of fimbrial operons allows for environmental regulators to change fimbriae expression and evade immune detection. The inactivation of single fimbrial operons in S Typhimurium did not affect overall virulence of the strain, but simultaneous inactivation of lpf pef and agf operons resulted in attenuated virulence ( 90 ) Understanding the redundant functionality of fimbriae in the animal host leads to the conclusion that the same may be true on the plant surface. Overall, the production of AgfD regulated cellulose in S Typhimurium 14028 has a signi ficant role in persistence on the tom ato environment, but little effect in the initial attachment of Salmonella under the tested condition s There was no significant reduction of initial attachment or persistence attributed to AgfD or thin aggregative fim briae deficiencies. In both the competitive index and persistence assays, the cellulose deficient mutants, S. Typhimurium MKF2 and WJZ2, were more sustainable on the surface of red, ripened tomatoes than green. This reduced persistence on green tomatoes indicates that both bacterial behavior, including cellulose production, and produce ripeness are factors in persistence of S. Typhimurium on the tomato surface.
78 Table 4 1. Statistical analysis of competitive index values of Salmonella recovered from the g reen tomato surface Sample comparison Sample recovery step p Value 14028:JS246 and 14028:MKF1 Initial rinse 0.769 Tomato swab 0.389 14028:JS246 and 14028:MKF2 Initial rinse 0.299 Tomato swab 0.331 14028:JS246 and 14028:WJZ2 Initial rinse 0.015 T omato swab 0.050 Table 4 2. Statistical analysis of competitive index values of Salmonella recovered from the red tomato surface Sample comparison Sample recovery step p Value 14028:JS246 and 14028:MKF1 Initial rinse 0.011 Tomato swab 0.103 14028: JS246 and 14028:MKF2 Initial rinse 0.151 Tomato swab 0.621 14028:JS246 and 14028:WJZ2 Initial rinse 0.062 Tomato swab 0.411 Table 4 3. Statistical analysis comparing competitive index values of Salmonella recovered from green versus red tomato surf ace
79 Table 4 4. Statistic al analysis of recovered Salmonella concentrations from the green tomato surface over 6 days Sample Day Significantly different strains Table 4 5. Statistical analysis of recovered concentrations of Salmonella from the red tomato s urface over 6 days Sample Day Significantly different strains
80 0 1 2 3 4 5 6 7 5 15 30 60 120 Log CFU/ml Recovered Time (min) 14028 MKF1 Figure 4 1 Desiccation of S almonella Typhimurium 14028 and MKF1 on the tomato surface over time. This study was performed to determine if significant desiccation of Salmonella was observed over 12 0 min on the tomato surface. The study assessed the recovered concentrations of S. Typhimurium at 5, 15, 30, 60, and 120 min. The recovered concentrations were obtained through plate counts. The study compiled data from three biological replicates, each performed in triplicate. The strain designations are labeled in the legend and t he error bars denote the standard error.
81 0 2 4 6 8 10 Initial Rinse Secondary Rinse Tomato Rinse Log CFU/ml Recovered 14028 MKF2 WJZ2 0 2 4 6 8 10 Initial Rinse Secondary Rinse Tomato Rinse Log CFU/ml Recovered 14028 MKF2 WJZ2 Figure 4 2 Recoverable Salmonella Typhimurium from the tomato surface through spot or dip inoculation. A) Illustrates the results from spot inoculation. B) Illustrates the results from dip inoculation. The study was designed to determine if there was a difference in recoverable Salmonella from the two inoculation procedures. The studies wer e performed with three biologica l replications in triplicate. The strain designations are labeled in the legends and t he error bars represent the standard error.
82 0 2 4 6 8 10 14028 MKF1 MKF2 WJZ2 Log CFU/ml Recovered Salmonella strain Initial Rinse Secondary Rinse Tomato Rinse Figure 4 3 The initial attachment of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 on the tomato surface. The recovery of Salmonella from the surface of green, unprocessed Florida 47 tomatoes was performed after 5 min (initial rinse) and 2 hr (secondary rinse), followed by abrasive removal of Salmonella through rubbing t he tomato (tomato rinse). Recovered Salmonella concentrations were obtained through plate counts on XLD. The represented data is the average of three biological r eplicates performed in triplicate The error bars indicate standard error.
83 Figure 4 4 Competitive c o infections of Salmonella strains on the tomato s urface. A) The CI values recovered from the initial rinse of green Florida 47 tomatoes. B) The CI values recovered the initial rinse of red Florida 47 tomatoes. C) The CI values recovered from the tomato swab of green Florida 47 tomatoes. D) The CI values recovered from the tomato swab of red Florida 47 tomatoes. The competitive index (CI) studies were performed to specifically measure the effect of t he deletion mutants on the fitness of the Salmonella on the tomato surface. CI values greater than zero indicate an increase in fitness for the deletion mutants in comparison to the wild type, while CI values less than zero indicate a decrease in fitness for the deletion mutant on the tomato surface. The data encompasses the results from five biological replicates, each performed in triplicate. The res tet res insertion of JS246 should not alter the competitive fitness and was used as a control for stati stical analysis. The box plots around the data points denote the 10%, 25%, 75% and 90% quantiles as well as the median. The diamonds represent the group mean as test were used to test for significance at p<0 .05. Statistical results represented in Tables 4 1, 4 2, and 4 3.
84 0 1 2 3 4 5 6 7 0 1 2 4 6 8 10 12 14 16 18 20 Log CFU/ml Recovered Day 14028 MKF1 MKF2 WJZ2 Figure 4 5 Persistence of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 in 0.1 M potassium phosphate buffer. The assay was performed to determine if the genetic mutants decreased in viability over a period of time. The mutants deficient in AgfD, thin aggregative fimbriae, an d/or cellulose production were sampled from the suspension media utilized for the persistence studies on the tomato surface for 20 days The data wa s averaged from three biological replicates. The strain designations are assigned in the legend; t he error bars indicate standard deviation.
85 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day 14028 MKF1 MKF2 WJZ2 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day 14028 MKF1 MKF2 WJZ2 Figure 4 6 Persistence of Salmonella Typhimurium 14028, MKF1, MKF2, and WJZ2 on the green and red tomato surface. A) Illustrates the persistence of Salmonella on green tomatoes. B) Illustrates the persistence of Salmonella on red tomatoes. The assay was performed to determine if deficiency of AgfD, thin aggregative fimbriae, and/or cell ulose production would alter the persistence of Salmonella on either green or red tomatoes. The data was averaged from six biological replicates. The strains are labeled in the legend. The error bars indicate standard error. The results were statistica lly analy zed through ANOVA and pairwise t t ests at a significance of p<0.05.
86 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day MKF1 pWSK29 MKF2 pCR2.1 WJZ2 pCR2.1 pBBR1 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day 0 1 2 3 4 5 6 7 0 1 2 4 6 Log CFU/ml Recovered Day MKF1 pMKF4 MKF2 pMKF2 WJZ2 pMKF2 pMKF5 Figure 4 7 Persistence of Salmonella complement and vector control strains on the surface of green and red tomatoes. A) Illustrates t he persistence of Salmonella complements on green tomatoes B) Illustrates the persistence of Salmonella empty vector controls on green tomatoes. C) Illustrates the persistence of Salmonella complements on red tomatoes. D) Illustrates the persistence of Salmonella empty vector controls on red tomatoes. The assay was performed to determine if the persistence deficiencies observed by the deletion mutants w ere the result of the gene deletion. The strain designations are detailed in the legends between the graphs, with A and C sharing a legend and B and D sharing the other legend. The data was obtained through six biological replicates and error bars indicate standard error.
87 CHAPTER 5 THE ROLES OF AgfD REGULATED THIN AGGREGATIVE FIMBRIAE AND CELLULOSE IN THE BIOFILM FORMATION OF Salmonella Typhimurium ON POLYSTYRENE, THE TOMATO SURFACE AND TOMATO SEGMENTS AND THE ROLE OF THE RDAR MORPHOLOGY IN ENVIRONMENTAL FITNESS Background The formation of biofilms facilitate bacterial survival in both host and non host environments. Biofilms are structured communities of bacterial cells enclosed in a se lf produced polymeric matrix adherent to either abiotic or biotic surfaces ( 80 ) The bacteria present in biofilms are generally protected from environmental stress, antibiotics, and disinfectants, making them difficult to eradicate. Salmonella spp. are known to form biofilm on numerous abiotic surfaces including plastic, rubber, glass, and stainless steel ( 58 76 80 ) Salmonella spp. may persist in the plant environment through the formatio n of biofilms on the plants. S. Thompson was visualized on the surface and within cuts of lettuce leaves through episcopic differential interference contrast microscopy coupled with epifluorescence to find what appeared to be biofilm formation on the lett uce ( 94 ) Salmonella have also demonstrated biofilm formatio n on the surface of parsley in studies that resulted in decreased biofilm from strains deficient in t hin aggregative fimbriae and cellulose ( 49 ) Human pathogens may integrate into the biofilms of the plant microflora for enhanced protection and survival ( 51 ) Salmonella also form biofilm on the hyphae of fungi, including Aspergillus niger, which may serve in the environmental survival of the pathogen ( 10 ) Biofilm formation may have an integral role in the persistence of human pathogens in the plant environment. Thin aggreg ative fimbriae and cellulose are important structural components of biofilm formation and Salmonella deficient in these components demonstrate reduced biofilm on glass, polystyrene, leafy greens, and Aspergillus niger ( 10 45 ) Transcription
88 of agfD is also greatest in conditions that would be encountered outside of the animal host, including 28 C, low osmolarity, and nutrient starvation during stationary growth ( 63 ) It has been implicated the environmental persi stence of pathogens is achieved through biofilm formation and that 30 80% of bacterial populations on plant surface are present in biofilms ( 51 80 ) It has been hypothesized that Salmonella exhibiting the rdar morphology are more resistant to desiccation and environmental stress, which is vital in environmental survival ( 100 ) It has also been proposed that non rdar morphology promotes fitness of S. Typhimurium within an infected tomato ( 104 ) If Salmonella spp. are forming biofilms on the plant surface to persist in the environmen t, there may be a corre lation between the rdar morphology and increased fitness in the plant environment. Of 204 Salmonella isolates acquired from food, environmental, and clinical samples, as well as some included laboratory strains, 66% resulted in the rdar morphology. In as sessing the strains for biofilm formation, only 3% were identified as biofilm deficient ( 76 ) Another research group determined the percent age of rdar positive Salmonella strains based on the source of the strain. The results included rdar strains in 3% of t he clinical strains 84% of strains obtained from meat samples, and 56% of produce related strains In assessing the biofilm formation of the produce related strains, all but one rdar strain exhibited strong biofilm formation in 96 well polystyrene microtiter plates; the non rdar strains lacked biofilm formation ( 78 ) Both the rdar morphology and biofilm formation of Salmonella may contribute to the environmental survival of the pathogen, which leads to the question of whether the rdar morphology corresponds to biofilm formation capabilities While evidence suggests that Salmonella displaying the rdar morphology may have increased biofilm formation
89 and protection from environmental stress, there were a high percentage of non rdar strains (44%) isolated from produce samples, suggesting that no n rdar strains are persisting on plants as well. S. Typhimurium 14028, MKF1, MKF2, and WJZ2 as well as genetic complements, negative controls, and produce outbreak non rdar Salmonella strains will be assessed for biofilm formation on both abiotic and gre en and red tomatoes to determine the roles of the AgfD regulator, AgfD regulated thin aggregative fimbriae and/or cellulose i n biofilm formation and evaluate the roles of the rdar morphology and biofilm formation in the environmental survival of Salmonell a Materials and Methods Biofilm F ormation of Salmonella Typhimurium on Polystyrene Maintenance and growth of the Salmonella strains, as well as tomato acquisition are as detailed in the previous chapter, unless stated otherwise. Overnight cultures of S Typhimurium 14028, MKF5, MKF1, MKF2, WJZ2 MKF1 pMKF4, MKF1 pWSK29, MKF2 pMKF2, MKF2 pCR2.1, WJZ2 pMKF2 pMKF5, and WJZ2 pCR2.1 pBBR1 MSC5 were each washed in sterile PBS three times. Cultures were diluted 1:100 in 1/10 t rypt ic s o y b roth (TSB) (Bect on Di ckson, Franklin Lakes, NJ) per well) into appropriate wells of a 96 well polystyrene plate (Costar, Washington, DC). The 96 well plate also contained wells containing 1/10 TSB media that acted as negative controls. The plates were enclosed in zip lock bag s and were incubated at 30 C. of 1% crystal violet (Acros Organics, Geel, Belgium) solubilized in 100% ethanol was added to each well of the polystyrene plates. Plates were incubated at ambient temperature for 15 min. The cry stal vio let solution was decanted from plates and plates were gently rinsed in DI water. The crystal violet was solubilized
90 in 120 (Fisher Scientific, Pittsburgh, PA). From each well, 100 of solubiliz ed crystal violet was transferred to a new 96 well polystyrene plate. Optical density of the samples was read at 595 nm in the microtiter plate reader, Victor 3 (Perkin Elmer, Fremont, CA) and the Wallac 1420 Manager work station software. The results were baselined to the negative control s amples. The study was performed with four technical replications for each of the three biological replications. The data was analyzed for significance in JMP 9.0 Pro through ANOVA and post hoc Tukey Kramer HSD analysis with a significance of p<0.05. Biof ilm F ormation of S almonella Typhimurium MKF 5, MKF1, MKF2, and WJZ2 on the Tomato S urface Overnight cultures of S. Typhimurium MKF5, MKF1, MKF2, WJZ2 MKF1 pMKF4, MKF1 pWSK29, MKF2 pMKF2, MKF2 pCR2.1, WJZ2 pMKF2 pMKF5, and WJZ2 pCR2.1 pBBR1 MSC5 were washed in sterile DI water three times. All cultures were diluted to ~10 6 CFU/ml in sterile DI water and were supplemented with the appropriate antibiotic(s) to reduce proliferation and biofilm formation of tomato microflora. Field grown, unwaxed, unprocessed green and red tomatoes cultivar Florida 47 or Quincy were wiped with ethanol and dried in th e biological hood for 10 min Small, pla stic containers that were washed with b leach and UV sterilized were each fil led wit h 70 ml of inoculum. T he blossom end s of field tomatoes were submerged in the inocul a with 70 90 mm of the blossom end of each tomato being submerged The tomatoes were incubated at 30 C for 24 hr. After 24 hr, the tomatoes were moved to a container of 2 ml of 1% crystal violet. The staining and measurement of biofilm was performed as previously described by staining the blossom ends of the tomato, rinsing in DI, and
91 solubilizing in 30 ml of 33% acetic acid. Negative controls were performed with each study. The study was performed in tripli cate, with three biological replications. For the biofilm formation of S Typhimurium MKF5, MKF1, MKF2, and WJZ2, significance was determined in JMP 9.0 Pro through pairwise t tests comparing each ts were completed with MKF5 as the control, one for the green tomato data and the other for the red tomato data. An effect test was also performed to determine the variable ( Salmonella strain and/ or tomato maturity) most significantly effecting the Salmon ella biofilm formation. For the biofilm formation of S. Typhimurium MKF1 pMKF4, MKF1 pWSK29, MKF2 pMKF2, MKF2 pCR2.1, WJZ2 pMKF2 pMKF5, and WJZ2 pCR2.1 pBBR1 MSC5 significance was determined in JMP 9.0 Pro. Pairwise t tests comparing the biofilm formati on of gen etic complements S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, and WJZ2 pMKF2 pMKF5, to the biofilm formation of MKF5 and the vector controls, S. Typhimurium MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 were compared to MKF1, MKF2, and WJZ2, re spectively. The green tomato and red tomato data were compared separately for the statistical analysis of the genetic complements and controls Biofilm F ormation of S almonella Typhimurium MKF 5, MKF1, MKF2, and WJZ2 on T omato Segments Overnight cultures o f S. Ty phimurium MKF5, MKF1, MKF2, WJZ2, MKF1 pMKF4, MKF1 pWSK29, MKF2 pMKF2, MKF2 pCR2.1, WJZ2 pMKF2 pMKF5, and WJZ2 pCR2.1 pBBR1 MSC5 were washed three times in sterile DI water All cultures were diluted to ~10 6 CFU/ml in either sterile DI water or ste rile DI water supplemented with MnCl 2 (5mg/L) and MgSO 4 (50mg/L) and supplemented with the appropriate
92 antibiotic(s) Two field grown, unwax ed, unprocessed tomatoes cultivar Florida 47 or Quincy one green and one red, were wiped with ethanol and dried in the biological hood for 5 min. Six 15 mm x 15 mm squares were cut from the surface of each tomato with a scalpel and tweezers that were flame sterilized. The tissue was removed from the 15 mm x 15 mm squares to result in a height of ~2 mm. The tomato s egments were added to each well of two 6 well polystyrene plate s (Costar, Washington, DC) to result in a plate of green tomato segments and a plate of red tomato segments. Diluted cult ures were pipetted into the wells of the polystyrene plate 4 ml per we ll Two wells were only inoculated with media to serve as negative controls. Plates were placed in zip lock bags and incubated for 24 hr at 30 C. After 24 hr, the tomato segments were removed from the wells and placed into the individual wells of new, sterile 6 six well polystyrene plates containing 1 ml of 1% crystal violet in each well. The tomato segments were stained and optical density readings were recorded as previously described. Statistical analysis was performed as described in the previous section with pairwise t tests tests, and an effect test. C haracterization of Salmonella Outbreak S trains Several produce associated Salmonella o utbreak strains were electroporated with plasmid pWSK29, as previously described, to add antibiotic resistance to the outbreak strains. The strains are detailed in Table 5 1. The strains were assessed for rdar morphology and cellulose production by streaking the Salmonella outbreak strains with and without the pWSK29 plasmid onto CR and CW agar plates CR agar was incubated at 30 C for two days, while CW agar plates were incubated without light exposure at room temperature for two days. After 48 hr, pictures were taken of the CR
93 plates and the cultures on the CW agar were suspended in sterile deionize d (DI) water and samples were loaded into a 96 well polystyrene plate along with a control of DI water from an uninoculated CW agar plate. The optical density and fluorescence readings of the cultures were measured and recorded as performed in Da Re and G higo ( 23 ) Growth curves were also pe rformed over 12 hr at 30 C, as previously described. Pairwise t tests with 5% significance were performed to determine if the plasmid insert altered cellulose production. Of the strains that were not significantly different in their ce llulose production, an ANOVA and pairwise t tests were performed with a significance of p<0.05 After analysis of cellulose production and growth curves, the strains were utilized for biofilm assays. The biofilm formation of each strain containing the pWSK29 plasmid was ass essed on tomato segments of green, Florida 47 tomatoes, as previously described The biofilm data was analyzed in JMP 9.0 Pro through an ANOVA with significance at p<0.05. Results The biofilm studies were conducted at 30 C rather than 37 C due to the enha nced transcription of agfD at 25 30 C and to more effectively simulate the environmental conditio ns as opposed the temperatures within an animal host As displayed in Figure 5 1 the biofilm formation of S. Typhimurium 14028 MKF5, JS246, as well as MKF1 pMKF4 were indicated as significantly greater in the Tukey Kramer HSD analysis This indicates that the biofilm formation of both MKF5 and JS246 were comparable to the wild type 14028 culture, making either suitable for use in place of 14028 in future bio film studies An antibiotic resista nt wild type surrogate culture wa s necessary for biofilm studies on tomatoes to allow for control of the natural to mato microflora during the studies S. Typhimurium MKF5 was chosen for use in the
94 remaining biofilm stud ies in place of 1 4028 because the strain was resistant to kanamycin, which was the same resistance carried by MKF1, MKF2, and WJZ2. S. Typhimurium MKF1, MKF2, and WJZ2 demonstrated significantly reduced biofilm expressing up to a 3 fold loss of optical d ensity in comparison to MKF5 The biofilm f ormation of MKF1 pMKF4 indicated a f ull restoration of biofilm formation capabilities with the genetic complementation of agfD on the extrachromo somal pWSK29 plasmid. The biofilm formation of MKF2 pMKF2 and WJZ2 pMKF2 pMKF5 indicate d partial restoration of biofilm formation capabilities with a 2 fold increase in OD in comparison to MKF2 and WJZ2, respectively The empty vector controls, S. Typhimurium MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 demonstr ate d similar biofilm formation as the MKF1, MKF2, and WJZ2 strains, respectively. The assessment of biofilm formation of S. Typhimurium MKF5, MKF1, MKF2, WJZ2, MKF1 pMKF4, MKF1 pWSK29, MKF2 pMKF2, MKF2 pCR2.1, WJZ2 pMKF2 pMKF5, and WJZ2 pCR2.1 pBBR1 MSC5 o n intact tomatoes and tomato segments were completed on green and red field tomatoes in DI water. The resulting formation of biofilm on intact tomatoe s is illu strated in Figure s 5 2 and 5 3 There we re no significant differences in the biofilm formation of each Salmonella strain on green versus red intact tomatoes or tomato segments, as determined through pairwise t tests with a significance of p<0.05, the biofilm formati on of S. Typhimurium MKF1, MKF2, and WJZ2 on both green and red intact tomatoes were significantly less than the biofilm formed by MKF5, with a 2 fold reduction in OD An effect test with significance of 5% indicated that the significant effector of the b iofilm formation was the Salmonella strain, as
95 opposed to the tomato maturity in the intact tomato studies The comparison of S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, and WJZ2 pMKF2 pMKF5 to the biofilm formation of MKF5 and the comparison of S. Typhimurium MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 to the respective deletion mutants on either green or red intact tomatoes resulted in no significant differences Figure s 5 4 and 5 5 illustrate the biofilm formation of Salmonella on green and red toma to segments suspended in DI water and the appropriate antibiotic. The biofilm formation of S. Typhimurium MKF5 wa s larger than the MKF1, MKF2, and WJZ2 deletion mutants, but the difference in biofilm formation wa s not significant on green tomatoes accordi iofilm formation on red tomato segments S. T yphimurium MKF2 and WJZ2 displayed significantly decreased biofilm in comparison to MKF5 with optical density levels measured below the negative controls The pairwise t test s for the strains on green or red tomato segments resulted in a significant difference for biofilm formation of S. Typhimurium MKF1 pMKF4 on green versus red tomatoes, with greater biofilm formation on green tomato segments. There were no significant diff erences in the biofilm formation of S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, and WJZ2 pMKF2 pMKF5 compared to MKF5 or S. Typhimurium MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 compared to MKF1, MKF2, and WJZ2, respectively, on either green or red t omato segments. An effect test indicated that the Salmonella strain was a significant factor in biofilm formation on tomato segments, while the tomato maturity was found to be insignificant. Analysis of Salmonella biofilm on green and red tomato segment s was also performed in DI water supplemented with MnCl 2 (5 mg/L) and MgSO 4 (50 mg/L) and the
96 appropriate antibiotic (s) to determine if the presence of the compounds altered biofilm formation ( Figure 5 6) Pairwise t tests were performed to determine that there was no significant difference in the biofilm formation of the Salmonella strai ns i n the presence or absence of MnCl 2 and MgSO 4 on either green or red tomato segment s with the exception of S. Typhimurium MKF1 pMKF4 on green tomatoes An effect test also indicated that in considering the factors of Salmonella strain, tomato maturity, and presence or absence of MnCl 2 and MgSO 4 the only significant factor effecting the formation of biofilm on tomatoes was the Salmonella strain. The biofilm studies sh owed that the non rdar S. Typhimurium MKF1, MKF2, and WJZ2 were deficient in biofilm formation on the tomato surface. The deficiency of targeted deletion mutants in the formation of biofilm does not establish there is a correlation between the rdar morpho logy and biof ilm formation. Several produce outbreak Salmonella strains of clinical and environmental origin were also used to study the correlation between the rdar morphology, biofilm formation, and environmental persistence (Table 5 1). The outbreak s trains were assessed for the rdar or non rdar morphology on CR agar along with three lab derived controls, S. Typhimurium 14028, the wild type strain, S. Typhimurium JTN176, an rdar like strain, and S. Typhimurium JTN212, a spontaneous rpoS mutant with saw (smooth and white) morphology The S. Typhimurium JTN176 and JTN212 strains were strains of 14028 passed through tomatoes three times (Table 3 2). The three controls, along with the outbreak strains underwent electroporations with pWSK29 to add extrachr omosomal antibiotic resistance to the strains for use in the biofilm studies
97 The strains were characterized for morphology on CR media growth curves were performed, and cellulose was quan tified through fluorescence on CW agar ( Figure s 5 7, 5 8, and 5 9 ) The S. Typhimurium 14028 pWSK29 and JTN1 76 pWSK29 strains demonstrated rdar or rdar like morphologies with the JTN212 pWSK29 demonstrating the saw morphology. The outbreak strains all expressed non rdar morphologies that were a slight pdar (pink, dry and rough) with irregular or curled edging ( Figure 5 7). The growth curves in Figure 5 8 showed no irregular or deficient growth patterns for the Salmonella strains. The cellulose quantification resulted in significant differences between the wild typ e and pWSK29 containing strains for S. Typhimurium JTN176, S. Braenderup 04E01347, and S. Michigan LJH521. The three strains were removed from the study. The cellulose quantifications for the remaining Salmonella strains were analyzed to determine the on ly strains that did not significant ly differ in cellulose production were S. Newport C4.2 and C6.3 ( Figure 5 9). The cellulose was quantified for each strain to determine if biofilm formation could be linked to cellulose production. The biofilm formation s of the strains were only assessed on green tomato segments. The ANOVA analysis of the biofilm data indicated there is no significant difference between the outbreak strains or the S. Typhimurium 14028 and JTN212 controls. Discussion The analysis of the biofilm formation of S. Typhimurium MKF5, MKF1, MKF2, and WJZ2 estab lished that the deletion of the AgfD regulator, AgfD regulated thin aggregative fimbriae, and/or the cellulose genes did significantly reduce biofi lm formation on polystyrene and intact t omato surface s of both gree n and red tomatoes ( Figure s 5 1 and 5 2). The TSB media utilized in the biofilm assays was diluted 1:10 because nutrient limitation has been shown to promote biofilm formation and
98 transcription of agfD ( 78 ) On the polystyrene surface, the deletion mutants had an approximate 3 fold reduction in biofilm formation (as measured by optical density ) ( Figure 5 1). While on the intact tomat o surface, the loss of biofilm formation was approximately 2 fold for each of the three delet ion mutants in comparison to S. Typhimurium MKF5 ( Figure 5 2). The formation of biofilm by S. Typhimurium MKF5 and reduced production from MKF1, MKF2, and WJZ2 on both polystyre ne and tomato surfaces indicated that t he AgfD regulator and cellulose have important roles in surface independent biofilm formation of Salmonella. The lack of significant difference between S. Typhimurium MKF2 and WJZ2 indicate that the th in aggregative fimbriae do not have a major role under these experimental conditions, as was observed in the previous initial attachment and persistence studies. The previously described persistence studies determined that the lack of the AgfD regulator did not decrease the persistence of Salmonella on the tomato surface, but the deletion of agfD alters the biofilm forming capacity of the strain on polystyrene and the intact tomato surface There are AgfD independent routes of cellulose production and nu merous fimbriae operons of Salmonella, which were proposed to contribute to the null effect of the deletion of agfD in the persistence studies. The deletion of agfD is hypothesized to result in the significa nt loss of biofilm formation because agfD is reg ulated by environmental cues of nutrient scarcity, oxygen tension, temperature, pH, osmolarity, and ethanol presence through numerous global regulators including RpoS, OmpR, MlrA, CxpR, H NS, and IHF ( 11 47 100 ) It has been proposed that the environmentally based regulation of agfD allows the regulator to trig ger the transition between planktonic behavior and the sessile community structure
99 of biofilm formation in response to changing environmental conditio n s ( 31 ) The Salmonella could not effectively form biofilm due to the loss of regulation by AgfD. The biofilm formation of Salmonella on tomato segm ents differed from the biofilm formed on the tomato surface. On the tomato segments, the Salmonella had contact with the tomato tissue and any exudates remaining on the exposed tissue and surface As previously discussed, the plant surface has minimal nu trients available and is a location where Salmonella may persist but do not typically thrive, while studies have demonstrated a 5 fold increase in Salmonella population s within red tomatoes in less than 1 2 days ( 56 ) The presence of the exposed tomato tissue may have provided nutrients to the bacterium, therefore, reducing nutrient limitation and biofilm formation S. Typhimurium MKF5 exhibited decreased biofilm formation on green tomato segments in DI water in comparison to the biofilm formation on the green intact tomato or the red tomato segment, resulting in a lack of significant differences between the biofilm formation of S. Typhimurium MKF5 and MKF1, MKF2, and WJZ2 ( Fig ure 5 4). The biofilm formation of S. Typhimurium MKF 5 on red tomato segments exceeded that on the intact tomato surface, but the biofilm formation of MKF1, MKF2, and WJZ2 we re minimal on the red tomato compared to the green tomato segments Biofilm stud ies in 96 well polystyrene microtiter plates were performed using rinsates from the tomato segments to establish if there were antagonistic or synergistic exudates causing the difference in biofilm formation of S. Typhimurium MKF5, but the biofilm in the r insates did not differ (data not shown). The data indicates that while biofilm formation on tomato surfaces and segments may be variable, there is strong biofilm formation on the intact tomato surface of bo th green and red tomatoes.
100 The biofilm assays o n the tomato segment were also performed with a suspension solution of DI water supplemented with MnCl 2 and MgSO 4 to determine if the compounds would alter the development of biofilms. The only significant difference was between the biofilm formation of S Typhimurium MKF1 pMKF4 on green tomato segments in DI versus the supplemented DI ( Figure 5 6) The biofilm formation of S. Typhimurium MKF1 pMKF4 was also the only strain to significantly differ on green versus red segments, so the significance is likel y attributed to the high optical density measured on the green tomato segments in DI water rather than an effect from the MnCl 2 and MgSO 4. The effect tests of the tomato surface, tomato segment, and variable suspension solution studies all only indicated that the Salmonella strain was significantly directing biofilm formation, indicating that the characteristics of the bacterial strain may be more important than the ripeness or presence of plant microflora in the environmental persistence of Salmonella in the produce environment. In all three biofilm studies on tomatoes, the genetic complements including S. Typhimurium MKF1 pMKF4, MKF2 pMKF2, WJZ2 pMKF2 pMKF5 demonstrated a restored phenotype to the biofilm capabilities of MKF5 with no significant differe nces between the complements and MKF5 for any of the assays. The biofilm formation of the vector controls, S. Typhimurium MKF1 pWSK29, MKF2 pCR2.1, and WJZ2 pCR2.1 pBBR1 MSC5 also did not significantly differ from their respective mutants, MKF1, MKF2, an d WJZ2 ( Figure s 5 3, 5 5) The results for the genetic complements and controls indicated that the decrease in biofilm formation of MKF1, MKF2, and WJZ2 was due to the deficiency of the AgfD regulator and AgfD regulated thin aggregative fimbriae, and/or c ellulose in the strains
101 The characterization and assess ment of biofilm formation of Salmonella outbreak strains was done to assess the correlation between the rdar morphology biofilm formation, and fitness in the plant environment. The cellulose for t he wild type and pWSK29 containing strains were compared and the strains significantly differing in cellulose production were removed from the study since the insertion of pWSK29 altered a phenotype that contributes to biofilm formation. The outbreak stra ins assessed for biofilm formation were linked to t omato outbreaks of either clinical or environmental origin (Table 5 1). The S. Typhimurium 14028 pWSK29 was the control for the rdar morphology, with JTN212 pWSK29 as a saw control. The Salmonella outbre ak strains exhibit environmental survival, so the characterization of the strains as non rdar and having decreased cel lulose production indicates that the rdar morphology does not directly correlate with environmental survival ( Figure s 5 5 and 5 9 ). The maximum b iofilm formation on the tomato segments was produced by S. Typhimurium 14028 pWSK29, the rdar control, but the biofilm formation was not significantly different from the saw control, S. Typhimurium JTN212 pWSK29 or the remaining outbreak strains ( Figure 5 10) The resulting studies with the outbreak strains i ndicate d that rdar morphology and biofilm format ion may correlate but not significantly so. It is difficult to correlate the rdar morphology with a specific phenotype, such as biofilm or ce llulose production. This lack of correlation may be due to the unknown mutation causing the loss of the rdar morphology. The loss of rdar can be an evolutionary loss, which is often seen in laboratories where strains are continuously passed through nutri ent rich media The loss of the rdar morphology may be due to a mutation in the cellulose biosynthesis pa thway, but it is more likely to be derived from a
102 mutation of one of the upstream regulators of agfD, which may alter cellulose production along with other crucial factors of biofilm formation. The complexity of the rdar morphology and the survival of Salmonella outside of the animal host results in the conclusion that t here is no in dication that rdar morphologies, biofilm formation, or cellulose produ ction can singularly dictate the environment fitness of Salmonella in the produce environment.
103 Table 5 1. Salmonella enterica cultures obtained from produce associated outbreaks S erovar and strain Outbreak event and charac teristics Source or reference Braenderup 04E00783 Isolated from 19 yr old male, stool sample, early in Roma Tomato outbreak ( 34 ) Braenderup 04E 01347 Isolated from 24 yr old female, stool sample, sample obtained in middle of Roma Tomato outbreak ( 34 ) Braenderup 04E01556 Isolated from 5 y r old male, stool sample, sample obtained late in Roma Tomato outbreak ( 34 ) Michigan LJH521 Spontaneous rifamycin resistant derivative of the st rain isolated from cantaloupe associated Salmonella outbreak L. J. Harris, University of California Davis Montevideo LJH519 Spontaneous rifamycin resistant derivative of the strain isolated from patient with salmonellosis in tomato outbreak M. D. Danylu k, University of Florida Newport C4.2 Isolated from tomato field in Delmarva area during round tomato associated outbreak, identical PFGE BlnI and XbaI pattern as C6.3 strain ( 33 ) Newport C6.3 Isolated from tomato field in Delmarva area during round tomato associated outbreak, identical PFGE BlnI and XbaI pattern as C4.2 strain ( 33 )
104 Figure 5 1. Biofilm formation of Salmonella on polystyrene The Salmonella st rains were diluted 1:100 in 1/10 trypt ic soy broth (TSB) and were incubated in polystyrene plates for 24 hr at 30 C. The plates were stained with crystal vi olet, which was solubilized in 33% acetic acid. The optical density of the samples was measured at 595 nm. The error bars denote the standard deviation. The statistical grouping of the data as determined by Tukey Kramer HSD analysis with significance se t to p<0.05 is denoted by the letters, with a significance difference in samples of different letters. 0.0 0.4 0.8 1.2 1.6 2.0 Absorbance Salmonella strain A A A B B B A B AB B AB B
105 0.0 0.4 0.8 1.2 MKF5 MKF1 MKF2 WJZ2 Absorbance Salmonella strain Green Tomato Red Tomato Figure 5 2. Biofilm formation of Salmonella Typhimurium MKF5, MKF1, MKF2, and WJZ2 on the surface of intact green and red to matoes The blossom end of field grown green and red, round tomatoes were submerged in Salmonella inocu la suspended in DI water for 24 hr at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 595 nm. The err or bars denote standard error. The stars denote significant difference in biofilm formation of the strains in comparison to MKF5 on both green or red
106 Figure 5 3 Biofilm formation of Salmonella Typhimurium complement and vector control strains on the surface of intact green and red tomatoes The blossom end of field grown green and red, round tomatoes were submerged in Salmonella inocul a in DI water for 24 h r at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 595 nm. The error bars denote standard error. 0.0 0.4 0.8 1.2 Absorbance Salmonella strain Green Tomato Red Tomato
107 -0.8 -0.4 0.0 0.4 0.8 1.2 1.6 2.0 MKF5 MKF1 MKF2 WJZ2 Absorbance Salmonella strain Green Tomato Red Tomato Figure 5 4 Biofilm formation of S almonella Typhimurium M KF5, MKF1, MK F2, and WJZ2 on green and red tomato segments. Tomato segments measuring 15 mm x 15 mm x 2 mm were submerged in Salmonella inocul a of DI water for 24 hr at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 5 95 nm. The error bars denote standard error. The stars denote significant difference in biofilm formation of the strains in comparison to MKF5 on both green or red tomatoes as indicated by a
108 0.0 0.4 0.8 1.2 1.6 2.0 Absorbance Salmonella strain Green Tomato Red Tomato Figure 5 5 Biofilm formation of Salmonella complement and vector control strains on green and red tomato segments. Tomato segments measuring 15 mm x 15 mm x 2 mm were submerged in Salmonella inocul a of DI water for 24 hr at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 595 nm. The error bars denote standard error.
109 -0.4 0.0 0.4 0.8 1.2 1.6 Absorbance Salmonella strain -0.4 0.0 0.4 0.8 1.2 1.6 Absorbance Salmonella strain Figure 5 6. Biofilm formation of Salmonella strains on green and red tomato segments in DI water supplemented with MnCl 2 and MgSO 4 A) Biofilm formation of Salmonella on green tomato segments. B) Biofilm formation of Salmonella on red tomato segments. Tomato segments measuring 15 mm x 15 mm x 2 mm were sub merged in Salmonella inocul a of DI water for 24 hr at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 595 nm. The error bars denote standard error.
110 Figure 5 7 Depictions of Salmonella outbreak s trains on Congo r ed media. A) S. Typhimurium 14028 pWSK29, B) JTN176 pWSK29, C) JTN212 pWSK29, D) Braenderup 0783 pWSK29, E) Braenderup 1347 pWSK29, F) Braenderup 1556 pWSK29, G) Michigan pWSK29, H) Montevideo pWSK29, I) Newport C4.2 pWSK29 J) Newport C6 .3 pWSK29 Salmonella cultures spotted onto Congo r ed media and incubated at 30C for 48 hr. An Olympus MVX10 confocal microscope and Olympus MVX TV1XC camera was used to photograph the strains. The Salmonella 14028 pWSK29 and JTN176 pWSK29 are rdar con trols, while JTN212 pWSK29 is a saw control. The outb reak strains all appear to be varying non rdar morphology.
111 Figure 5 8. Growth curves of produce related Salmonella outbreak strains for 12 hr at 30 C. Overnight cultures of each strain were dilute d to the approximate concentration of 10 4 CFU/ml and incubated at 30C, shaking at 200 rpm for up to 12 hr. Samples were diluted and plated once per hour, including hour 0. The final growth curves were the average of three replicates per strain with erro r bars indicating standard deviation. 0 2 4 6 8 10 0 1 2 3 4 5 6 7 8 9 10 11 12 Average Log CFU/ml Time (Hours) 14028 pWSK29 JTN212 pWSK29 JTN176 pWSK29 Braenderup 1374 pWSK29 Braenderup 0783 pWSK29 Braenderup 1556 pWSK29 Michigan pWSK29 Montevideo pWSK29 Newport C4.2 pWSK29 Newport C6.3 pWSK29
112 0 100 200 300 400 500 600 Relative Fluorescence Units Salmonella strain A F C B E D D Figu re 5 9. C ellulose quantification of Salmonella outbreak strains. The graph represents quantified relative fluorescence units of cellulose produced by each strain on the left axis. The st reaked colonies for each of the strains were resuspended on the CW agar in DI water and the fluorescence was recorded in 96 well plates at an excitation of 360 40 nm and emission of 460 40 nm. Optical density of the cultures was measured a t 600 nm. T he relative fluorescence is expressed as fluorescence per unit of OD 600 standard deviation from three replicate studies. The statistical grouping of the data as determined by t test analysis with significance set to p<0.05 is denoted by the letters, wit h a significance difference in samples of different letters.
113 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Absorbance Salmonella strain Figure 5 10. Biofilm formation of Salmonella outbreak strains on green tomato segments. Tomato segments measuring 15 mm x 15 mm x 2 mm were submerged in Salmo nella inocul a of DI water for 24 hr at 30 C. The biofilm was quantified through crystal violet staining and optical density readings at 595 nm. The error bars denote standard error.
114 CHAPTER 6 CONCLUSIONS AND FUTURE STUDIES In this work, the AgfD re gulator, AgfD regulated thin aggregative fimbriae and /or AgfD regulated cellulose were investigated for their roles in the interactions of Salmonella and the tomato surface The research concluded that AgfD regulated thin aggregative fimbriae were not fo und to have a significant role in initial attachment, persistence, or biofilm formation on the tomato surface under the se experimen tal conditions. Future research c ould be directed towards the role of thin aggregative fimbriae in the interactions between Salmonella and produce surfaces as there have been reports of thin aggregative fimbriae mutants (both targeted deletion mutant s and spontaneous mutants) deficient in attachment and biofi lm. A continuation of this research could assess the role and express ion of the fimbrial operons of Salmonella during the attachment to abiotic and biotic surfaces to determine which specific fimbriae are active in attachment to produce. Along with thin aggregative fimbriae, the AgfD regulator and AgfD regulated cellulose were also found to be insignificant in the initial attachment to the tomato surface, indicating that initial attachment may be directed more b y physiochemical properties than the biological factors studied in this research In the persistence of Salmonell a on the tomato surface, AgfD regulated cellulose was determined to be a significant factor in persistence over 6 days Understanding the components necessary for the environmental survival and persistence of Salmonella ma y provide targets for preharvest treatments to reduce contamination of produce a nd while the application of cellulase may reduce persistence of Salmonella o n plant surface, the cellulose may also damage the plants since cellulose is a component of cell walls in green plants. Despite
115 the lack of an effective target to reduce Salmonella persistence on plants, the research assessing the necessary components for human pathogen survival and persistence in the environment should be continued for future development of targeted preharvest prevent ion strategies. The establishment of Salmonella biofilms on the surface of both green and red tomatoes is also important in the food industry. A future direction of this work could be the investigation of current sanitizer concentrations against Salmone lla biofilms on the tomato surface and in simulated flume systems to establish if current standards are effective in removing the pathogen. The biofilm studies were performed by submerging the materials in solutions of Salmonella so these conditions are unlikely to occur in the field environment but it is possible for Salmonella to form biofilm on the surface of intact tomatoes without additional nutrient sources in the field during a flooding event The formation of biofilms should also be considered in the tomato packing facilities where flume systems are used The potential for biofilm formation may support the use of brush roller system rather than flumes The AgfD regulat or and AgfD regulated cellulose were found to be significant factors in the fo rmation of biofilm on the surface of polystyrene and intact tomatoes. The AgfD regulator is active under a variety of stimuli including temperatures of 25 30 C, low nutrient conditions, and in the presence of ethanol. Packing facilities should deter the use conditions in the flume systems that are favorable for transcription of agfD to reduce AgfD directed biofilm formation. Cellulase could also be used along with chemical sanitizers in the sanitation process of the flumes and other food contact surfaces to reduce the persistence of any viable Salmonella before biofilms are formed.
116 The persistence of Salmonella on the tomato surface was also found to be more favorable on red versus green tomatoes for the cellulose deficient strains. The 2 fold reduction in the persistence on green tomatoes indicates that both bacterial behavior, including cellulose production, and produce ripeness could be factors in the persistence of S. Typhimurium on the tomato surface. T his should be considered in determining the bes t time to harvest the produce. The taste of a product is important to consumers, and higher demand is being seen for fresher produce that are harvested closer to the peak maturity and ripeness. Additional research would be needed to determine if the pers istence of Salmonella is favorable as the tomatoes ripens, because if so, harvesting pink tomatoes to meet consumer demand may decrease the safety of the product. Salmonella has the ability to persist and form biofilm on the tomato surface under the condi tions of this research. The work should be applied to greenhouse and field studies to determine the persistence and potential biofilm formation in these environments The determination of environmental fitness through screening methods such as rdar or no n rdar classification could also be used to characterize Salmonella isolated from the environment. W hile this was attempted, the rdar morphology and environmental persistence did not correlate. The current direction of food safety is one of prevention. Understanding the biological factors necessary for Salmonella survival in the plant environment may provide targets for specific preharvest treatments in addition to the general safety strategies set forth by GAPs and FSMA.
117 APPENDIX A MEDIA COMPOSITION B uffered Peptone Water (0.1%) 1 .0 g peptone 0.5 g NaCl 0 35 g Na 2 HPO 4 0.15 g NaH 2 PO 4 1 .0 L DI H 2 O Autoclave sterilize Calcof luor White (CW) Agar 9.6 g tryptone 4.8 g yeast extract 960 ml DI H 2 O Autoclave sterilize Ad d: 40 ml of Calcofluor white with Evans B lue dye Colonization Factor Antigen (CFA) Medium 10 g casamino acids 1.5 g yeast extract 50 mg MgSO 4 5 .0 mg MnCl 2 1 .0 L DI H 2 O Adjust to pH 7.4 Autoclave sterilize Congo Red (CR) Agar 10 g tryptone 5 .0 g yeast extract 1 .0 L DI H 2 O Autoclave sterilize Add: 4 .0 ml of 100 mg/ml Congo r ed in methanol 2 .0 ml of 100 mg/ml Brilliant Blue in ethanol Evans Blue Uranine (EBU) Agar 10 g tryptone 5 .0 g yeast extract 5 .0 g NaCl 2.5 g glucose 15 g agar 960 mL DI H 2 O Autoc lave sterilize Add: 28.71 mL of 1M K 2 HPO 4 1.25 mL of 1% Evans Blue in DI H 2 O 2.5 mL of 1% Uranine in DI H 2 O Light sensitive, store properly
118 Luria Be rtani (LB) Lennox Agar 10 g tryptone 5 .0 g yeast extract 5 .0 g NaCl 15 g agar 1 .0 L DI H 2 O Autoclave sterilize Luria Be rtani (LB) Lennox Broth 10 g tryptone 5 .0 g yeast extract 5 .0 g NaCl 1 .0 L DI H 2 O Autoclave sterilize NZY+ Per 100 ml DI H 2 O 1 .0 g NZ amine 0.5 g yeast extract 0.5 g NaCl Autoclave sterilize Add: 1.25 mL of 1 M MgCl 2 2 .0 mL 20% glucose in DI H 2 O Potassium Phosphate B uffer (0.1M) 61.5 ml of 1 M K 2 HPO 4 in DI H 2 O 38.5 ml of 1 M KH 2 PO 4 in DI H 2 O Adjust pH to 7.0 Dilute 1:10 in DI H 2 O Autoclave sterilize PUM Buffer ( 65 ) 22.2 g K 2 HPO 4 H 2 O 7.26 g KH 2 PO 4 1.8 g urea 0.2 g MgSO 4 H 2 O 1 .0 L DI H 2 O Adjust to pH 7.1 Autoclave sterilize Sodium Phosphate Buffer (0.002 M) 46.3 ml of 1 M Na 2 HPO 4 in DI H 2 O 53.7 ml of 1 M NaH 2 PO 4 in DI H 2 O Adjust pH t o 6.8 Dilute 0.4 ml in 199.6 ml DI H 2 O Autoclave sterilize
119 Trypt ic Soy Broth (TSB) 1.7 g tryptone 0.3 g soy peptone 0.25 g dextrose 0.5 g NaCl 0.25 g Na 2 HPO 4 1 .0 L DI H 2 O Autoclave sterilize Xylose Lysine Desoxycholate (XLD) Agar 3 .0 g xylose 5 .0 g L ly sine 7.5 g lactose 7.5 g sucrose 5 .0 g NaCl 3 .0 g yeast extract 0.08 g phenol red 2.5 g sodium des oxycholate 6.8 g sodium thiosulfate 0.8 g ferric ammonium citrate 13.5 g agar 1 .0 L DI H 2 O Heat media until reaching boiling, do not autoclave
120 APPENDIX B AT TACHMENT AND BIOFILM FORMATION OF Salmonella STRAINS AFTER INCUBATION UNDER STATIC OR SHAKING CONDITIONS Before studying the attachment, persistence, and biofilm formation of Salmonella the incubation conditions for growing the Salmonella strains were as sessed. The cultures were grown at 30 C to encourage activation of agfD and the AgfD regulated system ( 64 ) Preliminary attachment and biofilm studies were performed with S. Typhimurium 14028, MKF2, and WJZ2 to determine if incubating the strains while shaking at 200 rpm would disrupt the extracellular components of the bacteria and alter the results in the tomato studies. The MKF1 strain was not included in the study because the scope of the assay was on the effect of incubation conditions on the condition of the extracellular components, so the culture with the hi ghest and lowest expression of thin aggregative fimbriae and cellulose were compared. The 14028, MKF2, WJZ2 strains were grown in parallel at 30 C either static or shaking at 200 rpm. After overnight incubation, the strains grown under shaking conditions were diluted to the measured OD 600 (optical density at 600 nm) of the static cultures. The strains were washed three times in PBS. The strains were then used to assess attachment to tomatoes and biofilm formation. For the attachment assay, t he blossom e nd of Campari, vine ripe tomatoes (obtained from Publix Grocery, Gainesville, FL) were spot inoculated with 10 spots of 10 of inoculum triplicate, with two biological replications. After inoculation, the tomatoes rested on the bench for five minutes before being rinsed in 10 ml of sterile DI water. The rinsate was diluted and spread plated onto XLD and the tomatoes were stored at 22C, 60 80%
121 relative humidity for 4 hr The blossom end s of the tomatoes were then rinsed in 10 ml of sterile DI water. The tomatoes were rubbed by hand for 60 sec in a stomacher bag containing 100 ml of PBS. The secondary rinsates and tomato rinse from the stomacher bags were diluted and plated onto XLD. Pl ates were incubated at 37C overnight to obtain plate counts of the recovered Salmonella concentrations ( Figure B 1). The recovered concentrations of 14028, MKF2, and WJZ2 in the initial rinse, secondary rinse, and tomato attachment do not differ. For the biofilm assay, the washed strains were diluted 1:100 in colonization factor antigen (CFA) medium and loaded into 96 well polystyrene plates (Appendix A) The plates were incubated at 30C for 24 hr. After incubation, the plates were treated with crystal violet, rinsed with DI water, and the crystal violet dye was solubi lized in 33% acetic acid for optical density measurement at 595 nm The results of the biofil m assay are illustrated in F igure B 2 T test analysis for each culture pair was performed wi th JMP 9.0 Pro statistical software to show there was no significant difference (with p<0.05) in the attachment or biofilm formation of S. Typhimurium 14028, MKF2 or WJZ 2 when grown at 30C in static or shaking conditions.
122 Figure B 1 Attachme nt of Salmonella on the surface of Campari tomatoes after static or shaking overnight growth. The assay was performed by incubating the S. Typhimurium 14028, MKF2, and WJZ2 strains overnight in duplicate at 30 C, either static or shaking at 200 rpm The cultures were recovered in an initial rinse after 5 min a secondary rinse after 4 hr, and a final recovery through physical abrasion of the tomato surface by rubbing. The three graphs depict the attachment results based on the resulting plate counts, with the growth conditions and strains labeled within the legends of each graph. 0 1 2 3 4 5 6 7 8 9 Initial Rinse Secondary Rinse Tomato Rinse Log CFU/ml Recovered 14028 Static 14028 Shaking 0 1 2 3 4 5 6 7 8 9 Initial Rinse Secondary Rinse Tomato Rinse Log CFU/ml Recovered MKF2 Static MKF2 Shaking 0 1 2 3 4 5 6 7 8 9 Initial Rinse Secondary Rinse Tomato Rinse Log CFU/ ml Recovered WJZ2 Static WJZ2 Shaking
123 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 14028 MKF2 WJZ2 14028 MKF2 WJZ2 Static Shaking Absorbance Growth Conditions and Salmonella Strain Figure B 2 Salmonella biofilm formation in polystyrene plates after static or shaking overnight growth. The assay was performed by incubating the S. Typhimur ium 14028, MKF2, and WJZ2 strains overnight in duplicate at 30 C, either static or shaking at 200 rpm. The cultures grown under shaking conditions were diluted to maintain similar starting concentrations, and the cultures were inoculated into polystyrene plates in parallel After overnight incubation at 30 C, the plates were stained and the optical density of the solubilized solution was recorded at an OD of 595 nm. The biofilm readings for each Salmonella strain were baselined to uninoculated samples wi thin each biofilm plate to account for any crystal violet binding to the polystyrene.
124 LIST OF REFERENCES 1. Andrews Polymenis, H., C. Santiviago, and M. McClelland. 2009. Novel genetic tools for studying food borne Salmonella. Curr. O pin. Biotechn 20:149 157. 2. Austin, J. W., G. Sanders, W. W. Kay, and S. K. Collinson. 1998. Thin aggregative fimbriae enhance Salmonella enteritidis biofilm formation FEMS Microbiol. Lett 162:295 301. 3. Barak, J., A. Liang, and K. Narm. 2008. Differe ntial attachment to and subsequent contamination of agricultural crops by Salmonella enterica. Appl. Environ. Microbiol 74:5568 5570. 4. Barak, J. D., L. Gorski, P. Naraghi Arani, and A. O. Charkowski. 2005. Salmonella enterica virulence genes are require d for baterial attachment to plant tissue Appl. Environ. Microbiol 71:5685 5691. 5. Barak, J. D., C. E. Jahn, D. L. Gibson, and A. O. Charkowski. 2007. The role of cellulose and O antigen capsule in the colonization of plants by Salmonella enterica. Mol. Plant Microbe Interact 20:1083 1091. 6. Baudart, J., J. Grabulos, J. P. Barusseau, and P. Lebaron. 2000. Salmonella spp. and fecal coliform loads in coastal waters from a point vs. nonpoint source of pollution J. Environ. Qual 29:241 250. 7. Berg, G., L. Eberl, and A. Hartmann. 2005. The rhizosphere as a reservoir for opportunistic human pathogenic bacteria Environ. Microbiol. 7:1673 1685. 8. Bhagwat, A. 2006. Microbiologica l safety of fresh cut produce. Where are we now? p. 121 165. In K. Matthews (ed .), Microbiology of Fresh Produce ASM Press, Washington, DC. 9. Brandl, M. 2004. Fitness of Salmonella enterica in the phyllosphere. Phytopathology 94:128. 10. Brandl, M., M. Carter, C. Parker, M. Chapman, S. Huynh, and Y. Zhou. 2011. Salmonella biofilm f ormation on Aspergillus niger involves cellulose chitin interactions PLoS ONE 6:e25553. 11. Brown, P., C. Dozois, C. Nickerson, A. Zuppardo, J. Terlonge, and R. Certiss. 2001. MrlA, a novel regulator of curli (AgF) and extracellular matrix synthesis by E schericha coli and Salmonella enterica serovar Typhimurium Mol. Microbiol 41:349 363. 12. C enters for Disease Control and Prevention 1991. Multi state outbreak of Salmonella Poona United States and Canada MMWR 40:549 552.
125 13. Centers for Disease Contr ol and Prevention 1999. Outbreak of Salmonella serotype Muenchen infections assoicated with unpasteurized orange juice United States and Canada MMWR 48. 14. Centers for Disease Control and Prevention 2002. Multistate outbreaks of Salmonella serotype Po ona infections associated with eating cantaloupe from Mexico United States and Canada, 2000 2002 MMWR 51:1044 1047. 15. Centers for Disease Control and Prevention 2007. Multistate outbreaks of Salmonella infections associated with raw tomatoes eaten in r estaurants United States, 2005 2006 MMWR 56:909 911. 16. Centers for Disease Control and Prevention 2013 Foodborne disease outbreak surveillance Available at: http://www.cdc.gov/outbreaknet/surveillance_data.html Accessed 3 April 2013. 17. Centers fo r Disease Control and Prevention 2008. Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce items United States, 2008 MMWR 57:929 934. 18. Centers for Disease Control and Prevention 2011. Incidence and trends of inf ection with pathogens transmitted commonly through food foodborne diseases active surveillance network, 10 U.S. sites, 1996 2010. MMWR 60:749 755. 19. Chaidez, C., M. Moreno, W. Rubio, M. Angulo, and B. Valdez. 2003. Comparison of the disinfection effic acy of chlorine based products for inactivation of viral indicators and pathogenic bacteria in produce wash water Int. J. Environ. Health Res 13:295 302. 20. Clegg, S., and G. Gerlach. 1987. Enterobacterial fimbriae J. Bacteriol. 169:934 938. 21. Cox, J 2000. Salmonella. p. 1928 1937. In B.C. Robinson RK, Patel PD (ed.), Encyclopedia of Food Microbiology, vol. 3. Academic Press, California. 22. Cummings, K., E. Barrett, J. Mohle Boetani, J. Brooks, J. Farrar, T. Hunt, A. Fiore, K. Komatsu, S. Werner, an d L. Slutsker. 2001. A multistate outbreak of Salmonella enterica serotype Baildon associated with domestic raw tomatoes Emerg Infect Dis 7:1046 1048. 23. Da Re, S., and J. M. Ghigo. 2006. A CsgD independent pathway for cellulose production and biofilm formation in Escherichia coli. J. Bacteriol. 188:3073 3087. 24. Davies, R., and W. C. 1996. Seasonal variations in the isolation of Salmonella typhimurium Salmonella enteritidis Bacillus cereus and Clostridium perfringens from environmental samples J. Vet. Med. B 43:119 127.
126 25. Dong, Y., A. Iniguez, B. Ahmer, and E. Triplett. 2003. Kinetics and strain specificity of rhizosphere and endophytic colonization by enteric bacteria on seedlings of Medicago sativa and Medicago truncatula. Appl. Environ. Micro biol. 69:1783 1790. 26. Ellermeier, C., A. Janakiraman, and J. Slauch. 2002. Construction of targeted mediated site specific recombination in bacteria Gene 290:153 161. 27. Florida Adminstrative Code 2010. Tom ato Good Agricultural Practices, Rule 5G 6. Florida Department of State, Tallahassee, FL. Available at: https://www.flrules.org/gateway/chapterhome.asp?chapter=5g 6. Accessed 12 December 2012 28. Foster, J. 1991. Salmonella acid shock proteins are require d for the adaptive acid tolerance response J Bacteriol 173:6896 6902. 29. Frenzen, P., T. Riggs, J. Buzby, T. Breuer, T. Roberts, D. Voetsch, S. Reddy and F. W. Group. 1999. Salmonella cost estimate update using FoodNet data Food Rev 22:10 15. 30. Ge rstal, U., and U. Romling. 2001. Oxygen tension and nutrient starvation are major signals that regulate agfD promoter activity and expression of the multicellular morphotype in Salmonella typhimurium. Environ. Microbiol 3:638 648. 31. Gerstal, U., and U. Romling. 2003. The csgD promoter, a control unit for biofilm formation in Salmonella typhimurium. Res. Microbiol 154:659 667. 32. Gibson, D., A. White, S. Snyder, S. Martin, C. Heiss, P. Azadi, M. Surette, and W. Kay. 2006. Salmonella produces an O Antige n capsule regulated by AgfD and important for environmental persistence J. Bacteriol 188:7722 7730. 33. Greene, S., E. Daly, E. Talbot, L. Demma, S. Holzbauer, N. Patel, T. Hill, M. Walderhaug, R. Hoekstra, M. Lynch, and J. Painter. 2008. Recurrent multi state outbreak of Salmonella Newport associated with tomatoes from contaminated fields, 2005 Epidemiol. Infect 136:157 165. 34. Gupta, S., K. Nalluswami, C. Snider, M. Perch, M. Balasegaram, D. Burmeister, J. Lockett, C. Sandt, R. Hoekstra, and S. Montgo mery. 2007. Outbreak of Salmonella Braenderup infections associated with Roma tomatoes, northeastern United States, 2004: a useful method for subtyping exp osures in field investigations. Epidemiol. Infect. 135:1165 1173. 35. Hammar, M., A. Arnqvist, Z. Bai n, A. Olsen, and S. Normark. 1995. Expression of two csg operons is required for production of fibronectin and Congo red binding curli polymers in Escherichia coli K 12 Mol Microbiol 18:661 670.
127 36. Hanning, I., J. Nutt, and S. Ricke. 2009. Salmonellos is outbreaks in the United States due to fresh porduce: sources and potential intervention measures Foodborne Path. Dis 6:635 648. 37. Heaton, J. C., and K. Jones. 2008. Microbial contamination of fruit and vegetables and the behaviour of enteropathogens in the phyllosphere: a review J. Appl. Microbiol 104:613 626. 38. Hedberg, C. W., F. J. Angulo, K. E. White, C. W. Langkop, W. L. Schell, M. G. Stobierski, A. Schuchat, J. M. Besser, S. Dietrich, L. Helsel, P. M. Griffin, J. W. McFarland, M. T. Osterhol m, and T. Invest. 1999. Outbreaks of salmonellosis associated with eating uncooked tomatoes: implications for public health Epidemiol. Infect 122:385 393. 39. Hughes, K., B. Olivera, and J. Roth. 1987. Rec dependence of mu transposition from P22 transduc ed fragments J. Bacteriol. 169:403 409. 40. Humphries, A., M. Raffatellu, S. Winter, E. Weening, R. Kingsley, R. Droleskey, S. Zhang, J. Figueiredo, S. Khare, J. Nunes, L. Adams, R. Tsolis, and A. Baumler. 2003. The use of flow cytometry to detect express ion of subunits encoded by 11 Salmonella enterica serotype Typhimurium fimbrial operons Mol. Microbiol. 48:1357 1376. 41. Inoue, H., H. Nojima, and H. Okayama. 1990. High efficiency transformation of Escherichia coli with plasmids Gene 96:23 28. 42. Isl am, M., J. Morgan, M. Doyle, S. Phatak, P. Millner, and X. Jiang. 2004. Persistence of Salmonella enterica serovar Typhimurium on lettuce and parsley and in soils on which they were grown in fields treated with contaminated manure composts or irrigation wa ter Foodborne Path Dis 1:27 35. 43. Iturriaga, M., M. Tamplin, and E. Escarton. 2007. Colonization of tomatoes by Salmonella Montevideo is affected by relative humidity and storage temperature J Food Prot 70:30 34. 44. Iturriaga, M. H., E. F. Escarti n, L. R. Beuchat, and R. Martinez Peniche. 2003. Effect of inoculum size, relative humidity, storage temperature, and ripening stage on the attachment of Salmonella Montevideo to tomatoes and tomatillos J. Food Prot 66:1756 1761. 45. Jain, S., and J. Che n. 2007. Attachment and biofilm formation by various serotypes of Salmonella as influenced by cellulose production and thin aggregative fimbriae biosynthesis J. Food Prot. 70:2473 2479. 46. Jeter, C., and A. Matthysse. 2005. Characterization of the bindin g of diarrheagenic strains of E. coli to plant surfaces and the role of curli in the interaction of the bacteria with alfalfa sprouts M ol. P lant Microbe Microbiol 18:1235 1242.
128 47. Jubelin, G., A. Vianney, C. Beloin, J. M. Ghigo, J. C. Lazzaroni, P. Leje une, and C. Dorel. 2005. CxpR/OmpR interplay regulates curli gene expression in response to osmolarity in Escherichia coli. J. Bacteriol 187:2038 2049. 48. Kovach, M., P. Elzer, D. Hill, G. Robertson, M. Farris, R. Roop, and K. Peterson. 1995. Four new de rivatives of the braod h ost range cloning vector pBBR1M S C 1, carrying different antibiotic resistance cassettes Gene 166:175 176. 49. Lapidot, A., and S. Yaron. 2009. Transfer of Salmonella enterica serovar Typhimurium from contaminated irrigation water t o parsley is dependent on curli and cellulose, the biofilm matrix components J. Food Prot 72:618 623. 50. Lee, K. K., and K. C. Yii. 1996. A comparison of three methods for assaying hydrophobicity of pathogenic vibrios Lett. Appl. Microbiol. 23:343 346. 51. Lindow, S. E., and M. T. Brandl. 2003. Microbiology of the phyllosphere Appl. Environ. Microbiol 69:1875 1883. 52. Ljungh, A., S. Hjerten, and T. Wadstrom. 1985. High surface hydrophobicity of autoaggregating Staphylococcus aureus strains isolated f rom human infections studied with the salt aggregation test Infect. Immun. 47:522 526. 53. Mahon, B. E., A. Ponka, W. N. Hall, K. Komatsu, S. E. Dietrich, A. Siitonen, G. Cage, P. S. Hayes, M. A. Lambert Fair, N. H. Bean, P. M. Griffin, and L. Slutsker. 1 997. An international outbreak of Salmonella infections caused by alfalfa sprouts grown from contaminated seeds J. Infect. Dis 175:876 882. 54. Mattos Guaraldi, A., L. Duarte Formiga, and A. Braga Andrade. 1999. Cell surface hydrophobicity of sucrose fer menting and nonfermenting Cornyebacterium diphtheriae strains evaluated by different methods Curr. Microbiol. 38:37 42. 55. Merighi, M., C. Ellermeier, J. Slauch, and J. Gunn. 2005. Resolvase in vivo expression technology analysis of the Salmonella enteri ca serovar Typhimurium PhoP and PmrA refulons in BALB/c mice J. Bacteriol. 187:7407 7416. 56. Noel, J., N. Arrach, A. Alagely, M. McClelland, and M. Teplitski. 2010. Specific responses of Salmonella enterica to tomato varieties and fruit ripeness identifi ed by in vivo expression technology PLoS ONE 5:e12406. 57. Painter, J., R. Hoekstra, T. Ayers, R. Tauxe, C. Braden, F. Angulo, and P. M. Griffin. 2013. Attribution of foodborne illnesses, hospitalizations, and deaths to food commodities using outbreak da ta, United States, 1998 2008 Emerg Infect Dis 19:407 415 58. Prouty, A., and J. Gunn. 2003. Comparative analysis of Salmonella enterica serovar Typhimurium biofilm formation on gallstones and on glass Infect. Immun. 71:7154 7158.
129 59. Public Law. 2011 FDA Food Modernization Act Public Law 111 353, 124 STAT. 3885. Available at http://www.gpo.gov/fdsys/pkg/PLAW 111publ353/pdf/PLAW 111publ353.pdf. Accessed 10 December 2012. 60. Reller, M., R. Tauxe, L. Kalish, and K. Molbak. 2008. Excess salmonellosis in women in the United States: 1968 2000 Epidemiol Infect 136:1109 1117. 61. Romling, U. 2005. Characterization of the rdar morphotype, a multicellular behaviour in Enterobacteriaceae Cell. Mol. Life Sci 62:1234 1246. 62. Romling, U., Z. Bain, M. Hammer, W. Sierralta, and S. Normark. 1998. Curli fibers are highly conserved between Salmonella typhimurium and Escherichia coli with respect to operon structure and regulation J. Bacteriol 180:722 731. 63. Romling, U., M. R hode, A. Olsen, S. Normark, and J. Reinkoster. 2000. AgfD, the checkpoint of multicellular and aggregative behaviour in Salmonella typhimurium reguates at least two independent pathways Mol. Microbiol 36:10 23. 64. Romling, U., W. Sierralta, K. Eriksson, and S. Normark. 1998. Multicellular and aggregative behavior of Salmonella typhimurium strains is controlled by mutations in the agfD promoter Mol. Microbiol 28:249 264. 65. Rosenberg, M., D. Gutnick, and E. Rosenberg. 1980. Adherence of bacteria to hyd rocarbons: a simple method for measuring cell surface hydrophobicity. FEMS Microbiol. Lett. 9:29 33. 66. Santiviago, C., M. Reynolds, S. Porwollik, S. H. Choi, F. Long, H. Andrews Polymenis, and M. McClelland. 2009. Analysis of pools of targeted Salmonella deletion mutants identifies novel genes affecting fitnessduring competitive infection in mice PLoS Pathog 5:e1000477. 67. Sapers, G. M., and D. M. Jones. 2006. Improved sanitizing treatments for fresh tomatoes J. Food Sci 71:M252 M256. 68. Scallan, E. P. Griffin, F. Angulo, R. Tauxe, and R. Hoekstra. 2011. Foodborne illness aquired in the United States Unspecified agents Emerg Infect Dis 17:16 22. 69. Scallan, E., R. Hoekstra, F. Angulo, R. Tauxe, M. Widdowson, S. Roy, J. Jones, and P. Griffin. 20 11. Foodborne Illness Acquired in the United States Major Pathogens Emerg Infect Dis 17: 7 15 70. Scher, K., U. Romling, and S. Yaron. 2005. Effect of heat, acidification, and chlorination on Salmonella enterica serovar Typhimurium cells in a biofilm f ormed at the air liquid interface Appl. Environ. Microbiol 71:1163 1168.
130 71. Schikora, A., A. Carreri, E. Charpentier, and H. Hirt. 2008. The dark side of the salad: Salmonella typhimurium overcomes the innate immune response of Arabidopsis thaliana and shows an endopathogenic lifestyle PLoS ONE 3:e2279. 72. Schuenzel, K. M., and M. A. Harrison. 2002. Microbial antagonists of foodborne pathogens on fresh, minimally processed vegetables J Food Prot 65:1909 1915. 73. Shi, X., A. Namvar, M. Kostrzynska R. Hora, and K. Warriner. 2007. Persistence and growth of different Salmonella serovars on pre and postharvest tomatoes J. Food Prot 70:2725 2731. 74. Sivapalasingam, S., E. Barrett, A. Kimura, S. Van Duyne, W. De Witt, M. Ying, A. Frisch, Q. Phan, E. Gould, P. Shillam, V. Reddy, T. Cooper, M. Hoekstra, C. Higgins, J. P. Sanders, R. V. Tauxe, and L. Slutsker. 2003. A multistate outbreak of Salmonella enterica serotype Newport infection linked to mango consumption: Impact of water dip disinfestation tec hnology Clin. Infect. Dis 37:1585 1590. 75. Sivapalasingam, S., C. Friedman, L. Cohen, and R. Tauxe. 2004. Fresh produce: a growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997 J Food Prot 67:2342 2353. 76. Solano, C ., G. Valle, C. Berasain, J. M. Ghigo, C. Gamazo, and I. Lasa. 2002. Genetic analysis of Salmonella enteritidis biofilm formation: critical role of cellulose Mol. Microbiol 43:793 808. 77. Solomon, E., M. Brandl, and R. Mandrell. 2006. Biology of foodbor ne pathogens on produce In K. Matthews (ed.), Microbiology of Fresh Produce ASM Press, Washington, DC. 78. Solomon, E., B. Niemira, G. Sapers, and B. Annous. 2005. Biofilm formation, cellulose production, and curli biosynthesis by Salmonella originating f rom produce, animal, and clinical sources J. Food Prot. 68:906 912. 79. Steele, M., and J. Odumeru. 2004. Irrigation water as source of foodborne pathogens on fruit and vegetables J. Food Prot 67:2839 2849. 80. Steenackers, H., K. Hermans, J. Vanderleyd en, and S. D. Keersmaecher. 2011. Salmonella biofilms: an overview on occurrence, structure, regulation, and eradication Food Res Int 45:502 531 81. Stine, S. W., I. Song, C. Y. Choi, and C. P. Gerba. 2005. Effect of relative humidity on preharvest surv ival of bacterial and viral pathogens on the surface of cantaloupe, lettuce, and bell peppers J. Food Prot 68:1352 1358. 82. Sukupolvi, S., R. Lorenz, J. Gordon, Z. Bain, J. Pfeifer, S. Normark, and M. Rhen. 1997. Expression of thin aggregative fimbriae promotes interaction of Salmonella typhimurium SR 11 with mouse small intestine epithelial cells Infect. Immun 65:5320 5325.
131 83. Tauxe, R., H. Kruse, C. Hedberg, M. Potter, J. Madden, and K. Wachsmuth. 1997. Microbial hazards and emerging issues associat ed with produce A preliminary report to the National Advisory Committee on Microbiologic Criteria for Foods J. Food Prot 60:1400 1408. 84. Teplitski, M., J. Noel, A. Alagely, and M. Danyluk. 2012. Functional genomics studies shed light on the nutrition and gene expression of non typhoidal Salmonella and enterovirulent E. coli in produce Food Res Int 45:576 586. 85. Tournas, V. H. 2005. Spoilage of vegetable crops by bacteria and fungi and related health hazards Crit. Rev. Microbiol 31:33 44. 86. To wnshed, S., N. Kramer, R. Edwards, S. Baker, N. Hamlin, and A. Baulmer. 2001. Salmonella enterica serotype Typhi possesses a unique repertoire of fimbrial gene sequences Infect. Immun. 69:2894 2901. 87. Ukuku, D. O., and G. M. Sapers. 2007. Effect of time before storage and storage temperature on survival of Salmonella inoculated on fresh cut melons Food Microbiol 24:288 295. 88. USDA. 1991. Standards for Grades of Fresh Tomatoes In USDA (ed.) Agr. Mktg. Serv., Washington, DC. 89. Van Beneden, C. A., W. E. Keene, R. A. Strang, D. H. Werker, A. S. King, B. Mahon, K. Hedberg, A. Bell, M. T. Kelly, V. K. Balan, W. R. Mac Kenzie, and D. Fleming. 1999. Multinational outbreak of Salmonella enterica serotype newport infections due to contaminated alfalfa sprout s Jama J. Am. Med. Assoc 281:158 162. 90. Van der Velden, A., A. Baulmer, R. Tsolis, and F. Heffron. 1998. Multiple fimbrial adhesins are required for full virulence of Salmonella typhimurium in mice Infect. Immun. 66:2803 2808. 91. Vidal, O., R. Longin C. Pringet Combaret, C. Dorel, M. Hooreman, and P. Lejeune. 1998. Isolation of an Escherichia coli K 12 mutant strain able to form biofilms on inert surfaces involvement of a new ompR allele that increases curli expression J. Bacteriol 180:2442 2449. 9 2. Vought, K., and S. Tatini. 1998. Salmonella enteritidis contamination of ice cream associated with 1994 multistate outbreak J Food Prot 61:5 10. 93. Wang, R., and S. Kushner. 1991. Construction of versatile low copy number vectors for cloning, sequen cing and gene expression in Escherichia coli. Gene 100:195 199. 94. Warner, J., S. Rothwell, and C. Keevil. 2008. Use of episcopic differential interference contrast microscopy to identify bacterial biofilms on salad leaves and track colonization by Salmo nella Thompson Environ Microbiol 10:918 25.
132 95. Wells, J. M., and J. E. Butterfield. 1997. Salmonella contamination associated with bacterial soft rot of fresh fruits and vegetables in the marketplace Plant Dis 81:867 872. 96. Whipps, J. M., P. Hand, D. A. C. Pink, and G. D. Bending. 2008. Human pathogens and the phyllosphere p. 183 221. In Advances i n Applied Microbiology, Vol 64 Elsevier Academic Press Inc, San Diego. 97. White, A., S. Collinson, P. Banser, D. Gibson, M. Paetzel, N. Strynadka, and W. Kay. 2001. Structure and characterization of AgfB from Salmonella enteritidis thin aggregative fimbriae J Mol. Biol. 311:735 749. 98. White, A., D. Gibson, S. Collinson, P. Banser, and W. Kay. 2003. Extracellular polysaccharides associated with thin a ggregative fimbriae of Salmonella enterica serovar Enteritidis J. Bacteriol. 185:5398 5407. 99. White, A., D. Gibson, G. Grassl, W. Kay, B. Finlay, B. Vallance, and M. Surette. 2008. Aggregation via the red, dry, and rough morphotype is not a virulence ad aptation in Salmonella enterica serovar Typhimurium Infect. Immun. 76:1048 1058. 100. White, A. P., and M. G. Surette. 2006. Comparative genetics of the rdar morphotype in Salmonella. J. Bacteriol 188:8295 8406. 101. Winfield, M., and E. Groisman. 2003. Role of nonhost environments in the lifestyles of Salmonella and Escherichia coli. Appl. Environ. Microbiol. 69:3687 3694. 102. Wright, A., L. Simpson, J. Oliver, and J. Morris. 1990. Phenotypic evaluation of acapsular transposon mutants of Vibrio vulnific us. Infect. Immun. 58:1769 1771. 103. Yuk, H., and K. Schneider. 2006. Adaptation of Salmonella spp. in juice stored under refrigated and room temperature enhaces acid resistance to simulated gastric fluid Food Microbiol 23:694 700. 104. Zaragoza, W., J. Noel, and M. Teplitski. 2012. Spontaneous non rdar mutations increase fitness of Salmonella in plants Environ Microbiol Reports 4:453 458. 105. Zhuang, R. Y., L. R. Beuchat, and F. J. Angulo. 1995. Fate of Salmonella Montevideo on and in raw tomatoes affected by temperature and treatment with chlorine Appl. Environ Microbiol 61:2127 2131. 106. Zogaj, X., M. Nimtz, M. Rohde, W. Bokranz, and U. Romling. 2001. The mulitcellular morphotypes of Salmonella typhimurium and Escherichia coli produce cellulose as the second component of the extracellular matrix Mol. Microbiol. 39:1452 1463.
133 BIOGRAPHICAL SKETCH Marianne Kathryn Fatica was born in 1986 in Albany, New York and grew up in Cape Coral, Florida. After graduating from Mariner High School in 2004 a s valedictorian Marianne attended the University of South Florida in Tampa, Florida. During her undergraduate studies, she became involved in laboratory research through the Research Experience for Undergraduates (REU) program. Through the Department of Chemistry, Marianne analyzed protein composition of European and Africanized Honeybees as part of a project to develop an enzyme linked immunosorbent assay (ELISA) test for the rapid identification of Africanized Honeybees. She was also involved in the A nthropology department, where she completed her undergraduate The Application of the Anthropological Perspective in Medicine to Address Problems within Physician the Uni versity of Sou th Florida in 2008 with a Bachelor of Science in microbiology and a minor in biological a nthropology. Marianne began her graduate degree in the Department of Food Science and Human Nutrition at the University of Florida in 2008. As someone new to the fiel d of Food Science, Marianne felt it was important to become certified in Good Agricultural Practices and HACCP for juice processors to better understand the application of these safety programs. Along with her doctoral program, Marianne also divers ified h er studies by earning a certification in public h ealth, with an emphasis in environmental health, from the University of Florida Public Health Department in 2011. Along with her graduate work Marianne was highly involved in teaching microbiology courses within the department. Marianne has also been honored with several awards throughout her
134 Association of Fo od Protection Scholarships. After completing her doctoral degr ee in May 2013, Marianne will continue her professional interests in food microbiology.