<%BANNER%>

Improved Sample Preparation for the Molecular Detection of Shigella sonnei in Foods


PAGE 1

IMPROVED SAMPLE PREPARATION FO R THE MOLECULA R DETECTION OF Shigella sonnei IN FOODS By BENJAMIN RAY WARREN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Benjamin Ray Warren

PAGE 3

To Nikki, for your undying love and support; to Zachary, for bringing so much love into our lives; to my parents, for never losing faith in me; and to all my friends along the way, for without all of you this w ould not have been possible.

PAGE 4

iv ACKNOWLEDGMENTS First and foremost, I would like to thank my committee chair, Dr. Keith Schneider, for all of his support and guidance over my gr aduate studies. Additionally, I would like to thank Dr. Mickey Parish for seeing my poten tial and inspiring me to return to the University of Florida to pursue my Ph.D., and Dr. Schneider for always having an open door and encouraging my questions and ideas. I would further like to thank the remaining members of my graduate committee, Dr. Douglas Archer, Dr. Eric Triplett and Dr. Keith Lampel, for all of their assistance with th is project; each of them brought a unique perspective to this project a nd provided excellent support. Furthermore, I would like to thank my wife Nicole, whose hard work and sacrifice during my graduate studies has made all of this possible; she will always have my admiration and love. I would also like to tha nk my parents, Dennis and Linda Warren, for their never-ending love and support not only in recent years but throughout my life; I thank them for setting such a good examples. Statistical assistance was provi ded by University of Florid a, IFAS Statistics, with special thanks to Meghan Brennan. This project was funded in part by the USDACSREES IFAFS Grant number 00-52102-9637.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES...........................................................................................................xi ABSTRACT......................................................................................................................x ii CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................6 Shigella as a Foodborne Pathogen................................................................................6 Epidemiology of Shigella ......................................................................................7 Demographic Variability of Infections with Shigella spp.....................................8 Recently Identified Serotype of S. dysenteriae .....................................................9 Foodborne Outbreaks Involving Shigella ..............................................................9 Prevalence of Shigella : Food and Food Handlers...............................................11 Survival of Shigella ....................................................................................................12 Environmental Factors on Survival of Shigella ...................................................12 Survival of Shigella on Fomites..........................................................................14 Survival of Shigella in Food and Water..............................................................15 Current Understanding of Shigella Pathogenesis.......................................................17 Shigella Invasion of Epithelial Cells...................................................................18 Potential Roles of the Ip aH Effector Proteins.....................................................18 IpaH7.8 facilitates escape fr om endocytic vacuoles....................................18 Subversion of host cell signaling by IpaH9.8..............................................19 Blockage of Autophagy by IcsB..........................................................................20 Genetic Relationship Between Shigella and Escherichia coli ....................................21 Detection Methods for Shigella in Foods...................................................................23 Conventional Culture Methods for Shigella ........................................................23 Traditional microbiological media fo r enrichment and isolation of Shigella ...................................................................................................23 The FDA Bacteriological Analytical Manual culture method for detection of Shigella in foods.................................................................27 Other culture methods for the detection of Shigella in foods......................28

PAGE 6

vi Immunological Methods for Shigella Detection.................................................29 Differences in S. sonnei form I and form II lipopolysaccharide..................30 Immunological Detection Methods for Bacteria.................................................31 Latex agglutination methods for Shigella ....................................................33 Enzyme immunoassay methods for Shigella ...............................................35 Immunomagnetic separation methods for Shigella detection......................36 Immunomagnetic separati on using the Pathatrix.........................................37 Additional technologies for immunological detection of Shigella ...............38 Molecular Microbiological Methods for Shigella Detection in Foods................39 Polymerase chain reaction detection of Shigella in foods............................39 Improved PCR detection of Shigella by FTA filtration...............................41 Additional molecular microbiologi cal techniques for detection of Shigella ...................................................................................................43 Detection of Pathogens by Sequence Capture............................................................46 3 MATERIALS AND METHODS...............................................................................50 Preliminar y Studies.....................................................................................................50 Preparation of Micr obiological Media................................................................51 Acquisition and Maintenance of Shigella sonnei Cultures..................................50 Adaptation of Cultures to Rifampicin.................................................................50 Acquisition/Preparati on of Food Matrices..........................................................51 Acquisition and Maintenance of AntiShigella Antibodies.................................52 Binding of Antibodies to Paramagnetic Beads....................................................53 Evaluation and Optimization of Immunocapture Using AntiShigella Beads....54 Crude DNA Extraction from Bacteria by Boiling...............................................55 DNA Extraction from AntiShigella Beads using the DNeasy Kit.....................56 Preparation of HeLa Cell Extracts.......................................................................56 RNA Extraction Using the RNeasy Kit...............................................................57 DNase Treatment of RNA Extracts Prior to RT-PCR.........................................58 Induction and Expression of ipaH RNA in S. sonnei ..........................................59 Identification of Shigella -Specific Genetic Loci.................................................59 Development of Primers/Probes for the Detection of Shigella ...........................60 Evaluation of Primer/Probe Specificity...............................................................60 Binding of Biotinylated Capture Probes to Streptavidin-Coated Paramagnetic Beads................................................................................................................64 Inoculum Preparation..........................................................................................65 Calculation of Generation Time of S. sonnei in Shigella Broth..........................65 Preliminary Experiments with AntiShigella Beads............................................66 Separation of S. sonnei from Food Matrices Using Low-Speed Centrifugation.66 Survival Studies..........................................................................................................67 Sample Inoculation and Subsequent Recovery...................................................67 Three-Tube Most Probable Numb er Estimation of Survivors............................68 Evaluation of Detection Methods...............................................................................68 Inoculation of Samples a nd Subsequent Recovery.............................................70 Modified BAM Culture Method for S. sonnei .....................................................70 Flow-Through Immunocapture (FTI) Using the Pathatrix..................................71

PAGE 7

vii Sequence Capture of Shigella DNA....................................................................72 Real-Time PCR and Reverse Transcriptase (RT) PCR.......................................74 Recording of Data and Statistical Analysis.........................................................75 4 RESULTS...................................................................................................................77 Preliminar y Studies.....................................................................................................77 Calculation of Generation Time of S. sonnei in Shigella Broth..........................77 Growth curve of S. sonnei ATCC 9290.......................................................78 Growth curve of S. sonnei ATCC 29031.....................................................78 Growth curve of S. sonnei ATCC 29030.....................................................78 Growth curve of S. sonnei ATCC 25931.....................................................80 Growth curve of S. sonnei ATCC 29930.....................................................80 Evaluation AntiShigella Antibodies for Use with Flow-Through Immunocapture................................................................................................82 Preliminary Experiments with AntiShigella Beads............................................84 Optimization of AntiShigella Bead Concentration for Flow-Through Immunocapture of S. sonnei ............................................................................85 Identification of Potentially Shigella -Specific Genetic Loci...............................87 Specificity of Primers Developed for Potentially Shigella -Specific Genetic Loci..................................................................................................................87 Separation of S. sonnei from Food Matrices by Low-Speed Centrifugation......90 Development of DNA Sequence Captur e (DSC) for the Detection of S. sonnei ...............................................................................................................91 Expression of ipaH RNA in Log and Stationary Phase S. sonnei .......................95 Survival Studies..........................................................................................................96 Survival of S. sonnei on Smooth Tomato Surfaces.............................................96 Survival of S. sonnei in Potato Salad...................................................................97 Survival of S. sonnei in Ground Beef..................................................................99 Evaluation of Detection Methods.............................................................................101 Detection of S. sonnei in Selected Foods by a Modified FDA Bacteriological Analytical Manual (BAM) Shigella Culture Method....................................101 Detection of S. sonnei in Selected Foods by Flow-Through Immunocapture (FTI)...............................................................................................................102 Detection of S. sonnei in Selected Foods by DNA Sequence Capture (DSC)..104 5 DISCUSSION AND CONCLUSIONS....................................................................106 Preliminar y Studies...................................................................................................106 Growth Characteristics of S. sonnei in Shigella Broth (SB).............................107 Expression and Induction of the ipaH Gene of S. sonnei ..................................108 Identification of Potentially Shigella -Specific Genetic Loci.............................109 Survival Studies........................................................................................................110 Rapid S. sonnei Inactivation on Tomato Surfaces.............................................111 S. sonnei Survives in Potato Salad and Ground Beef........................................112 Evaluation of Detection Methods.............................................................................112 Recovery of S. sonnei by the BAM Shigella Culture Method..........................112

PAGE 8

viii Analysis of lowest detection levels of the BAM Shigella culture method.113 Evaluation of Flow-Through Immunocapture for the Detection of S. sonnei in Food...............................................................................................................115 Operational issues with antiShigella beads in flow-through immunocapture.....................................................................................115 Non-specific immunocapture of Enterobacter cloacae and Escherichia coli ........................................................................................................117 Analysis of lowest detection le vels of the FTI-MAC and FTI-PCR methods.................................................................................................119 Future research involving flow-thr ough immunocapture for the detection of S. sonnei ............................................................................................122 Evaluation of DNA Sequence Captur e (DSC) for the Detection of S. sonnei in Food...............................................................................................................123 Evaluation of hybridization buffers for DSC.............................................123 Non-specific adsorption of DNA to CPShigella beads.............................124 Analysis of lowest detecti on levels of the DSC method............................126 Future research to improv e DSC for the detection of S. sonnei in food....128 Sources of Variation Among Inoculated Studies..............................................129 Practical Applications of Flow-T hrough Immunocapture and DNA Sequence Capture...........................................................................................................130 Conclusions...............................................................................................................131 APPENDIX A PREPARATION OF BUFFERS AND SOLUTIONS.............................................132 Binding/Washing (B/W) Buffer (2X).......................................................................132 Congo Red Solution..................................................................................................132 Dynabeads Solution A..............................................................................................133 Dynabeads Solution B..............................................................................................133 Hybridization Buffer 1..............................................................................................133 Hybridization Buffer 2..............................................................................................134 Low-Salt Wash Buffer..............................................................................................135 Phosphate Buffered Saline, pH 7.4...........................................................................135 Sodium Acetate Buffer, pH 4.0................................................................................136 Sodium Bicarbonate Buffer, pH 8.6.........................................................................136 Wash Buffer..............................................................................................................136 B ALIGNMENT OF CHROMOSOMALLY-LOCATED ipaH GENES OF Shigella sonnei ........................................................................................................................138 LIST OF REFERENCES.................................................................................................143 BIOGRAPHICAL SKETCH...........................................................................................156

PAGE 9

ix LIST OF TABLES Table page 2-1. Percentage of Shigella isolates in the United States reported by PHLIS in recent years.......................................................................................................................... .9 2-2. Selected foodbor ne outbreaks involving Shigella .....................................................10 3-1. Antibodies investigated for immunocapture of S. sonnei ...........................................54 3-2. Primers designed for the detection of Shigella. ..........................................................61 3-3. Shigella and nonShigella strains tested for specificity..............................................62 4-1. Evaluation of antiShigella antibodies for flow-through immunocapture (FTI) of S. sonnei ...................................................................................................................83 4-2. Optimization of antiShigella bead concentration for flow-through immunocapture (FTI) of S. sonnei ...........................................................................86 4-3. Genetic targets identified with potential specificity for Shigella spp. or for S. sonnei alone..............................................................................................................88 4-4. Evaluation of primer specificity among stock Shigella cultures and by comparative analysis against previously sequenced Shigella genomes...................89 4-5. Effects of low-speed centrifugation on S. sonnei populations in sample supernatant...............................................................................................................91 4-6. Evaluation of hybridization buffers for DSC for the detection of S. sonnei ..............92 4-7. Comparison of paramagnetic beads for use with CPShigella beads.........................94 4-8. Specific capture of ipaH DNA in the presence of non-target DNA by CPShigella beads.........................................................................................................................9 5 4-9. Sensitivity of DNA sequence capture method............................................................96 4-10. Transcriptional induction of the ipaH gene using HeLa cell extracts and the dye Congo red.................................................................................................................97 4-11. Number of samples positive for S. sonnei by various detection methods..............102

PAGE 10

x 5-1. Lowest detection levels of the BAM Shigella culture method.................................114 5-2. Lowest detection levels of the FTI-MAC and FTI-PCR methods............................120 5-3. Lowest detection leve ls of the DSC method............................................................127

PAGE 11

xi LIST OF FIGURES Figure page 2-1. Invasion of epithelial cells by Shigella spp................................................................19 2-2. Structural differences between hexose regions of Shigella sonnei and Shigella flexneri lipopolysaccharides.....................................................................................30 2-3. O-antigen repeating subunits of S. sonnei and S. flexneri lipopolysaccharide...........31 2-4. Lipopolysaccharides of S. sonnei ...............................................................................32 2-5. The Pathatrix system for flow -through immunomagnetic separation........................38 3-1. Design of the capture probe and CPShigella beads...................................................65 3-2. Flow diagram of the experiments involving inoculated food samples.......................69 3-3. Flow diagram of the DNA sequence capture method.................................................73 4-1. Growth curve: S. sonnei ATCC 9290 in Shigella broth.............................................79 4-2. Growth curve: S. sonnei ATCC 29031 in Shigella broth...........................................79 4-3. Growth curve: S. sonnei ATCC 29030 in Shigella broth...........................................80 4-4. Growth curve: S. sonnei ATCC 25931 in Shigella broth...........................................81 4-5. Growth curve: S. sonnei ATCC 29930 in Shigella broth...........................................81 4-6. Survival of a five-strain S. sonnei cocktail on the smooth surfaces of tomatoes.......98 4-7. Survival of five-strain S. sonnei cocktail in potato salad...........................................99 4-8. Survival of five-strain S. sonnei cocktail in ground beef.........................................100 5-1. Location of antiShigella beads in the capture phase of the Pathatrix during FTI...118

PAGE 12

xii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IMPROVED SAMPLE PREPARATION FO R THE MOLECULA R DETECTION OF Shigella sonnei IN FOODS By Benjamin Ray Warren August 2006 Chair: Keith R. Schneider Major Department: Food Science and Human Nutrition Shigella the causative agent of bacillary dysentery, was the third most reported foodborne bacterial pathogen in the U.S. fo r 2005, and most of the isolates were identified as S. sonnei Methods for the detection of Shigella in food, however, remain problematic. In preliminary studies, a chromo somally-located genetic target for specific detection of Shigella RNA was investigated by comparing the available genomes of Shigella and E. coli All DNA sequences identified with potential specificity for all Shigella spp. or with potential specificity for only S. sonnei using database searches tested positive for E. coli strains that had been isolated from ground beef. Additionally, the surviv al of a five-strain S. sonnei cocktail on tomato surfaces, in potato salad and in ground beef was investig ated using a most probable number (MPN) method. Inoculated tomatoes were stored at 13C at 85% relative humidity, while potato salad and ground beef samples were stored at 2.5C and 8C. On tomato surfaces, S. sonnei populations declined to undetectable le vels by day 3. In potato salad and ground

PAGE 13

xiii beef samples, S. sonnei populations detected at the end of the product shelf-life (28 days and 11 days, respectively) were not significantly different ( P > 0.05) than initial populations. These stud ies suggest that S. sonnei survives well in foods when not desiccated. Flow-through immunocapture (FTI) follo wed by analysis of recovered antiShigella beads by spread-plate using MAC (FT I-MAC), FTI followed by analysis of recovered antiShigella beads by real-time PCR (FTI-PCR) and DNA sequence capture (DSC) were compared to the Shigella culture method of the FDA Bacteriological Analytical Manual (BAM) for the detection of S. sonnei on tomato surfaces, in potato salad and in ground beef. FTI-MAC was significantly better ( P > 0.05) than the BAM Shigella culture method for the analysis of tomatoes, but not potato salad or ground beef. FTI-PCR and DSC were significantly better ( P > 0.05) than the BAM Shigella culture method for the analysis of tomatoes and potato salad, but not ground beef.

PAGE 14

1 CHAPTER 1 INTRODUCTION The ability to analyze food products for th e presence of pathogenic bacteria is essential for verifying the sa fety of foods, identifying agents of foodborne illness and determining sources of foodborne outbr eaks. Conventionally, the microbiological analysis of food involves cultu re enrichment followed by isolation on selective media. Confirmation of presumptive isolates is ge nerally through biochemical characterization and/or serology. Conventional culture methods however, are often problematic, in that many are time-consuming and require several days to complete, appropriate selective media are not currently available for all ba cterial foodborne pathogens, some bacterial pathogens require specific atmospheric or ot her growth conditions which may be difficult to simulate in the laboratory and some bacterial pathogens may not be culturable by currently available methods. In addition, most culture enrichment procedures used for the detection of bacterial foodborne pathogens de tect only the presence or absence of the target pathogen. For enumeration, a most probable number (MPN) method must be employed; however media required for MP N analysis for some bacterial foodborne pathogens is not currently available. For bacterial foodborne pathogens whos e conventional culture methods are problematic, alternative sample preparation me thods may be used to improve sensitivity. Alternative sample preparation methods provi de a means of separating and concentrating bacterial pathogens or component s (proteins, nucleic acids, etc.) of bacterial pathogens from food matrices. If molecu lar-based detection is to fo llow, the alternative sample

PAGE 15

2 preparation method must also c ontend with any potential inhi bitors which may be present in the food. Unfortunately, there is no a lternative sampling method suitable for the separation and concentration of all types b acterial pathogens from all forms of food; therefore each combination of sampling met hod, food matrix and bacterial pathogen must be investigated independently. Once bacterial pathogens are separated a nd concentrated from food, there are many options for detection by rapid methods, such as immunoassays (enzyme linked immunosorbent assay (ELISA) and lateral flow devices), DNA hybridizations or the polymerase chain reaction (PCR). Most ra pid methods require the presence of 103 colony forming units (CFU)/ml of the target bacterial pa thogen for consistent and dependable detection (Stevens and Jaykus, 2004); therefore efficient separation and concentration of bacterial pathogens from f ood are critical. Because bacterial pathogens may be present in food at low, yet potent ially infectious populations, selective or nonselective enrichment is often pe rformed prior to sample analysis to increase the bacterial population. For some sample preparation methods very short enrichment times (4-5 hr) may be sufficient to increase the bacterial populations to de tectable levels, allowing the assay to be performed in a same-day format (Yuk et al ., 2006; Schneider and Warren, unpublished data). The detection of Shigella spp. in food is one example of where the use of alternative sample preparation methods may be used to improve analysis over conventional culture methods. The most commonly used Shigella culture method in the U.S. is found in the U.S. Food and Drug Administration’s Bacteriological Analytical Manual (BAM). The enrichment medi a recommended in the BAM is Shigella broth (SB),

PAGE 16

3 a low-carbohydrate medium used to limit th e decrease in pH associated with acid production from the microbiological meta bolism of sugars. SB, however, does not provide adequate specificity for Shigella spp. and other members of the family Enterobacteriaceae have been reported to out-compete and/or overgrow Shigella spp. during enrichment (Uyttendaele et al ., 2001; Warren, 2003; Warren et al ., 2005a). Other enrichment media, such as Enterobacteriaceae Enrichment (EE) broth (Uyttendaele et al ., 2000) and Gram-negative (GN) broth (C MMEF), have been suggested for the enrichment of Shigella spp.; however EE broth has been reported to be inhibitory to S. boydii serotype 18 (Warren, 2003; Warren et al ., 2005b) and GN broth contains bile salts and sodium desoxycholate, which have been sh own to inhibit stressed shigellae (Tollison and Johnson, 1985; Uyttendaele et al ., 2001). Isolation media are also problematic for Shigella spp. The BAM recommends MacConkey agar (MAC), which is selective fo r Gram-negative bacteria and differential based on the utilization of lactose. Typically, Shigella spp. are lactose negative; however some serotypes of S. boydii have been reported as lactose positive. During the Gulf War in the early 1990’s, lactose positive S. sonnei were isolated from U. S. military personnel (Dr. D.J. Kopecko, FDA, personal communi cation). Upon further investigation, it was discovered that typical lactose negative S. sonnei mutate at high frequency to lactose positive phenotypes during stationary phase on lactose-containing microbiological media (Dr. D.J. Kopecko, FDA, pers onal communication). It has yet to be determined whether lactose positive S. sonnei mutants can form in contaminated food products and whether this mutation would allow S. sonnei a means to evade detection by conventional culture methods, such as the BAM.

PAGE 17

4 Previously, a filtration-based sample preparation method using FTA filters (Whatman, Clifton, NJ) in combination with nested PCR was investigated for the detection of S. sonnei and S. boydii on tomato surfaces (Warren, 2003; Warren et al ., 2005b). When a sample is applied to FTA filters, the moisture from the sample activates chemical denaturants, chelating agent buffers and free radical traps embedded in the filter which lyse cells, denature enzymes, inac tivate pathogens and immobilize genomic DNA (Whatman, 2006). Using a tandem filter funnel sy stem in which the first filter funnel was for size exclusion of sample material and the second filter funnel contained an FTA filter, 100 ml PBS rinses of tomatoes were analyzed and the captured Shigella DNA was amplified by nested PCR. This FTA filtrati on-nested PCR assay was able to detect S. sonnei and S. boydii on tomato surfaces at inocula tion levels as low as 6.2-7.4 CFU/tomato. Unfortunately, the FTA filtration system was not as successful when applied to other types of produce, such as strawberries, cantaloupes or retail Valencia oranges. The analysis of these types of pr oduce in the FTA filtration system resulted in clogged filters and poor detec tion limits as compared to those performed on tomato rinses. Recently, a novel device for flow-through immunomagnetic separation (IMS), the Pathatrix, has been develope d (Matrix MicroScience, Inc., Golden, CO). IMS methods involve the coupling of specific antibodies to paramagnetic beads, which exhibit magnetic qualities only when placed in a ma gnetic field. In IMS methods, the antibodies are used to capture target bacteria and then the bead-bacteria complexes are separated and concentrated from the food matrix using a magnet. Traditional IMS methods analyze small sample volumes (1.0 ml), whereas the Pathatrix allows the analysis of a much

PAGE 18

5 larger sample volume (250 ml) by using a tubing system in which IMS beads are immobilized in a capture phase and the sa mple is continuously pumped through the capture phase using a peristaltic pump. Once recovered, the IMS beads can be analyzed by spread-plate for the isolati on of viable colonies or appl ied directly to DNA extraction methods for subsequent analysis by PCR. Finally, sample preparation methods based on the specific isolation of bacterial or viral DNA/RNA from various types of matrices usi ng oligonucleotide probes immobilized on paramagnetic beads have b een reported. Most commonly, biotinylated probes are attached to parama gnetic beads pre-coated with streptavidin. The bead-probes are then used to sample mixtures contai ning sample DNA/RNA in hybridization buffer. Following hybridization, the captured DNA/RNA can be rem oved from the bead-probes using heat and used as template in PCR, RT-PCR or other molecular-based detection assay. The hypothesis of this study was that specific antibodies and/or specific oligonucleotide probes may be attached to pa ramagnetic beads and used for the detection of S. sonnei with increased sensitiv ity over the current conventional culture methods. The specific objectives of th is study were as follows: 1. To investigate the survival of S. sonnei on the tomato surfaces, in potato salad and in ground beef when held under st andard refrigerated conditions. 2. To investigate chromosomally-located gene tic targets for the sp ecific detection of all Shigella spp. or only S. sonnei 3. To develop and compare flow-through imm unocapture, using the Pathatrix, for the detection of S. sonnei on the tomato surfaces, in potato salad and in ground beef with the BAM Shigella culture method. 4. To develop a nucleotide sequence capture method for the rapid detection of S. sonnei on the tomato surfaces, in pot ato salad and in ground beef.

PAGE 19

6 CHAPTER 2 LITERATURE REVIEW Shigella as a Foodborne Pathogen Background Information Shigella spp. are the causative agents of shigellosis, or “bacillary dysentery,” first discovered over 100 years ago by Kiyoshi Sh iga, a Japanese scientist (Anonymous, 2002). Shigellae are members of the bacterial family Enterobacteriaceae and are nearly identical genetically to Escherichia coli and are also closely related to Salmonella and Citrobacter spp. (Lampel, 2001). Shigellae are characterized as Gram-negative, facultatively anaerobic, non-sporeforming, non-mo tile rods that typically do not ferment lactose. In addition, they are lysine-decar boxylase, acetate, and mucate negative and do not produce gas from glucose, although some S. flexneri six serotypes have been reported to produce gas (Echeverria et al ., 1991; International Co mmission on Microbiological Specifications for Foods (ICMSF), 1 996). There are four serogroups of Shigella : S. dysenteriae (serogroup A; 15 serotypes), S. flexneri (serogroup B; eight serotypes divided into 11 subserotypes), S. boydii (serogroup C; 20 serotypes), and S. sonnei (serogroup D; one serotype) (Centers for Disease Co ntrol and Prevention (CDC), 2004). Shigella has been classically ch aracterized as a waterborne pathogen (Smith, 1987) and outbreaks have been reported from contamin ated community water sources that were unor under-chlorinated (Blostein, 1991; Fleming et al ., 2000; CDC, 2001). Foodborne outbreaks of Shigella are also common, especially with foods that are subjected to processing or preparation by hand, are exposed to a limited heat treatment, or are

PAGE 20

7 served/delivered raw to the consumer (Wu et al. 2000). Some examples of food products from which Shigella spp. have been isolated are potato salad, ground beef, bean dip, raw oysters, fish, and raw vegetables. The infective dose for Shigella is very low: 10 cells of S. dysenteriae to 500 cells of S. sonnei (Kothary and Babu, 2001). At-risk popul ations, such as the very young, very old, or persons with decreased immune func tion, may be more sus ceptible to infection. Due to the low infective dose of Shigella person-to-person transmission is common, especially in daycare settings ( S. sonnei ) where toddlers commonly practice poor personal hygiene. Typical symptoms of in fection include bloody diarrhea, abdominal pain, fever, and malaise. Although the m echanism is unknown, seizures have been reported in 5.4% of shigellosis cas es involving children (Galanakis et al ., 2002). Chronic sequelae from S. dysenteriae serotype 1 infections can include hemolytic uremic syndrome (HUS), while S. flexneri infections are associated with later development of reactive arthritis, especially in persons with the genetic marker HLA-B27 (CDC, 2006a). Reactive arthritis is characterized by joint pain, eye irritation, and painful urination (CDC, 2006a). Epidemiology of Shigella The four serogroups of Shigella differ in epidemiology (Ingersoll et al ., 2002). S. dysenteriae is primarily associated with epidemics (Ingersoll et al ., 2002) with serotype 1 associated with the highest fa tality rate (5-15%) (CDC, 2003). S. flexneri predominates in areas of endemic infection, while S. sonnei has been implicated in source outbreaks in developed countries (Hale, 1991). S. boydii has been associated with source outbreaks in Central and South America but is most commonly restricted to the Indian subcontinent. S.

PAGE 21

8 boydii is rarely isolated in North America; how ever slight increases in the numbers of isolates have been observed in both 2003 and 2004 (CDC, 2004). According to the CDC Emerging Infections Program, Foodborne Diseases Active Surveillance Network (FoodNet), Shigella was the third most reported foodborne bacterial pathogen in 2005 (CDC, 2006b) Of 16,614 laboratory-diagnosed cases, Shigella accounted for 2,078 cases (12.5% of total cases) behind only Salmonella (6,471 cases) and Campylobacter (5,655 cases) (CDC, 2006b). From 1996-1998 to 2005 the estimated annual incidence of Shigella spp. in the U.S. decreased by 43% (CDC, 2006b). While S. sonnei continued to be the most isolated serogroup in the U.S. in 2004, the rate of isolation has declined while the rate of isolation of S. flexneri and S. boydii have increased slightly (Table 2-1). Demographic Variability of Infections with Shigella spp. FoodNet data on shigellosis in the U.S. collected from 1996 to 1999 were recently analyzed for trends in dem ographic variability (Shiferaw et al ., 2004). The overall incidence of shigellosis wa s highest among the following groups: children aged 1-4 years, male patients, blacks, Hispanics and Native Americans (Shiferaw et al ., 2004). There were also marked demographic differences between infection with S. sonnei and S. flexneri with respect to age, sex and race. While the incidence of both S. sonnei and S. flexneri were higher among those aged 1-4 ye ars, there was a second peak of S. flexneri infection among those aged 30-39 years (Shiferaw et al ., 2004). The incidence of S. sonnei among men and women were simila r; however the incidence of S. flexneri among men was almost twice th at of women (Shiferaw et al ., 2004). In addition, the incidence of S. sonnei among blacks and whites was higher than that of S. flexneri while the incidence of S. flexneri was higher among Native Americans (Shiferaw et al ., 2004).

PAGE 22

9 Table 2-1. Percentage of Shigella isolates in the United States reported by PHLIS in recent years. Serogroup 2002 2003 2004 S. sonnei 83.5% 80.2% 68.9% S. flexneri 12.2% 14.4% 17.2% S. boydii 0.8% 1.1% 1.8% S. dysenteriae 0.3% 0.4% 0.4% Ungrouped 3.2% 3.9% 11.7% Total isolates 12,992 11,552 9,343 (CDC, 2002; CDC, 2003; CDC, 2004) Recently Identified Serotype of S. dysenteriae During 2001 to 2003, six biochemically, sero logically and genetically identical Shigella strains were isolated in geographically distant locations in Canada. When analyzed biochemically, the suspect strains di splayed reactions consis tent with that of Shigella spp. (Melito et al ., 2005). When analyzed serol ogically, the suspect strains produced weak reactions with S. dysenteriae serovars 4 and 16 and E. coli O159 and O173 antisera, however antisera prepared from one of the suspect isolates was completely absorbed by antigens from S. dysenteriae serotype 4 and E. coli O159 (Melito et al ., 2005). In addition, all six stra ins tested PCR positive for the ipaH gene and the invasion associated locus. Mol ecular typing by PCR-RFLP of the rfb gene produced a S. dysenteriae serovar 2 and E. coli 0112ac pattern (Melito et al ., 2005). Based on these analyses, the authors proposed the six susp ect isolates represen ted a novel serovar of S. dysenteriae Foodborne Outbreaks Involving Shigella A summary of selected recent foodborne sh igellosis outbreaks is given in Table 22. Characteristic of foodborne shigellosis, seve ral recent outbreaks have been associated with foods consumed raw (Martin et al ., 1986; Fredlund et al ., 1987; Davis et al ., 1988;

PAGE 23

10 Table 2-2. Selected f oodborne outbreaks involving Shigella Year Serogroup Food Product(s) Implicated Reference(s) 1983 S. sonnei Tossed salad Martin et al ., 1986 1986 S. sonnei Shredded lettuce Davis et al ., 1988 1986 S. sonnei Raw oysters Reeve et al ., 1989 1987 S. sonnei Watermelon Fredlund et al ., 1987 1988 S. sonnei Uncooked tofu salad Yagupsky et al ., 1991; Lee et al ., 1991 1989 S. flexneri 4a German potato salad Lew et al ., 1991 1992 S. flexneri 2 Tossed salad Dunn et al ., 1995 1994 S. sonnei Iceberg lettuce Long et al 2002; Kapperud et al ., 1995; Frost et al ., 1995 1995-6 S. sonnei Fresh pasteurized milk cheese Garca-Fulgueiras et al ., 2001 1996 S. flexneri Salad vegetables PHLS, 1997 1998 S. sonnei Uncooked, chopped curly parsley CDC, 1999 1998 S. flexneri Restaurant-associated, source unknown Trevejo et al ., 1999 1999 S. boydii 18 Bean salad (parsley or cilantro) CDPH, 1999 2000 S. sonnei Five layer bean dip CDC, 2000; Kimura et al ., 2004 2001 S. sonnei Raw oysters Terajima et al ., 2004 2002 Shigella spp. Greek-style pasta salad TPH, 2002 Cook et al ., 1995; Dunn et al ., 1995; Frost et al ., 1995; Kapperud et al ., 1995; Public Health Laboratory Service (PHLS), 1997; CDC, 1999) and processed or prepared by hand (Martin et al ., 1986; Lee et al ., 1991; Lew et al ., 1991; Yagupsky et al ., 1991; Dunn et al ., 1995; Chicago Department of Pub lic Health (CDPH), 1999; Trevejo et al ., 1999; Toronto Public Health (TPH), 2002). In 2000, a multi-state outbreak of shigellosis was traced to a commercially prepared five-layer bean dip (Kimura et al ., 2004). This outbreak involved 406 persons (14 hospitaliz ations, 0 deaths) across 10 states. After extensive epidemiological i nvestigation, numerous problem s were identified in the manufacturing process and investigators dete rmined that the source of the outbreak was most likely an infected food-handler (Kimura et al ., 2004). This outbreak demonstrates

PAGE 24

11 the vulnerability of our food supply to point-source contamination with Shigella followed by wide distribution and subsequent infection of many consumers (Kimura et al ., 2004). Prevalence of Shigella : Food and Food Handlers Despite the high incidence of shigellosis, there is limited data on the prevalence of Shigella among food handlers or on food products. A study investigating the presence of enteropathogens among food handler s in Irbid, Jordan, isolated Shigella from the stools of four out of 283 examined food handlers (al-Lahham et al ., 1990). Mensah et al (2002) evaluated 511 food items from the str eets of Accra, Ghana, from which S. sonnei was isolated from one sample of macaroni. It wa s noted that the macaroni was served using bare hands instead of clean uten sils, which may have led to the S. sonnei contamination. Wood et al (1983) examined foods from Mexican homes, commercial sources in Guadalajara, Mexico, and from restaurant s in Houston, TX, for contamination with bacterial enteropathogens. While no Shigella was isolated from foods sampled from 12 Houston restaurants or from food commercially prepared in Guadalajara, Mexico, four isolates were obtained from meals prepared in Mexican homes. These studies demonstrate the importance of proper food ha ndling and the role food handlers in the transmission of Shigella In response to President Clinton’s Nationa l Food Safety Initiative (January 1997) and Produce & Imported Foods Safety Initiati ve (October 1997), the U.S. Food and Drug Administration (FDA) has investigated the presence of foodborne pa thogens, including Shigella on imported and domestic produce (FDA, 2001a; FDA, 2001b; FDA, 2003). An FDA survey of imported broccoli, cantaloupe celery, parsley, s callions, loose-leaf lettuce, and tomatoes found Shigella contamination in nine of 671 total samples: three of 151 cantaloupe samples, two of 84 celery sample s, one of 116 lettuce samples, one of 84

PAGE 25

12 parsley samples, and two of 180 scallion sa mples (FDA, 2001a). Another FDA survey of domestically grown fresh cantaloupe, celer y, scallions, parsley and tomatoes found Shigella contamination in five of 665 total samp les: one of 164 cantaloupe samples, three of 93 scallion samples, and one of 90 parsley samples (FDA, 2003). An additional survey of imported produce was conducted; however at the time of th is publication results were not publicly available (FDA, 2001b). Survival of Shigella Environmental Factors on Survival of Shigella Shigella spp. are heat sensitive, acid resistant, salt tolerant bacteria that can withstand low levels of organic acids (Z aika, 2001; Zaika, 2002a; Zaika 2002b). Zaika (2001) studied the survival of S. flexneri strain 5348 in brain hear t infusion (BHI) broth as a function of pH (2 to 5) and temperature (4 to 37C). When inoc ulated into BHI broth adjusted to pH 5, S. flexneri demonstrated growth when held at 19, 28, and 37C, while counts declined over time at temperature of 12C or lower. When inoculated into BHI broth adjusted to pH 2, 3, or 4, inoculated S. flexneri counts declined over time at all temperatures tested. S. flexneri in BHI broth adjusted to pH 2 reached undetectable levels in 1 to 3 days when held at temper ature of 19C or lower. In general, S. flexneri survival was greater in BHI broth incubated at lower te mperatures and adjusted to higher pH. This study suggests that S. flexneri is acid resistan t and that acidic foods may support the survival of Shigella over a long period of time (Zaika, 2001). Zaika (2002a) studied the survival of S. flexneri strain 5348 in BHI broth (pH 4 to 6) containing 0.5 to 8% NaCl In BHI adjusted to pH 6, S. flexneri grew in the presence of < 6% NaCl when held at 19 and 37 C, and in the presence of < 7% NaCl when held at 28C. Growth of S. flexneri was also observed in BHI brot h adjusted to pH 5 containing

PAGE 26

13 < 2, < 4, < 4, < 0.5% NaCl when held at 37, 28, 19, a nd 12C, respectively (Zaika, 2002a). S. flexneri populations gradually declined in BH I adjusted to pH 4 at all incubation temperatures and all levels of NaCl tested (Zaika, 2002a). Results from this study suggest that S. flexneri is salt tolerant and may survive in salty foods such as pickled vegetables, caviar, pickled herring, dry cured ham, and cer tain cheeses for extended periods of time (Zaika, 2002a). S. flexneri survival was also studied in BHI broth supplemented with organic acids commonly found in fruits and vegetables (citri c, malic, and tartaric acid) or fermentation acids commonly used as preservatives (acetic and lactic acid) at 0.04M and adjusted to pH 4 with HCl or NaOH (Zaika, 2002b). Fermen tation acids (acetic and lactic acid) had a greater effect on survival th an citric, malic, and tartaric acids (Zaika, 2002b). When incubated at 37C, S. flexneri survived for 1 to 2 days in the presence of each organic acid tested. At 4 C, S. flexneri survived in the presence of a ll the organic acids tested for longer than 55 days (Zaika, 2002b). This study s uggests that organic acids may aid in the inactivation of Shigella however foods with low leve ls of acids stored at low temperatures may support the survival of the bacterium for extended periods of time (Zaika, 2002b). Temperature is an important factor in su rvival of shigellae. Freezing (-20C) and refrigeration (4C) temperatures su pport survival, but not growth, of Shigella (International Commission on Microbiological Specifications for Foods (ICMSF), 1996). When studied in nutrient broth, the observed temperature ranges which permitted growth for S. sonnei and S. flexneri were 6.1 to 47.1C and 7.9 to 45.2C, respectively (cited in ICMSF, 1996); however Shigella can survive for extended peri ods of time when stored at

PAGE 27

14 -20C or at 4C. Elevated temp eratures are less permissive for Shigella survival, and traditional pasteurization and cooking temperat ures are sufficient for inactivation. Evans et al (1970) calculated the decima l reduction time (D-value) of S. dysenteriae in pasteurized whole milk to be 0.0008 sec at 82.2 C. When studied in nutrient broth, most strains of S. sonnei and S. flexneri were inactivated within 5 min at 63C (cited in ICMSF, 1996). Sublethal heat exposure can sensitize Shigella to selective components of microbiological media. Tollison and J ohnson (1985) demonstrated that S. flexneri sublethally heat-stressed by exposure to 50.0C for 30 minutes in phosphate buffer became sensitive to 0.85% bile salts and 0.50% sodium desoxycholate. Since these compounds are ingredients in several enrichme nt and isolation media used for detection of Shigella the thermal history of the food sample to be analyzed should be known. Survival of Shigella on Fomites Inanimate objects, or fomites, can serve as vectors for transmission of Shigella and there have been several re ports on the survival of Shigella on various surfaces (Spicer, 1959; Nakamura, 1962; Islam et al ., 2001). Spicer (1959) stud ied the survival of S. sonnei dried on cotton threads at room temperat ure and under refrigeration at various levels of relative humidity. In general, surviv al was better at refrigerated temperatures and at high (84%) and low (0 %) relative humidity (RH). S. sonnei remained detectable on the cotton threads after 12 days at 5-10C (84% RH) (Spicer, 1959). Similar results were observed using 10 different strains of S. sonnei inoculated on cotton, glass, wood, paper, and metal at various temperatures (-20C to 45C) (Nakamura, 1962). When held at 20C, most of the strains survived for more than 14 days on each surface, however surfaces held at 45C did not support surv ival of most strains (Nakamura, 1962). Investigations on the survival of S. dysenteriae serotype 1 on cloth, wood, plastic,

PAGE 28

15 aluminum, and glass objects suggest that 1.5 to 4 hours post-inoculation, S. dysenteriae serotype 1 enters a viable but non-culturable (VBNC) state (Islam et al ., 2001). While no S. dysenteriae serotype 1 could be recovered af ter 5 days by conventional culture methods, viable cells could be observed using fluorescent antibody techniques (Islam et al ., 2001). Whether or not Shigella is able to achieve a true VB NC state or if these results demonstrate the inadequacy of pl ating media in recovering viable Shigella has not been fully investigated. Nevertheless, these studies demonstrate that fomites can sustain viable Shigella for an extended time and serve as vehi cles in transmission of the pathogen. Survival of Shigella in Food and Water Shigella can survive in water w ith little decline in populat ion. Rafii and Lunsford (1997) inoculated distilled water with S. flexneri at 2.8 x 108 colony forming units (CFU)/ml and held the samples at 4C. After 26 days, 9.2 x 107 CFU/ml of S. flexneri survived. This high survival rate of S. flexneri in water supports the historical association of shigellosis outbreak s and water sources. Fruits and vegetables can suppor t the growth or survival of Shigella Escartin et al (1989) artificially contaminated fresh cut papaya, jicama, and watermelon with S. sonnei S. flexneri or S. dysenteriae and within 6 hours at room temperature, growth was observed. Rafii and Lunsford (1997) inocul ated raw cabbage, onion, and green pepper with S. flexneri and although counts decrea sed slightly at 4C, survival was observed after 12 days on onion and green pepper, at which time sampling was terminated due to spoilage (Rafii and Lunsford, 1997). S. flexneri survival was observed on the cabbage for 26 days. Wu et al (2000) studied survival of S. sonnei on whole and chopped parsley leaves. At 21 C, growth on chopped parsley was observe d at a similar rate to that in

PAGE 29

16 nutrient broth (Wu et al ., 2000). At 4 C, populations declined on both chopped and whole parsley throughout the 14 -day storage period, however S. sonnei survived regardless of initial population (Wu et al ., 2000). These studies demonstrate Shigella survival on refrigerated produce for periods of time that exceed expected shelf life. Low pH foods can support survival of Shigella when held at refrigerated temperatures. Bagamboula et al (2002) demonstrated S. sonnei and S. flexneri survival in apple juice (pH 3.3 to 3.4) and tomato juice (pH 3.9 to 4.1) held at 7 C for 14 days. No reduction was observed in the tomato juice, while a 1.2 to 3.1 log10 CFU reduction was observed in apple juice over the 14 day study. Rafii and Lunsford (1997) observed S. flexneri survival in carrot salad (p H 2.7 to 2.9), potato salad (pH 3.3 to 4.4), coleslaw (pH 4.1 to 4.2), and crab salad (pH 4.4 to 4.5) held at 4 C. Sampling was terminated at day 11 for the carrot and the potato salads, at which time S. flexneri counts decreased from 4.3 x 106 to 4.2 x 102 CFU/g and from 1.32 x 106 to 8.5 x 102 CFU/g, respectively. Sampling of the coleslaw and the crab salads ceased due to product spoilage on days 13 and 20, respectively, however S. flexneri survived at levels of 2.16 x 104 and 2.4 x 105 CFU/g, respectively. These studies indicate that Shigella survived at refrigerated temperatures despite the presence of background microflora and low pH. Prepared foods can also support the survival of Shigella Islam et al (1993b) investigated the grow th and survival of S. flexneri in boiled rice, lentil soup, milk, cooked beef, cooked fish, mashed potato, mashed brin jal, and raw cucumber. All food samples, except raw cucumber, were autoclaved prior to inoculation. Ten gram or 10 ml samples of each food were inoculated with 105 cells of S. flexneri incubated at 5, 25, or 37C and sampled over 72 hr. All of the foods tested supported growth up to 108 to1010 cells per g

PAGE 30

17 or ml within 6 to 18 hr after inoculation at 25C and 37C (Islam et al ., 1993b). Initial inoculum levels were maintained throughout the 72 hr holding period for all foods, except rice and milk. S. flexneri counts in rice decreased by approximately 1 log after 72 hr at 5C, whereas counts in milk dropped afte r 48 hr but then retu rned to the initial inoculum level by 72 hr. These resu lts demonstrate the ability of Shigella to grow and survive in a variety of prep ared foods that may be contaminated by an infected food handler. Shigella is able to survive on produce p ackaged under vacuum or modified atmosphere. Satchell et al (1990) investigated the survival of S. sonnei in shredded cabbage packaged under vacuum or in a m odified atmosphere of nitrogen and carbon dioxide when stored at room temperature or under refrigeration. When test samples were stored at room temperature, counts of inoculated S. sonnei remained at level for up to three days, at which time populations began to drop. When test samples were stored under refrigeration, S. sonnei counts did not drop after seve n days. In the latter study, refrigerated samples maintained a constant pH throughout the study, while samples stored at room temperature had a drop in pH that may have contributed to the decline in S. sonnei cell numbers. Current Understanding of Shigella Pathogenesis This discussion will be limited to an overview of epithelial cell invasion by Shigella potential roles of th e IpaH effector proteins and blockage of autophagy by IcsB. The reader is directed to the following reference for a more complete review of Shigella pathogenesis and the toxins produced by Shigella spp.: Warren et al ., 2006.

PAGE 31

18 Shigella Invasion of Epithelial Cells The invasion of the local epithelium of th e colon (large intestine) is presented in Figure 2-1. Once ingested, shigellae move throu gh the gastrointestinal tract to the colon, where they translocate the epith elial barrier via M cells that overlay the solitary lymphoid nodules (Suzuki and Sasakawa, 2001). Upon r eaching the under side of the M cells, Shigella infect macrophages and induce cell a poptosis (Suzuki and Sasakawa, 2001). Once released from the macrophage, Shigella enters neighbori ng epithelial cells. Shigella first forms a membrane bound protrusion into the adjacent cell. This protrusion must distend two membranes: one from the donor cell, and another from the recipient cell (Parsot and Sansonetti, 1996). As the protrusion pushes further in to the recipien t cell, it is taken up by the recipient cell resulting in th e bacteria enclosed in a double-membrane vacuole (Monack and Theriot, 2001). Inte rcellular spread is completed when Shigella escape from the double-membrane vacuole, rel easing it into the cytosol of the recipient cell. In response to invasion, epithelial cell s produce pro-inflammatory cytokines that contribute to inflammation of the colon (Suzuki and Sasakawa, 2001). Potential Roles of the IpaH Effector Proteins IpaH7.8 facilitates escape from endocytic vacuoles Fernandez-Prada et al (2000) reported that IpaH7.8 of S. flexneri was required for efficient escape from endocytic vacu oles. Human monocyte-derived macrophages (HMDM) and the J744 mouse macrophage cell line were infected with S. flexneri 2457T and pWR700, an ipaH 7.8 deletion mutant of S. flexneri 2457T. After the infected HMDM and J744 cells were incubated in the presence of gentamicin and chloroquine, results showed that more pWR700 than 2457T was present within endocytic vacuoles, suggesting that IpaH7.8 is requ ired for escape from the vacu ole. The contrast between

PAGE 32

19 Figure 2-1. Invasion of epithelial cells by Shigella spp. In the large intestine, Shigella enters through M cells that overlay th e solitary lymphoid nodules, infect the resident macrophage and induce cell apoptosis. Once released from the macrophage, Shigella enters the neighbor ing enterocytes and escape from the double membrane vacuole that encompasses them. Shigella multiply in the cytoplasm of the host cell and polymerize actin for motility. IcsB is required to evade autophagic recognition; therefore icsB mutants are degraded once they escape from the vacuole. Figur e reproduced from Ogawa and Sasakawa, 2006. pWR700 and 2457T localization within endocy tic vacuoles was more pronounced in the J744 cell line. One explanation for this was th at the HMDM in tissue culture represented a heterogenous population of cells, at vari ous stages of differentiation. The authors further suggested that the ipaH genes may play a bigger role in monocytes than macrophages (Fernandez-Prada et al ., 2000). It is noteworthy to mention that ipaH 4.5 and ipaH 9.8 mutants behaved like the wild-type 2457T in both HMDM and J744 cells, suggesting their role in viru lence differs from that of ipaH 7.8. Subversion of host cell signaling by IpaH9.8 Toyotome et al. (2002) investigated the secre tion of IpaH proteins from S. flexneri in broth cultures and determined that IpaH proteins are exported by type III secretion

PAGE 33

20 after entry into the host cell. Further investigation showed that once secreted, IpaH9.8 accumulates in the nucleus, while small amount s are present in the cytoplasm. IpaH9.8 has similar structure to the Salmonella Typhimurium protein SspH1, which belongs to the bacterial LPX repeat protein family. Upon inf ection of the local ep ithelium, intracellular pathogens, such as Shigella and Salmonella elicit the secretio n of proinflammatory cytokines, such as interleukin 8 (IL-8) (Har aga and Miller, 2003). Pr oduction of IL-8 and other cytokines in response to bacterial invasion are dependan t, in part, on activation of transcription factor NF-kappa B. Haraga a nd Miller (2003) demonstr ated that SspH1 and IpaH9.8 both localize to the ma mmalian nucleus and inhibit nu clear factor kappa B (NFkappa B)-dependent gene expression (Har aga and Miller, 2003). In this way, IpaH9.8 serves to subvert host cell si gnaling events involved in th e immune response to epithelial invasion. Blockage of Autophagy by IcsB Autophagy is a critical process in eukar yotic cells in which undesirable cellular components or organelles, including invadi ng microbes, are degraded. Recently, Ogawa et al (2005) identified IcsB as critical in th e camouflage against au tophagic recognition. IcsB is an effector protein exported by t ype III secretion and is located on the cell surface. IcsB mutants are fully invasive and capable of escaping from the vacuole, but defective in its ability to multip ly within the host cell (Ogawa et al ., 2003). In the absence of IcsB, the autophagy protein Atg5 recognizes and binds to IcsA (VirG), thus initiating autophagosome formation. Ogawa et al (2005) demonstrated the IcsA (VirG) binding region for both Atg5 and IcsB is the same, a nd Atg5 binding to IcsA (VirG) is inhibited by IcsB in a dose-dependent manner. By bloc king the binding if IcsA (VirG) by Atg5,

PAGE 34

21 IcsB inhibits the aut ophagic recognition of Shigella within the host cell cytoplasm, thus contributing to intracellular survival. Genetic Relationship Between Shigella and Escherichia coli While it has been generally accepted that Shigella are within the species E. coli recent studies have indicated that Shigella like the other forms of pathogenic E. coli derived from different evolutionary origin s, suggesting converge nt evolution of the Shigella phenotype (Pupo et al ., 2000). Rolland et al (1998) used restriction fragment length polymorphism of rDNA (ri botyping) to group 75 strains of Shigella 13 strains of enteroinvasive E. coli (EIEC) and 72 E. coli strains of the E. coli Reference (ECOR) Collection, which have been classified into four phylogenic groups (A, B1, B2 and D). The S. sonnei S. flexneri and most S. dysenteriae ribotypes were closely related to phylogenic group D, while S. dysenteriae serotype 1 strains we re closely related to phylogenic group B1 and S. boydii strains were spread between phylogenic group D and B1 (Rolland et al ., 1998). In contrast, the ribotypes of EIEC strains were widely distributed among phylogenic groups A, B1 and B2. This evidence suggests that Shigella and EIEC derived from different origins. Pupo et al (2000) sequenced eight housekeeping genes from four regions of the chromosome for 46 strains of Shigella representing all four se rotypes. Three distinct clusters of Shigella were identified and although S. sonnei and S. dysenteriae serotype 1, 8 and 10 did not group in the main three cl usters, they fell well within the species E. coli (Pupo et al, 2000). As with the study by Rolland et al (1998), S. boydii serotype 13 was distantly related to the other Shigella strains. Cluster 1 contained most of the S. boydii and S. dysenteriae strains along with S. flexneri serotypes 6 and 6A. Cluster 2 contained seven S. boydii strains and S. dysenteriae serotype 2. Cluster 3 contained S. flexneri

PAGE 35

22 serotypes 1-5 and S. boydii serotype 12. Unlike the results from ribotyping, the use of multiple genes for phylogenic analysis reve aled greater genetic diversity among the strains of Shigella further suggesting that Shigella derived from differe nt evolutionary origins. Fukiya et al (2004) used comparative genomic hybridization microarray analysis to compare the gene content of E. coli K-12 with that of 22 pathogenic E. coli and Shigella strains. When compared to the E. coli K-12 genome, the genomes of S. sonnei S. boydii and S. flexneri 2a were missing only 613, 533 and 716 open reading frames (ORFs). The genomes of the other pathogenic E. coli strains were missing similar numbers of ORFs. Subsequent phylogenic analys is revealed a close relationship between three of four EIEC strain s and the three strains of Shigella which suggests EIEC and Shigella form a single E. coli pathovar (Fukiya et al 2004; Yang et al 2005). Providing further to the body of evidence that Shigella and EIEC derived from different origins, the complete genomes of S. boydii serotype 4 (strain 227), S. dysenteriae serotype 1 (strain 197) and S. sonnei (strain 046) have recently been sequenced (Yang et al ., 2005). Comparative genomics among the newly sequenced Shigella genomes and the previous ly sequenced genomes of S. flexneri 2a (strain 301) and E. coli K-12 (strain MG1655) supported previous work by Fukiya et al (2004). While the genomes of Shigella share most of their genes with that of E. coli K-12, the Shigella phenotype is a result of the gain and loss of functions through bacteriophagemediated gene acquisition, insertion se quence (IS)-mediated DNA rearrangements and formation of pseudogenes (Yang et al ., 2005). For example, the chromosome and virulence plasmid of S. sonnei strain 046 contained 327 and 28 intact IS elements and 67

PAGE 36

23 and 68 partial IS elements, re spectively. In contrast, the E. coli K-12 genome contained only 37 intact IS elements and seven partial IS elements. These studies taken together, demonstrate that the Shigella have evolved from distinct E. coli ancestors through convergent evolution. Detection Methods for Shigella in Foods Conventional Culture Methods for Shigella Traditional microbiological media fo r enrichment and isolation of Shigella Traditional microbiological techniques make use of selective and differential media for the enrichment and isolation of Shigella Many variants of enrichment and plating media have been investigated for optimal recovery, often with conflicting results. Although Shigella is readily isolated from clinical samples, food samples are more problematic. Isolation of Shigella from food samples can be inhibited by indigenous microflora, especially the coliform bacteria and Proteus spp. (ICMSF, 1996). The addition of the antibiotic novobiocin to liquid and solid media has been shown to improve the isolation of S. flexneri and S. sonnei from investigated foods (cited in ICMSF, 1996). Contamination of food products with Shigella results primarily through a food handler with poor personal hygiene; ther efore the concentration of Shigella may be very low compared to that of the indigenous microf lora (Lampel and Maurelli, 2001). Currently, selective media are not available that ad equately suppress the growth of background microflora, therefore Shigella is often overgrown by comp eting microorganisms (Lampel and Maurelli, 2001). More research is needed to determine more appropriate selective media and enrichment conditi ons for the is olation of Shigella from food samples. Two enrichment broths initially used for the isolation of Shigella were Selenite-F (SF) and Tetrathionate (TT) broth. These broths were orig inally designed for the isolation

PAGE 37

24 of Salmonellae, but due to the lack of specific enrichment media for shigellae they were used as all-purpose enteric enrichment br oths (Taylor and Schelhart, 1969). Sodium selenite, although selective for salmonellae, is toxic to Shigella (and most enterics); therefore it is no longer used in enrichment procedures for Shigella TT is a peptonebased broth with bile salts and sodium thiosu lfate that inhibits growth of most Grampositives and Enterobacteriaceae While TT is routinely used for the enrichment of Salmonella, it is rarely used for Shigella Gram-negative (GN) broth is a peptone-based broth with glucose and mannitol. The concen tration of mannitol in GN broth is higher than glucose to promote mannitol fermentors, like Shigella Both TT and GN broths contain bile salts, which can be inhibitory to stressed cultures (Tollison and Johnson, 1985). Furthermore, GN broth contains sodium desoxycholate, which has been shown to inhibit heat-stressed shigellae (Uyttendaele et al ., 2001). Regardless, GN broth is listed as an alternate enrichment me dium for the detection of Shigella from food by some standard methods (Lampel, 2001). Current enrichment procedures (FDA, 1998; Lampel, 2001, Health Canada, 2004) use a low carbohydrate medium, Shigella broth (SB) with addition of novobiocin, for the detection/isolation of Shigella One study reported that acids produced by other Enterobacteriaceae during the fermentation of carbohydr ates were toxic to shigellae (Mehlman et al ., 1985); however, other stud ies have shown that Shigella can grow at pH 4.5 to 4.75 (Bagamboula et al ., 2002) and survive at pH 4.0 (Zaika, 2002). Nevertheless, the use of SB limits the production of acids, and thereby limits the introduction of low pH, during enrichment. SB is also less st ringent than TT broth and GN broth for the enrichment of Shigella since it contains neither bile salts nor sodium desoxycholate. A

PAGE 38

25 recent study investigated SB, GN broth, tryptic soy broth, and Enterobacteriaceae Enrichment (EE) broth with the addition of novobiocin for enrichment/detection of shigellae (Uyttendaele et al ., 2001). When incubated in GN broth, S. sonnei was unable to grow to comparable levels as observed in SB and EE broths. EE broth, however, has been reported inhibitory to S. boydii (Warren, 2003; Warren et al ., 2005b). Multiple plating media with differing sele ctivity can be used to increase the chances of Shigella isolation. The most common low se lectivity medium used for plating Shigella is MacConkey Agar (MAC). Eosin methyl ene blue (EMB) or Tergitol-7 (T7) agars can also be used. Since different iation on MAC is solely based on lactose fermentation, Shigella colonies look similar to th ose of many lactose negative competitors (Uyttendaele et al ., 2001). On MAC, Shigella colonies are translucent or slightly pink, with or without rough edges. Shigella produce colorless colonies on EMB and bluish colonies on the yellowi sh-green T7 agar (Lampel, 2001). Intermediate selectivity media useful in isolating Shigella are desoxycholate citrate agar (DCA) and xylose lysine desoxycholate agar (XLD). Shigella spp. produce colorless colonies on both DCA and XLD. Bhat and Rajan (1975) repo rted XLD superior to DCA for the isolation of Shigella since DCA required a 48 hr in cubation to show clear colony morphology as opposed to overnight incuba tion for XLD. Unfort unately, D-xylose, which serves as a differentiating agent on XLD agar, is fermented by some strains of S. boydii while most Shigella cannot ferment xylose (Bhat and Rajan, 1975). Thus some strains of Shigella will be missed if XLD is used as the sole plating medium. Highly selective media for Shigella spp. include Salmonella-Shigella agar (SSA) and Hektoen Enteric agar (HEA). A problem a ssociated with SSA and HEA is that they

PAGE 39

26 may be too stringent for some strains of Shigella especially if the culture is stressed (Lampel, 2001; Uyttendaele et al ., 2001). Shigella spp. produce colorl ess, translucent colonies on SSA and green colonies on HEA. A newly developed plati ng medium, Chromogenic Shigella spp. Plating Medium (CSPM; R&F Laboratories, West Chicago, IL), offers medium selectivity (bile salts, antibiotic supplementation) with an altern ative to differentiation based on lactose fermentation. Instead, differentiation on C SPM is based on proprietary components consisting of select carbohydrates, pH indicat ors, and chromogens (Dr. Larry Restaino, personal communication). Shigella spp. produce white to clear colonies on CSPM while competitors produce colored colonies. CSPM has been compared to MAC and SSA for the recovery of S. boydii and S. sonnei from tomato surfaces with no significant differences in recovery observed (Warren, 2003; Warren et al ., 2005a). Further evaluation of CSPM against other strains of S. boydii and S. sonnei as well as the other serogroups of Shigella is needed. Recent studies at the FDA, Laboratory for Enteric and Sexually Transmitted Diseases have demonstrated stable lactos e-positive mutations in stationary phase S. sonnei (Dr. D.J. Kopecko, FDA, personal co mmunication). DNA sequencing experiments have revealed slip-strand mutations within the lac repressor ( lacI ) that are responsible for the lactose-positive phenotype. These mutations are significant for detection of S. sonnei in food since typical coloni es on plating media (MAC) are differentiated based on the utilization of lactose, the typical S. sonnei phenotype being lactose-negative.

PAGE 40

27 The FDA Bacteriological Analytical Manual culture method for detection of Shigella in foods The FDA Bacteriological Analytical Manual (BAM) culture method for the isolation and detection of Shigella spp. from food utilizes a combination of low carbohydrate enrichment, anaerobic conditions and elevated temperature (FDA, 1998). Briefly, a 25 g sample of the food pr oduct is transferre d to 225 ml of Shigella broth (SB) to which novobiocin (0.5 g/ml for S. sonnei ; 3.0 g/ml for other Shigella spp.) has been added. Samples are held at room temperature for 10 min and periodically shaken. Sample supernatants are transferred to an Erlenmeyer flask and the pH adjusted to 7.0 0.2 with sterile 1 N NaOH or 1 N HCl. Flasks are in cubated anaerobically for 20 hr (44C for S. sonnei ; 42C for all other Shigella spp.) and the enrichments are streaked on MAC. Confirmation of suspicious coloni es involves tests for motility, H2S, gas formation, lysine decarboxylase, and fermentation of sucrose or lactose. Isolates negative for all confirmatory tests are further tested usi ng biochemical reactions including adonitol, inositol, lactose, potassium cyanide, malonate citrate, salicin, and methyl red. Shigellae are negative for all except methyl red. Antisera agglutination is then used to identify any culture displaying typical Shigella characteristics. June et al (1993) evaluated the effectivene ss of the BAM culture method for Shigella. Two strains of S. sonnei strains 9290 and 25931, were inoculated on potato salad, chicken salad, cooked shrimp salad, le ttuce, raw ground beef, and raw oysters. Using either unstressed or chilled stresse d cells, acceptable recovery was achieved for both strains from the potato salad, chicke n salad, cooked shrimp salad and lettuce samples, but not from the ground beef and ra w oyster samples. An approximate 4-log unit difference in recovery from ground beef samp les was observed between the two strains,

PAGE 41

28 suggesting high strain variability. Given the low infective dose of Shigella (as low as 10 cells), the BAM was considered ineffective for th e evaluation of raw ground beef and raw oysters. In 2002, the BAM culture method for Shigella was evaluated using two strains of unstressed, chill-stressed, or freeze-stressed S. sonnei (strains 357 and 20143) on selected types of produce (Jacobson et al ., 2002). Acceptable recovery of unstressed cells (<1.0 x 101 CFU/25g) and chill-stressed or freeze-stressed cells (<5.2 x 101 CFU/25g) were observed for all produce types tested (Jacobson et al ., 2002). More recently, similar tests of a modified BAM protocol showed unacceptable recovery of unstressed S. sonnei (patient isolate from an outbreak involving an unknown source) and S. boydii (patient isolate from an outbrea k involving bean salad) on tomatoes at 1.9 x 102 CFU/tomato and >5.3 x 105 CFU/tomato, respectively (Warre n, 2003). These results support the observation that signifi cant variation exists among strains of S. sonnei and demonstrate the importance of including the other serogroups of Shigella when evaluating culture methods for detection in food. Other culture methods for the detection of Shigella in foods Alternate culture methods for the detection of Shigella in foods can be found in Health Canada’s Compendium of Analyti cal Methods (Health Canada, 2004) and the American Public Health Association’s Comp endium of Methods for the Microbiological Examination of Foods (CMMEF) (Lampel, 2001). The Health Canada method is based on the BAM culture method with a few modifi cations; most notably enrichments for all Shigella spp. (including S. sonnei ) are supplemented with 0.5 g/ml novobiocin and incubated anaerobically at 42C. The CMMEF cu lture method is also similar to the BAM culture method, however the level of novobiocin in S. sonnei enrichment media is lower

PAGE 42

29 (0.3 g/ml) and enrichment conditions are ae robic at 37C. The CMMEF further suggests that two to three plates of va rious selective media be used to streak the enriched cultures: MAC for low selectivity, XLD for intermediate selectivity, and HEA for high selectivity. Confirmation of suspicious colonies is sim ilar to methods described above for the BAM culture method. Recently, the CMMEF culture method and an enrichment procedure involving EE broth have been compared to the BA M culture method for the detection of S. boydii and S. sonnei on tomato surfaces (Warren, 2003). Natu ral tomato microflora was found to have a great impact on recovery of S. sonnei and completely inhibited recovery of S. boydii in all three culture methods. No significant differences ( P > 0.05) were observed among the culture methods for detection of S. sonnei or between the BAM and CMMEF culture methods for the detection of S. boydii EE broth was found to be inhibitory to S. boydii These results suggest the need for more selective enrichment protocols for the evaluation of Shigella spp. in food. Immunological Methods for Shigella Detection Lipopolysaccharide (LPS) of Shigella Gram-negative bacterial LPS consists of three distinct regions: lipid A, core oligosaccharide, and a serotype-specific Opolysaccharide chain (O-antigen) (Neidhardt, 2004). The lipid A portion anchors the LPS molecu le to the outer membrane. The core oligosaccharide is composed of two regions: the inner core and the outer core. The inner core is common to many enterob acterial species and is composed of heptose and 2 keto-3 deoxyoctulosonate (KDO), while the outer core is rich in hexose and tends to be more species-specific (Tsang et al ., 1987). Studies of the core structure of S. flexneri have indicated that serotypes 1 to 5 and variants X and Y share the E. coli R3 core (Carlin and

PAGE 43

30 Figure 2-2. Structural differen ces between hexose regions of Shigella sonnei and Shigella flexneri lipopolysaccharides. A) S. sonnei and S. flexneri serotype 6 share the hexose region of the E. coli R1 core. B) S. flexneri serotypes 1-5 and X and Y variants have a hexose region identical to the E. coli R3 core. (Structures compiled from Jansson et al ., 1981). Abbr.: Gal = gala ctose, Glc = glucose, GlcNAc = 2-acetamido-N-deoxyglucose. Lindberg, 1986), while S. flexneri serotype 6 and S. sonnei share the E. coli R1 core (Gamian and Romanowska, 1982; Carlin and Lindberg, 1986; Viret et al ., 1992) (Figure 2-2). Antibodies have been developed w ith specificity to the inner core of Shigella spp. (Rahman and Stimson, 2001). The O-antigen is a polymer of re peating saccharides that is highly variable among species. Strains sharing identical O-antigen re peating units are of the same serotype. The O-antigens of all S. flexneri serotypes are a repeating tetrasaccharide (Carlin and Lindbe rg, 1986), while the O-antigen of S. sonnei is a repeating disaccharide (Kenne et al ., 1977) (Figure 2-3). Differences in S. sonnei form I and form II lipopolysaccharide Virulent S. sonnei produce smooth colonies, termed form I, which result from expression of the O-antigen. Unlike other Shigella species, the LPS genes ( rfc and rfb ) of S. sonnei are located on a large virulence plasmid, which can be spontaneously lost at

PAGE 44

31 Figure 2-3. O-antigen re peating subunits of S. sonnei and S. flexneri lipopolysaccharide. (A) The disaccharide re peating subunit of the S. sonnei O-antigen, (B) the tetrasaccharide repeating subunit of S. flexneri serotype 6, and (C) the tetrasaccharide repeating subunit of S. flexneri serotypes1-5 and X and Y variants. (Structures compiled from Carlin and Lindberg, 1986; Dmitriev et al ., 1979; and Gamian and Romanowska, 1982). Abbr: LAltNAcA = N -acetyl-L-altrosaminuronic acid, 4N -DFucNac = 2-acetamido-4-amino-2,4,6-trideoxy-D-galactose, GalA = galact uronic acid, GalNAc = N -acetylgalactosamine, Ac3Rha = 3O -acetylrhamnose, Rha = rhamnose, GlcNAc = 2-acetamido-Ndeoxyglucose. high frequency. Sansonetti et al (1981) investigated the stab ility of form I plasmids and observed 1 to 45% plasmid loss from re-s treaking form I colonies onto MAC and incubating 24 hr at 37C. When the large vi rulence plasmid is lo st, rough (avirulent) colonies, termed form II, are produced that express the Enterobacteriaceae R1 lipopolysaccharide core (Sansonetti et al ., 1981; Gamian and Romanowska 1982). A defective mutant of form II S. sonnei LPS, termed R-form, is characterized by an incomplete core region (Gamian and Romanowska, 1982) (Figure 2-4). As antibodies specific for S. sonnei O-antigen will bind form I, but not form II or R-form LPS, immunological detection methods with specificity for the O-antigen of Shigella are compromised when the virulence plasmid is lost. Immunological Detection Methods for Bacteria Immunological methods, such as latex a gglutination (LA), enzyme immunoassay (EIA), or immunomagnetic separation (IMS), ha ve been utilized for the detection of

PAGE 45

32 Figure 2-4. Lipopolysaccharides of S. sonnei A) Form I, B) form II, and C) R-form. (Structures compiled from Ga mien and Romanowska, 1982). many foodborne bacterial pathog ens. LA assays involve la tex particles coated with antibodies specific for target bacteria. Bindi ng causes a visible clumping of the latex particle-bacteria complexes that can be seen with the naked eye. EIA is a term used to describe many assay formats in which enzyme-labeled antibodies are used to bind antigens with detection via a colorimetric reaction using the enzyme label. Microplate

PAGE 46

33 readers are generally used for detection of the colorimetric signal and these assays can be performed in high-throughput format. IMS i nvolves small paramagnetic beads coated with antibodies against the surface antigens of bacterial cells. Paramagnetic beads exhibit magnetic properties only when placed with in a magnetic field and show no residual magnetism when removed from this fiel d. After target bacteria are bound by the antibody-coated beads, a magnetic field is used to separate the bead-bacteria complexes. Once the magnetic field is removed, the bead -bacteria complexes return to suspension (Olsvik et al ., 1994). Detection of bead-bacteria comp lexes can be performed via direct plating on microbiological media, direct mi croscopy, or nucleic acid amplification. LA, EIA and IMS methods require the use of monoclonal or po lyclonal antibodies to bind specific antigens present on the surface of bacterial cells. Fo r this reason, development of antibodies with sufficient specificity is cr itical for the perfor mance of immunological methods. A summary of these and other rapid methods for Shigella detection is presented in Table 2-3. Latex agglutination methods for Shigella LA requires the prior isolati on of a suspect colony on so lid media. For this reason LA methods cannot be used for detection of Shigella directly from a food sample, but rather can confirm or aid in characterization of suspected Shigella colonies. Several latex agglutination serotyping kits for Shigella are commercially availa ble, however only one representative kit will be discussed in this review. The Wellcolex Colour Shigella Test (WCTShigella Remel Inc., Lenexa, KS) allows the identification of isolates to the species level using only two reagents, each consisting of a mixture of red and blue latex particles coated with antibodies specific for

PAGE 47

34Table 2-3. Selected rapid methods for detection of Shigella Serogroup Target Matrix Sample Preparation Detection Detection Limit Reference S. sonnei S. boydii ipaH Tomatoes FTA filtration Nested PCR < 101 CFU/tomato Warren, 2003 S. flexneri ipaH Water SE, boiled extract Semi-nested PCR 1.1 x 101 CFU/ml Theron et al ., 2001 S. dysenteriae 1 virA Mayonnaise CEP PCR 102 -103 CFU/ml Villalobo and Torres, 1998 S. dysenteriae virA Mussels PE, CEP PCR 101 -102 CFU/ml Vantarakis et al ., 2000 S. flexneri plasmid DNA Lettuce Alkaline denaturation PCR 1.0 x 104 CFU/g Lampel et al ., 1990 S. flexneri ial Various foods SE, BDC Nested PCR 1.0 x 101 CFU/g Lindqvist, 1999 S. sonnei form I LPS, ial Feces IMS PCR 1.0-1.5 x 101 CFU/ml Achi-Berglund and Lindberg, 1996 S. dysenteriae S. flexneri S. boydii S. sonnei LPS Rectal swabs PE, DPSM EIA ND Sonjai et al ., 2001 S. dysenteriae LPS Various foods None Biosensor 4.9 x 104 CFU/ml Sapsford et al ., 2004 S. dysenteriae 1 S. flexneri S. sonnei LPS, ial Feces IMS PCR ND Achi et al ., 1996 S. dysenteriae 1 S. flexneri LPS, ial Feces IMS PCR 1.0 x 101 CFU/g Islam and Lindberg, 1992 S. dysenteriae S. flexneri S. boydii S. sonnei LPS, 16S rRNA Sewage Immunocapture UPPCR 5.0 x 101 CFU/ml Peng et al ., 2002 S. dysenteriae 1 S. flexneri LPS Feces IMS EIA 1.0 x 103 CFU Islam et al ., 1993a Abbreviations: SE – selective enrichment PE – pre-enrichment, ND – not determined, LPS – lipopolysaccharide, EIA – enzyme immunoassay, BDC – buoyant density centrifugati on, UPPCR – universal primer PCR, DPSM – direct plate on selective media, CEP – chemical extraction/ ethanol precipitation.

PAGE 48

35 one of the four different Shigella serogroups. Each reagent is added to one of two identical sample spots of the suspect isolate. The presence of homologous antigen results in the agglutination of one color coupled with a change in background color. Color change combinations due to any of the four Shigella serogroups are easily distinguishable as is the negative reaction in which the pa rticles remain in sm ooth purple suspension. Non-specific agglutination results in a pur ple agglutination with a clear background. Bouvet and Jeanjean (1 992) tested the WCTShigella against 100 Shigella isolates from human stools and observed specific ity and sensitivity at 100% and 98%, respectively. The two Shigella strains that did not give visible aggregates were S. dysenteriae type 2 and S. flexneri type 4. These results were supported by Kocka et al (1992). Using the WCTShigella 42 of 42 clinical Shigella isolates and seven of eight stock Shigella cultures were correctly grouped (Kocka et al ., 1992). The stock Shigella culture that was not correctly grouped had b een repeatedly passed in culture that may have resulted in the loss of some antigenicity. Lefebvre et al (1995) evaluated WCTShigella and six commercial slide agglutination Shigella serogrouping kits for accuracy. The WCTShigella was easily performed and interpreta tion of results was less subjective than the other tests. WCTShigella met a performance standard of 90% accuracy in these evaluations. Enzyme immunoassay methods for Shigella Although EIA methods for the detection of foodborne pathogens are common, the only commercially available kits for Shigella spp are the Shigel-Dot A (for S. dysenteriae ), B (for S. flexneri ), C (for S. boydii ), and D (for S. sonnei ) test kits (Science Development and Management. Ltd., Bangkok, Th ailand). The Shigel-Dot kits are membrane dot-blot enzyme-linked immunosorbent assays (ELISA). These test kits have

PAGE 49

36 been validated using 500 rectal swabs and ha ve been compared to conventional culture isolation and Western blot anal ysis. The diagnostic accuracy of the Shigel-Dot A, B, C, and D was 99.2%, 95.0%, 94.0%, and 96.4%, re spectively, when compared to conventional culture isolation, and all were 100% when compared to the conventional culture isolation and Western blot combined (Sonjai et al ., 2001). Monoclonal antibodies included in the Shigel-Dot D k it are reported to detect both S. sonnei form I and form II LPS. Immunomagnetic separation methods for Shigella detection IMS has been investigated for the con centration and purific ation of bacterial pathogens from food samples (Cudjoe and Kr ona, 1997; Hsih and Tsen, 2001; Drysdale et al ., 2004; Lynch, et al ., 2004) and are reportedly more sensitive than comparable conventional culture methods (Cudjoe a nd Krona, 1997; Hsih and Tsen, 2001). AntiSalmonella antiE. coli O157:H7, antiCampylobacter and antiListeria IMS beads are commercially available (Matrix MicroScien ce, Cambridgeshire, UK; Dynal Biotech, Oslo, Norway) however, antiShigella IMS beads are not yet commercially available. IMS has been used as a technique to concentrate Shigella from clinical samples for downstream detection proce sses such as EIA (Islam et al ., 1993a) or PCR (Islam and Lindberg, 1992; Achi et al ., 1996; Achi-Berglund and Li ndberg, 1996). By using IMS, PCR inhibitors inherent to fecal samples we re successfully eliminated. Briefly, IMS was used to concentrate S. sonnei (Achi et al ., 1996; Achi-Berglund and Lindberg, 1996), S. flexneri (Islam and Lindberg, 1992; Achi et al ., 1996), and S. dysenteriae serotype 1 (Islam and Lindberg, 1992; Achi et al ., 1996) from feces prior to PCR. The detection limit for the IMS-EIA and IMS-PCR methods were 1.0 x 103 CFU/ml (Islam et al ., 1993a) and 1.0-1.5 x 101 CFU/ml (Islam and Lindberg, 1992; Achi-Berglund and

PAGE 50

37 Lindberg, 1996), respectively. In addition, the IMS-PCR method was more than two times as effective then the conventional cultur e method for the diagnosis of shigellosis in children with severe diarrhea (Achi et al ., 1996). These studies ha ve significance with respect to Shigella detection in foods, however furthe r investigations are needed to evaluate the suitability of the IMS techniques for use with food samples. Immunomagnetic separati on using the Pathatrix Recently, a novel device allowing flow-t hrough IMS, the Pathatrix, has been developed (Figure 2-5). The Pathatrix allows a 250 ml sample to be continuously pumped through a tubing system in which IMS beads are immobilized in a capture phase, thereby allowing the analysis of a complete 25 g sample + 225 ml buffer/enrichment broth homogenate with or without prior incubation. Each Pathatri x unit can analyze up to five samples at one time. The incubation pots in wh ich the sample (stomacher bag) is placed are temperature controlled from room temper ature up to 45C. After analysis, the tubing system is disconnected from the sample and c onnected to a vessel containing wash buffer which is then pumped over the capture phase to wash away any food particles or unbound microorganisms. Once washing is complete the beads are recovered in buffer and the captured microorganisms may be detected using selective plating, colorimetric assays or molecular assays. Arthur et al. (2005) compared the Pathatrix met hod in combination with selective plating and PCR using the Lightcycler (Roche Applied Science, I ndianapolis, IN) with the BAX system (DuPont Qualicon, Wilm ington, DE) and the Assurance GDS (BioControl Systems, Inc., S eattle, WA) for detection of E. coli O157:H7 in ground beef. When samples were inoculated at 17 CFU/ 65 g ground beef, all of the investigated methods detected E. coli O157:H7 in 57 of 57 samples. However, when samples were

PAGE 51

38 Figure 2-5. The Pathatrix system for flow -through immunomagnetic separation (IMS). The Pathatrix base unit (A) can hold up to five samples at one time. The capture phase (B) utilizes a magnet to immobilize the IMS beads within the tubing system to allow binding to speci fic antigen as they pass through. inoculated at 1.7 CFU/65 g ground beef, the Pa thatrix in combination with selective plating or PCR (4 hr pre-enrichment) detected E. coli O157:H7 in 98% of samples whereas the BAX system (8 hr pre-enrichment ) and the Assurance GDS system (8 hr preenrichment) detected E. coli O157:H7 in 66% and 73% of samples, respectively (Arthur et al ., 2005). Additional technologies for immunological detection of Shigella In a recent report, Sapsford et al (2004) describe detection of S. dysenteriae in buffer and on chicken carcasses using an ar ray biosensor developed at the Naval Research Laboratory. The array biosensor meas ures total internal reflection fluorescence using a 25-min sandwich immunoassay for antigen detection. An advantage of the array biosensor is that little or no sample preparati on is required prior to analysis. The detection limit of the array biosensor for S. dysenteriae in ground turkey, ch icken carcass wash, buffered milk, and a lettuce leaf wash was observed at 7.8 x 105 CFU/g, 4.9 x 104 CFU/ml, 7.8 x 105 CFU/ml, and 2.0 x 105 CFU/ml, respectively. When tested in buffer, the array biosensor did not respond as e fficiently to the other serogroups of Shigella and

PAGE 52

39 cross-reacted with E. coli suggesting the specificity of the polyclonal antibodies was a limiting factor. The use of a monoclonal anti body with higher specificity may improve the diagnostic ability of this testing format. Molecular Microbiological Methods for Shigella Detection in Foods Polymerase chain reaction detection of Shigella in foods Polymerase chain reaction (PCR) methods for detection of Shigella in food have previously demonstrated higher sensitivity than comparable culture methods (Warren, 2003). PCR assays for Shigella spp. have targeted the i nvasion associated locus ( ial ) (Islam and Lindberg, 1992; Lindqvist, 1999), the virA gene (Villalobo and Torres, 1998; Vantarakis et al ., 2000), or the ipaH gene (Sethabutr et al ., 1993; Sethabutr et al ., 2000; Theron et al ., 2001; Lampel and Orlandi, 2002; Wa rren, 2003) for amplification. The same PCR primers are used to de tect each of the serogroups of Shigella and enteroinvasive E. coli The ial and virA genes are located on the large virulence plasmid (sometimes referred to as the invasion plasmid), however sequencing of the S. flexneri genome (Jin et al ., 2002) revealed the ipaH gene to be encoded multiple times on both the chromosome and the large virule nce plasmid. Thus, detection of the ipaH gene is possible in the event the large virulence plas mid has been lost, which has been shown to occur when food samples are stored for a pr olonged periods of time prior to analysis (Lampel and Orlandi, 2002). The discussion be low will be limited to PCR detection of Shigella in food samples, however se veral reports of detection in fecal samples appear in the literature. These methods are summarized along with other rapid methods in Table 23. Vantarakis et al (2000) developed a multiplex PCR method to detect both Shigella spp. and Salmonella spp. in mussels. Multiplex PCRs involve the use of two or more sets

PAGE 53

40 of primers in the same PCR such that multiple targets can be amplified in the same reaction. Artificially inoculated S. dysenteriae and S. Typhimurium were recovered by homogenizing mussel meat with peptone wate r. DNA from an aliquot of the homogenate was purified using a guanidi ne isothionate method and concentrated via ethanol precipitation. The virA gene of Shigella spp. and the invA genes of Salmonella spp. were amplified. Sample homogenates were inoculated at various concentrations to establish the lowest detection limit for the method. When samples were not preenriched prior to analysis, the multiplex PCR me thod was able to detect S. dysenteriae at 1.0 x 103 CFU/ml in the homogenate. Following 22 hour incubation in buffered peptone water the multiplex PCR methods was able to S. dysenteriae detect levels as low as 1.0 x 101 to 1.0 x 102 CFU/ml in the homogenate. Sim ilar results were observed for S. Typhimurium. Villalobo and Torres (1998) investig ated PCR for the detection of S. dysenteriae serotype 1 in mayonnaise. Samples were hom ogenized in buffered peptone water and artificially contaminated with various concentrations of S. dysenteriae serotype 1. Bacterial cells were lysed with detergent, the DNA extracted with phenol-chloroform and precipitated with ethanol. A multiplex PCR wa s then used to amplify regions from the virA gene and the 16S rRNA gene. The lowest level of inoculation detected by this method was 1.0 x 102 to 1.0 x 103 CFU/ml in the homogenate. In a study by Lindqvist (1999), nested PCR was investigated for the detection of the ial locus of Shigella spp. from spiked lettuce, shrimp, milk, and blue cheese samples. Nested PCR involves the use of two sequential PCRs in which the target sequence in the second PCR lies within the amplified sequence in the first PCR. Nested PCRs can be used to achieve extreme low sensitivities. Food samples inoculated were homogenized in

PAGE 54

41 physiological saline and bacteria were is olated by buoyant density centrifugation (separation of components base d on density). Single PCR, us ing only the external primer sets, was able to detect S. flexneri in aqueous solution at 0.5-1.0 x 105 CFU/ml. Nested PCR was more sensitive and was able to detect 1.0 x 103 CFU/ml. The nested PCR assay in combination with buoyant density centrifugation was able to detect S. flexneri inoculated onto all four foods at 1.0 x 101 CFU/g (Lindqvist, 1999). Theron et al (2001) investigated a semi-neste d PCR for the detection of the ipaH gene of S. flexneri in spiked environmental water samples. The detection limits in the various environmental water samples were 2.0 x 103 CFU/ml for well water, 1.4 x 101 CFU/ml for lake water, 5.8 x 102 CFU/ml for river water, 6.1 x 102 CFU/ml for treated sewage water, and 1.1 x 101 CFU/ml for tap water. Variab ility in results among the water samples was attributed to the presence of humic substances that inhibited PCR. Humic substances are transformed products from so il organic matter that do not belong to the known classes of biochemistry. These include humic acids, fulvic acids, and humins. Preenrichment in GN broth served to dilu te humic substances while allowing the S. flexneri to multiply, thereby increasing the concentration of target DNA. Improved PCR detection of Shigella by FTA filtration FTA filters (Whatman, Clifton, NJ) have been used in template preparation for PCR detection of pathogenic microorganisms (Lampel et al ., 2000; Orlandi and Lampel, 2000; Lampel and Orlandi, 2002; Warren, 2003; Warren et al ., 2005b). Moisture in the sample activates chemicals in the FTA filter that lyse cells, denature enzymes, inactivate pathogens, and immobilize genomic DNA (Wha tman website, 2004). Chemicals in the FTA filters also protect DNA and RNA from light, free radicals, enzymes, and pathogens during dry, room-temperature storage. FTA filtration as sample preparation for PCR has

PAGE 55

42 been developed for detection of Cyclospora and Cryptosporidium in water (FDA, 1998). Recently, this sample preparation technique has been investigated for detection of S. boydii and S. sonnei from tomato rinses by nested PCR (Warren, 2003; Warren et al ., 2005b). Briefly, tomato rinses were passed thr ough a two-stage filter where the first stage contained filter paper for size exclusion and the second stage contained FTA filter paper. After purification and washing, 6-mm punches ar e taken from the FTA filters and used directly as template in the first step of the nested PCR. The FTA filtration/nested PCR (FTA-PCR) assay detected S. boydii and S. sonnei at 6.2 x 100 CFU/tomato and 7.4 x 100 CFU/tomato, respectively (Warren, 2003; Warren et al ., 2005b). Although FTA-PCR had excellent sensitivity when used to analyze tomato rinses, which are relatively clean, similar results were not observed when recovery of inoculated S. boydii and S. sonnei was attempted from cantaloupes, st rawberries, or retail Valencia oranges (Schneider and Warren, unpublished data). FTA-PCR was unable to detect inoculated Shigella from strawberries, suggesting th at the purification and washing protocol used on the FTA filters may not be adequate to remove PCR inhibitors (Warren, 2003). Despite several size exclusion filtering techniques applied prior to FTA filtration, the FTA filters were routinely clogged by cantaloupe fibers or the wax coating from retail oranges, which entered the buffer rinse dur ing recovery procedures (Warren, 2003). The advantage of the FTA-PCR technique is that no enrichment was necessary to obtain low detection limits due to the large volume of sa mple analyzed by the filters. The technique warrants further investigation fo r the analysis of foods for Shigella contamination as improvements in pre-filtration techni ques could improve the sensitivity.

PAGE 56

43 Additional molecular microbiologi cal techniques for detection of Shigella An interesting new method, immunocapture universal primer PCR (iUPPCR) has been recently reported (Peng et al ., 2002). Universal primer PCR employs primers designed against highly conserve d regions, such as 16S rRNA genes, thus the resulting PCR can be used for amplification of almost any bacteria. Typically, the sequence of the resulting amplicon would be further analyzed for identification of the bacteria. In iUPPCR however, immunocapture is employed for specificity and universal primer PCR is used for detection of captured bacteria Monoclonal antibodies sp ecific for individual serotypes of S. dysenteriae S. flexneri S. boydii or S. sonnei were coated into the wells of a 96-well polystyrene plate. Cross-reactivity tests demonstr ated the specificity of the monoclonal antibodies to the strain level. Wells were challenged with bacteria and captured bacteria were detected using unive rsal primer PCR, i.e. PCR with primers specific for 16S rRNA sequences conserve d among closely related bacteria. The detection limit for shigellae in broth was 5.0 x 102 CFU/ml (Peng et al ., 2002). Presumably, detection of up to 96 pathogens sharing the conserved 16S rRNA sequence could be accomplished in the same plat e using specific monoclonal antibodies for bacterial capture (differentiation of cell types) and the universal primer PCR for detection. Further investigation of this met hod with respect to food samples as opposed to bacteria in brot h is required. More recently, Ji et al (2006) reported iUPPCR in combination with denaturing gradient gel electrophoresis (DGGE) for the detection and identification of Shigella spp. In contrast to the iUPPCR method desc ribed above, genus-specific polyvalent monoclonal antibodies were used to coat the wells of a 96-well pl ate, such that all Shigella spp. may be captured in each well. Following iUPPCR to amplify 16S rRNA

PAGE 57

44 gene fragments, DGGE was used to identify the specific serotype. Unfortunately, the iUPPCR-DGGE method was only tested agai nst pure cultures in laboratory conditions. Although the authors specula te that it would be useful fo r the detection and identification of Shigella in food and environmental samples, further testing in the presence of food/environmental matrices and in digenous microflora are required. Advances in oligonucleotide microarrays have further enabled the simultaneous detection and identification of bacter ial foodborne pathogens. Oligonucleotide microarrays involve the ordered immobiliza tion of specific probes to a solid surface followed by hybridization with labeled sample DNA. After hybridization, development of the label can allow identification/enumeration of complementary sample DNA sequences. In a recent report, Kakinuma et al (2003) describes an oligon ucleotide microarray using probes specific for the gyrB gene to detect and differentiate E. coli Shigella and Salmonella PCR amplified regions of the gyrB gene were fluorescently labeled and hybridized to detection probes immobilized on glass slides. Based on reaction patterns, three species of Shigella seven serovars of Salmonella and one strain of E. coli were correctly identified at the species level. Identification at the subspecies level of Salmonella was problematic when multiple serovars were present in the same sample due to the overlap of microarray patterns. Assay formats such as this could be expanded to potentially detect and identify a wide range of enteric pathogens. Song et al (2005) reported an alternativ e DNA amplification method, loopmediated isothermal amplificati on (LAMP), for the detection of Shigella spp. in clinical samples. In the LAMP assay, four specialized primers designed to specifically recognize six distinct regions on the target gene ar e used in combination with a DNA polymerase

PAGE 58

45 with strand displacement activit y. The amplification reaction o ccurs at 65C, therefore no thermocycler is necessary as with PCR. Fo r a more complete description of the LAMP assay, the reader is directed to the Ei ken Chemical Co., Ltd. (Tokyo, Japan) website: http://loopamp.eiken.co.jp/e/lamp/index.html (Accessed 07 Jun 2006). Stool samples homogenized in sterile wa ter were inoculated w ith various titers of S. flexneri YSH6000 and the DNA was extracted using a boiling me thod. The sensitivity of the LAMP assay for S. flexneri in the inoculated stool samples wa s eight CFU per reaction whereas the sensitivity of a PCR reaction was determined to be 800 CFU per reaction. A distinct advantage of the LAMP assay is that amplif ied DNA may be visualized by the naked eye as turbidity in the sample, eliminating the need for post-amplification processes, such as gel electrophoresis. Finally, repetitive se quence-based PCR (rep-PCR) has been demonstrated for the identification and molecular typing of members of the family Enterobacteriaceae including Shigella (Raza et al ., 2003; Lising et al ., 2004). In rep-PCR, primers are designed complementary to bacterial interspe rsed repetitive seque nces and the regions lying outward of the primers are amplif ied resulting in DNA fragments of varying lengths. These fragments can then be separa ted by electrophoresis to form a bacterial fingerprint unique to indivi dual bacterial strains. The Di versiLab System (Spectral Genomics, Inc., Houston, TX) generates su ch fragments by rep-PCR that are then analyzed using microfuidics lab-on-a-chip and the Agilent 2100 Bioanalyzer (Agilent Technologies, Inc. Palo Alto, CA). Results ca n be visualized in several formats including dendograms, electropherograms, gel-like image, or scatte r plots (Lising et al ., 2004). Rep-PCR technology has been shown to be repr oducible and can allo w the differentiation

PAGE 59

46 of species, subspecies and strains in as little as 4 hours. Some Shigella serogroups show >95% similarity with the DiversiLab Enteri c Kit Beta version, however the DiversiLab Shigella kit offers greater differentiation among serogroups and strains (Lising et al ., 2004). Strains of E. coli (non-pathogenic, enterohemorragi c, and enterotoxigenic) were effectively differentiated from Shigella spp. using rep-PCR (Raza et al ., 2003; Lising et al ., 2004), however no enteroinvasive E. coli strains were included in the studies. Detection of Pathogens by Sequence Capture Sequence capture methods have been inve stigated for the detection of a wide variety of bacterial, viral, and fungal pa thogens. Typically, a sp ecific oligonucleotide probe is immobilized on a solid surface and DNA/RNA prepared from the target microorganism is hybridized to the probe. Zammatteo et al (1997) compared the sequence capture of human cytomegalovi rus using DNA probes immobilized on 96-well microtiter plates and paramagnetic beads and found that faster hybridization kinetics were obtained with the use of beads as the solid support. Although Zammatteo et al (1997) used DNA probes covalently bound to the solid supports, it is more common to use oligonucleotide probes labe led with biotin on their 5 end with paramagnetic beads or microtiter plates with wells pre-coated with streptavidin. As with more traditional nucleic acid hybridization, variables such as salt concen tration, temperature, contact time and GC content of the probe sequence all have infl uence on hybridization efficiency. Del Gallo et al (2005) investigated steric factors affecting the hybrid ization of PCR amplified sequences to DNA probes immobilized on the sc reen-printed gold surface of disposable electrodes. The amount of probe coverage on the surface as well as the relative position of the probe on the target sequence was found to partially control hybridization efficiency. When the probe coverage was approx. 2.9 x 1012 molecules/cm2 and when the

PAGE 60

47 probe sequence was located at on e of the termini of the target sequence the efficiency of hybridization was increased (Del Gallo et al ., 2005). In addition, the use of spacer molecules between the solid support and the probe sequence has been shown to drastically improve hybridization efficiency (Shchepinov et al ., 1997; Amagliani et al ., 2006). The method used to purify the probe during synthesis may also affect hybridization efficiency, with HPLC purification preferred over desalting (Amagliani et al. 2006). Finally, sequence capture methods may be used to prepare DNA/RNA from samples that contain PCR inhibitors, sin ce these inhibitors are removed during posthybridization washing (Maher et al ., 2001; Tsai et al ., 2003). While there are numerous reports of sequen ce capture for the analysis of clinical samples for bacterial and viral pathogens, this discussion is limited to the few reports of sequence capture used for the analysis of f oods or environmental samples. In a study by Chen et al (1998), sequence capture in combinati on with PCR was investigated for the detection of verotoxigenic E. coli (VTEC) in brain heart infu sion (BHI) broth cultures and in artificially contaminated ground beef. Biotin-labeled probes were used to form hybrids with specific DNA segments and then the hybrids were bound using streptavidincoated paramagnetic beads. In BHI broth, det ection of initial VTEC concentrations of 103, 102 and 100 CFU/ml was achieved after 5, 7 and 10 hr enrichment at 37C, respectively. In ground beef samples, the se quence capture method was able to detect VTEC at levels of 100 CFU/g after 15 hr incubation in BHI broth. For both BHI broth and ground beef samples, DNA was extracted from one ml aliquots of the enrichments by the boiling method and the prepared DNA was used as template in the sequence capture method. In a subsequent study, Chen and Griff iths (2001) modified this sequence capture

PAGE 61

48 method for the simultaneous detection of Salmonella and Shiga-like toxin-producing E. coli Again using cultures in BHI broth, the detection limit was determined to be 100 CFU/ml after 10 hr incubation at 37C (Chen and Griffiths, 2001). This study demonstrates that multiple probes, each speci fic for a different pathogen, may be used in combination on solid supports in order to test for different pathogens at the same time. Tsai et al (2003) investigated sequence capture in combination with PCR for the detection of enterotoxigenic E. coli (ETEC) associated with cattle in environmental water samples. Biotin-labeled probes with specifi city for the enterotoxin gene LTIIa were attached to streptavidin-coated paramagne tic beads and used to form hybrids with prepared DNA samples. The detection limit of the assay was determined to be 2.5 attogram/l DNA. In addition, some of the extracted DNA samples were spiked with humic acids, known inhibitors of PCR, to de termine if the sequence capture method was effective at removing these inhi bitors prior to PCR. In th e presence of humic acids, the sequence capture method increased sensitiv ity 10,000-fold over conventional PCR (Tsai et al ., 2003). Amagliani et al (2006) developed sequence capture in combination with PCR for the detection of Listeria monocytogenes in milk. Using NH2-labeled DNA probes specific for the hlyA gene coupled to paramagnetic beads, the detection limit of the assay was 101 CFU/ml. The sensitivity was achieved using (CH2)12 spacers between the NH2-label and the DNA probe sequence. In contrast to ma ny previously report ed sequence capture methods, the method of Amagliani et al (2006) required no DNA ex traction/purification as capture was performed directly in the milk sample. For this reason a two-step sequence capture in which biotin-labeled probes form ed hybrids with target sequences followed by

PAGE 62

49 isolation of the hybrids using streptavidin-c oated paramagnetic beads was not successful. It was hypothesized that indigenous biotin in the milk samples interfered with the biotinlabeled hybrids binding to the streptavidin-coated beads (Amagliani et al ., 2006).

PAGE 63

50 CHAPTER 3 MATERIALS AND METHODS Preliminary Studies Acquisition and Maintenance of Shigella sonnei Cultures The following cultures were purchased from the American Type Culture Collection (ATCC; Manassas, VA): S. sonnei ATCC 9290, S. sonnei ATCC 29031, S. sonnei ATCC 29030, S. sonnei ATCC 25931 and S. sonnei ATCC 29930. Each strain was resuscitated as instruct ed in the package inserts a nd the resulting growth was streaked for isolation on MAC plates. One typical S. sonnei colony from each plate was transferred to a TSA slant and stored at 4 C. A second typical S. sonnei colony was transferred per product instruc tions to Protect ™ Bacterial Preservers (Scientific Device Laboratory, Inc., Des Plaines, IL) and stored at -70 C. Adaptation of Cultures to Rifampicin Subcultures of each of the five S. sonnei strains were adapted to the bactericidal agent rifampicin by spontaneous mutation. A 10,000 g/ml (1%) stock solution of rifampicin was prepared by dissolving 2.0 g ri fampicin (Fisherbrand, Fisher Scientific, Pittsburg, PA) in 200 ml methanol. The stock so lution was then filter st erilized and stored in the dark at 4C. Stock cultures were grown overnight in 10 ml TSB (37 C, 30 rpm). Overnight cultures were transferred to 10 ml TSB s upplemented with 2.5 g/ml rifampicin and grown overnight (37 C, 30 rpm). With each transfer, th e concentration of rifampicin increased until the final concentration was 200 g/ml rifampicin. Once the cultures were

PAGE 64

51 adapted to 200 g/ml rifampicin, cultures were grown overnight (37 C, 30 rpm) three consecutive times in TSB supplemented with 200 g/ml rifampicin to ensure well adapted populations. Once adaptation was complete, the final overnight culture was streaked for isolation on MAC and inc ubated overnight at 37 C. One typical colony from the overnight MAC plate was transferred to a TSA slant supplemented with 200 g/ml rifampicin and stored at 4 C. Another typical colony was tr ansferred per manufacturer’s instructions to a Protect™ Bacter ial Preserver and stored at -76 C. Preparation of Mi crobiological Media Wild-type cultures were grown in Tryp tic Soy Broth (TSB; BD Diagnostics, Franklin Lakes, NJ) containing 10 M Congo red (TSCR) and maintained on Tryptic Soy Agar (TSA; BD Diagnostics) slants contai ning 10 M Congo red. For survival studies only, bacterial strains were adapted to 200 g/m l of the bactericidal agent rifampicin and experiments were conducted in TSB supplemen ted with 100 g/ml rifampicin (TSB rif+). All dilutions, unless otherw ise specified, were performed using 0.1% Peptone (BD Diagnostics) water. Phosphate Buffered Salin e (PBS; pH 7.4) was prepared using the following formula per liter: 1.2 g NaHPO4, 8.2 g Na2PO4 and 5.0 g NaCl. Shigella Broth (SB) was prepared accor ding to the U.S. Food and Drug Administration’s (1998) Bacteriological Analytical Manual (BAM) and supplemented with novobiocin (ICN Biomedicals Inc., Aurora, OH) at 3.0 g/ml. MacConkey Agar (MAC; BD Diagnostics), Triple Sugar Iron Agar (TSI; BD Diagnostics), Lysine Iron Agar (LIA; BD Diagnostics) and Motility Medium (MM; BD Diagnostics) were all prepared according to manufacturers instructions TSI and LIA were prepared as slants in

PAGE 65

52 screw cap tubes. MM was supplemented with 0.005% triphenyltetr azolium chloride (TTC; BD Diagnostics). When necessary, medi a pH was adjusted using filter sterilized 1N NaOH. Acquisition/Preparation of Food Matrices Mature green, unwashed, unwaxed tomatoes (Florida 47 cultivar) were obtained from a nearby packinghouse. Tomatoes were held at 4C prior to use. ‘Picnic Potato and Egg Salad’ (hereafter referred to as potato salad) was prepared the night before each experiment using a publicly available recipe ( http://southernfood.about.com/od/ potatosalads/r/bl00624c.htm?terms =picnic+potato+and+egg+salad ). Briefly, 6.0 lbs potatoes were skinned, cubed, boiled for 15 min, and then cooled. Eight large eggs were hard boiled for 15 min, cooled, and chopped. Af ter straining the water from the cooked potatoes, the chopped eggs, cup chopped re d onion, 1 cup chopped fresh celery, 1 and cup mayonnaise (Hellman’s Lite, Unilever, Englewood Cliffs, NJ), 3 tbs yellow mustard (Publix Supermarket brand, Lakeland, FL), 1 tsp salt, and 1 tsp black pepper were added and mixed well with a large se rving spoon. The potato salad was stored at 4C until use. Ground sirloin (90% lean/10% fat; hereafter referred to as ground beef) was purchased at a local grocer the morning of each experiment. Acquisition and Maintenance of AntiShigella Antibodies All polyclonal and monoclonal antibodies used in the following experiments are commercially available and are listed in Table 3-1. AntiShigella polyclonal antibodies AB01 and AB04 were purchased from Virost at, Inc. (catalog nu mber 0901; Portland, ME) and AbCam, Inc. (catalog number Ab19988; Cambridge, MA), respectively. Goat anti-rabbit IgG (H+L chain specific) antibodie s (AB03) were purchased from Southern Biotech Associates, Inc. (catalog nu mber 4050-01; Birmingham, AL). AntiS. sonnei

PAGE 66

53 monoclonal antibodies (AB02) were purchased from Novus Biologicals (catalog number BM1316; Littleton, CO). Goat anti-mouse IgG (H+L chain specific) antibodies (AB05) were purchased from Southe rn Biotech Associates, Inc. (catalog number 1031-01). Upon receipt, all antibodies were stored at 4.0C if they were to be used within two months otherwise aliquots of 100 l were stored frozen at -20C. Binding of Antibodies to Paramagnetic Beads The binding of antibodies to MagaCell bead s was performed according to Cortex Biochem, Protocol 503 (available at http://www.cortex-biochem.com/commerce/ccc1010 -protocols.htm ) except that all volumes were re duced proportionately for a starting volume of 500 l beads. The MagaCell bead s were mixed by inverting the bottle by hand until no beads were visible as a pellet on the bo ttom of the bottle. A 500-l aliquot of the beads were transferred to a clean, sterile 1.5 ml microcentr ifuge tube. All washing steps were performed using a magnetic particle concentrator (MPC-S; Dynal Biotech) or a magnetic separator (CD3002; Cortex BioChem) to draw the beads to th e side of the tube allowing the supernatant to be removed us ing a pipette. The beads were washed two times in 1000 l de-ionized water, followe d by four times in 1000 l acetone. The beads were then resuspended in 250 l acetone containing 0.12 g 1,1 carbonyldiimidazole per 10 ml. The microcentrifuge tubes were placed on a rugged rotator and mixed by gentle end-over-end rotation for 1 hr. The beads we re then washed four times in 1000 l acetone, four times in 1000 l de-ionized water and four times in 1000 l 0.1M sodium bicarbonate buffer, pH 8.6. The beads were then resuspended in 400 l 0.1M sodium bicarbonate buffer, pH 8.6, an appropriate amount of antibod y solution (Table 3-1) was added and the volume was adjusted to 500 l using 0.1M sodium bicarbonate buffer, pH

PAGE 67

54 Table 3-1. Antibodies investig ated for immunocapture of S. sonnei Designation Clonality Immunogen Volume added per 500 l MagaCell beads AB01 Polyclonal Mixture of S. boydii, S. flexneri, S. dysenteriae 50 AB02 Monoclonal S. sonnei NCTC 9774 (Wheeler phase I) 100 AB03 Polyclonal Pooled antisera from goats hyperimmunized with normal rabbit IgG 100 AB04 Polyclonal Membrane extract of S. sonnei and S. flexneri 50 AB05 Polyclonal Pooled antisera from goats hyperimmunized with mouse IgG paraproteins 100 8.6, when necessary. The beads were again placed on a rugged rotator and mixed with gentle end-over-end rotation for 18-24 hr at room temperature. The next day, the beads were washed two times in 500 l 0.1M s odium bicarbonate buffer, pH 8.6. The beads were then resuspended in 500 l 0.1M sodi um bicarbonate buffer, pH 8.6, containing ethanolamine (3 ml/liter) and mixed by gentle end-over-end rotation for 1 hr. The beads were then washed once in 500 l 0.1M sodium acetate buffer, pH 4.0, resuspended in 500 l 0.1M sodium acetate buffer, pH 4.0, and mixed by gentle end-over-end rotation for 1 hr. Finally the beads were washed three ti mes in 500 l PBS containing 0.1% sodium azide and stored at 4 C until used. Evaluation and Optimization of Immunocapture Using AntiShigella Beads Preliminary challenges of antiShigella beads for immunocapture of S. sonnei strains were performed in PBS. A 100-l aliquot of an overnight S. sonnei culture in TSCR was used to inoculate sterile stomacher bags containing 250 ml PBS such that the final cell titer was approximately 2.0 x 105 CFU/ml. The stomacher bags were then placed into Pathatrix incubation pots and the samples were circulated for 30 min at 37C

PAGE 68

55 using 25 l aliquots of antiShigella beads for immunocapture. AntiShigella beads prepared using various antiShigella antibodies and various concentrations of antiShigella beads were investigated for FTI. Afte r circulation, each sample was washed in PBS and the beads were recovered in approxi mately 250 l PBS. A 100-l aliquot of the recovered beads was analyzed by spread plate using MAC. MAC plates were incubated at 37C for 24 hr and the resulting colonies were counted. Antibodies AB01, AB02 and AB04 were bound directly to MagaCell beads and tested for primary capture in FTI. In addition, AB01 and AB02 were te sted for immunocapture of S. sonnei in broth culture followed by secondary capture using MagaCe ll beads coated with antibodies AB03 and AB05, respectively. Crude DNA Extraction fro m Bacteria by Boiling Stock cultures frozen on Pr otect ™ Bacterial Pr eservers were retrieved from frozen (-70 C) storage and allowed to thaw. One bead was aseptically transferred from the Protect ™ Bacterial Preserver into 10 ml TSB and grown overnight (37 C). Overnight cultures were plated for isolation on an appr opriate selective and differential medium and incubated overnight at 37 C. Overnight plates were observed for typical colony morphologies. One typical colony was transf erred to 10 ml TSB and grown overnight (37 C). A 1.0-ml aliquot of the overnight culture was transferred to a clean, sterile 1.5 ml microcentrifuge tube and the bacterial cel ls were harvested by centrifugation (3,220 x g for 10 min). The resulting supe rnatant was discarded and th e pellet was re-suspended in 200 l de-ionized, sterilized water. Samples were then boiled for 10 min in a dry bath incubator (Fisher Scientific, IsoTemp 125D ). The supernatant (DNA template) was

PAGE 69

56 aseptically transferred to a cl ean, sterile 1.5 ml microcentrif uge tube and stored at -20 C. The pellet was discarded. DNA Extraction from AntiShigella Beads using the DNeasy Kit Extraction of DNA from bacterial cells captured on antiShigella immunomagnetic beads was performed using a modified protocol with DNeasy spin columns (Qiagen, Valencia, CA). A 100-l aliquot of concentrated antiShigella beads was transferred to a clean, ster ile 1.5 ml microcentrifuge tu be and heated on a dry bath incubator at 95C for 10 min to lyse bacterial cells. After incubation, 100 l sterile water, 200 l buffer AL (Qiagen) and 200 l absolu te ethanol was added to the sample. The resulting mixture was vortexed for 5 sec and pl aced in a MPC-S for at least 1 min to draw the magnetic particles to the side of the tube Without disturbing the magnetic particles, the mixture was transferred to a DNeas y spin column and centrifuged at 6,000 x g for 1 min. The DNeasy spin column was transferre d to a clean collec tion tube and 500 l buffer AW1 (Qiagen) was passed through the column by centrifugation at 6,000 x g for 1 min. The DNeasy spin column was transferre d to a clean collec tion tube and 500 l buffer AW2 (Qiagen) was passed through the column by centrifugation at 16,000 x g for 3 min. The DNeasy spin column was transferre d to a clean, sterile 1.5 ml microcentrifuge tube and 100 l buffer AE (Qiagen) was a dded to elute bacterial DNA by centrifugation at 6,000 x g for 1 min. The DNeasy spin column was discarded and the eluted DNA was placed on ice or stored at -20C pr ior to analysis by real-time PCR. Preparation of HeLa Cell Extracts Fresh HeLa cells were obtained from the laboratory of Dr. F. Southwich, University of Florida, resuspended in 1.0% Triton X-100 supplemented with Complete Mini Protease Inhibitor cock tail tablets (Roche Applied Sc iences, Indianapolis, IN).

PAGE 70

57 HeLa cells were lysed using three cycles of freeze-thaw. One hundred l aliquots of the lysed cells were transferred to clean, ster ile 1.5 ml microcentrifuge tubes and proteins were extracted using a modified method of Wessel and Flgge (1984). Briefly, the samples were vortexed and centrifuged at 5,000 x g for 2 min and the sample supernatant was transferred to a clean, ster ile 1.5 ml centrifuge tube a nd saved as ‘HeLa cell extract 1’. The remaining pellet was resuspended in 400 l methanol (Fishe rbrand), vortexed and centrifuged at 9,000 x g for 10 sec. To the solution, 200 l chloroform (Fisherbrand) was added and the resulting solution wa s vortexed and centrifuged at 9,000 x g for 10 sec. To the solution, 300 l water was added and th e resulting solution was vortexed vigorously and centrifuged at 9,000 x g for 1 min. The upper phase in the resulting supernatant was transferred to a clean, sterile 1.5 ml centrifuge tube and saved as ‘HeLa cell extract 2’. To the lower phase, 300 l methanol was added and the solution was vortexed and centrifuged at 9,000 x g for 2 min. The resulting supernat ant was transferred to a clean, sterile 1.5 ml centrifuge tube and saved as ‘HeLa cell extract 3’. The pellet was dried using forced air (generated using a transfer pipette) and saved as ‘HeLa cell extract 4’. RNA Extraction Using the RNeasy Kit Extraction of RNA from bacterial cells was performed using RNeasy spin columns (Qiagen). A 250-l aliquot of bacterial cells were transfer red to clean, sterile 1.5 ml microcentrifuge tubes containing 500 l RNAbacterial Protect Reagent (Qiagen), incubated at room temperature fo r 5 min and centrifuged at 5,000 x g for 10 min. The resulting supernatant was discarded and pell et was resuspended in 100 l TE buffer, pH 8.0, containing 1 mg/ml lysozyme (Sigma, St Louis, MO) and incubated at room temperature for 5 min. To the mixtures, 350 l buffer RLP (Qiagen), containing 10 l/ml 2-mercaptoethanol (Sigma), were added and the solutions was vortexed vigorously for 5

PAGE 71

58 sec. To the mixture, 200 l absolute etha nol was added and the solution was mixed by gentle action of the pipette The resulting mixture (approx. 700 l) was transferred to clean, sterile RNeasy spin co lumns and centrifuged at 8,000 x g for 15 sec. In a fume hood, the flow-through was discarded, 700 l bu ffer RW1 (Qiagen) was added to the RNeasy spin column and the sample was centrifuged at 8,000 x g for 15 sec. In the fume hood, the collection vessel and flow-through were discarded and the RNeasy spin column was transferred to a clean, sterile collection vessel. A 500 l aliquot of buffer RPE (Qiagen) was added to the RNeasy spin co lumn, the sample was centrifuged at 8,000 x g for 15 sec and the flow-through was discarde d. A second 500 l ali quot of buffer RPE (Qiagen) was added to the RNeasy spin co lumn, the sample was centrifuged at 8,000 x g for 2 min and the collection vessel and flow -through were discarded. The RNeasy spin column was transferred to a clean, sterile 1.5 ml microcentrifuge tube and 50 l water was added directly onto the RNeasy silica-gel membrane to elute RNA. The RNeasy spin column inside of the 1.5 ml microcen trifuge tube was centrifuged at 8,000 x g for 2 min and the RNeasy spin column was discarded. The collected RNA sample was placed on ice or stored at -70C pr ior to further analysis. DNase Treatment of RNA Extracts Prior to RT-PCR DNase treatment of collected RNA samples was performed using the DNase I Amplification Grade Kit (Invitrogen, Carlsb ad, CA). To a clean, sterile 0.5 ml microcentrifuge tube, 1 l 10X DNase I Reac tion Buffer (Invitrogen ), 2 l DNase I Amp Grade (1 U/l; Invitrogen) and 2 l DEPC-t reated water (ICN Biomedicals Inc.) was added. A 5 l aliquot of RNA was added and the solution was mixed using the action of the pipette and incubated for 15 min at room temperature. After incubation the reaction was stopped by addition of 1 l 25 mM EDTA Solution (Invitr ogen) and the mixture was

PAGE 72

59 heated at 65C for 10 min. DNase I-treated RNA samples were placed on ice prior to further analysis by RT-PCR. Induction and Expression of ipaH RNA in S. sonnei A five-strain S. sonnei cocktail was prepared as described below in the section ‘Inoculum Preparation’. The following test solu tions were prepared in clean, sterile 1.5 ml microcentrifuge tubes: 900 l SB + 100 l HeLa cell extract 1, 900 l SB + 100 l HeLa cell extract 2, 900 l SB + 100 l HeLa cell extract 3, 900 l SB + 100 l HeLa cell extract 4, 900 l SB + 100 l Congo red solution (see Appendix A) and 1000 l SB (control). A 10-l aliquot of the S. sonnei cocktail was added to each of the test solutions and the solutions were vortexed. Total RNA wa s extracted from a 250-l aliquot of each test solution and analyzed by RT-PCR using ipaH gene-specific primers. The remaining portions of the test solutions were incubated at 37C for 4 hr, after which the total RNA was extracted from a second 250-l aliquot of each test solution and analyzed by RTPCR using ipaH gene-specific primers. Identification of Shigella -Specific Genetic Loci Homologous gene cluster tables were created using the Microbial Genome Database for Comparative Analysis (MBGD; National Institute for Basic Biology, National Institutes of Natural Sciences, Japan; available at: http://mbgd.genome.ad.jp/ ) using the available genomes (as of December 2005) for Escherichia coli and Shigella spp. ( E. coli O157:H7, E. coli CFT073, E. coli K-12 W3110, E. coli K-12 MG1655, E. coli EDL933, S. flexneri 2a 2457T, S. flexneri 2a 301 and S. sonnei Ss046). Homologous genes were analyzed using ClustalW (protein -protein alignment f unction included in the MBGD) to identify potential ge nes for the detection of Shigella spp. or S. sonnei Specifically, those genes that co ntained sequences conserved among Shigella spp. or

PAGE 73

60 sequences unique to S. sonnei were identified as potential ge netic targets. The nucleotide sequences were obtained for the identified pot ential genetic targets and analyzed using the nucleotide-nucleotide Basic Local Alignmen t Search Tool (BLASTn; available at the National Center for Biotechnology Information website: http://www.ncbi.nlm.nih.gov/ ) to further identify se quences specific to Shigella spp. or S. sonnei The identified sequences were then used for primer/probe development as described below. Development of Primers/Probes for the Detection of Shigella The primers/probes investigated in th is study are listed in Table 3-2. All primers/probes were developed using the Beacon Designer 5 software (PREMIER Biosoft International, Palo Alto, CA). Fo r design of the 01-023 primers and probe, the conserved sequence from the chromosomally-located ipaH genes of S. sonnei Ss046 was identified using ClustalW. Default primer sett ings were used with one exception; the 3' maximum G was adjusted from 10 –kcal/mol to 4.0 –kcal/mol. For primers designed for the detection of all Shigella spp., regions with cross homology to the E. coli K-12 genome were avoided. For primers desi gned for the specific detection of S. sonnei regions with cross homology to the S. flexneri 2a 2457T genome were avoided. All hydrolysis probes were designed with the reporter dye FAM on the 5' end and the quenching dye TAMRA on the 3' end. All prim ers/probes were purchased from SigmaGenosys (The Woodlands, TX). Evaluation of Primer/Probe Specificity All primers and probes were evaluated for specificity in silico using BLASTn in addition to in-house testing against DNA from stock bacterial cultures (Table 3-3). Primer specificity was initially tested using real-time PCR followed by melt curve

PAGE 74

61 Table 3-2. Primers designed for the detection of Shigella. All primers were designed using Beacon Designer 5 software with gene sequences obtained from the genome of S. sonnei Ss046 (accession number CP000038). Designation Gene ID Sequence (5' 3') Product length (bp) 01-023F ipaH GTGAAGGAAATGCGTTTCTATG 106 01-023R ACCAGTCCGTAAATTCATTCTC 01-023P AGTGACAGCAAATGACCTCCGCA 01-024F SSO_0670 TTTCTAAAGTTGCAGTCACCTTTG 143 01-024R GGTGCCTAAAAC GATATTGCTTTG 01-024P AGCTCAGCGAAAACCACCGGCG 01-025F SSO_2067 GCCCCGCTACGCATGTC 80 01-025R GTGATCTCCA GTTCCGCAAATG 01-026F SSO_2071 ACAATC GAAGACATCGCGTTTC 179 01-026R CCAATCACTTTCTTGCCACTTTTC 01-027F SSO_1019 ACGCTTACAAGGCCATTATGAATG 96 01-027R CCTCAGCTTC AGATGCTTTATCAC 01-028F SSO_2059 AAACCACTCATCAAATACGAGAG 106 01-028R TTCGCAATGACCAGACCTAC 01-029F SSO_2863 CGGCTGGTTTGGCAAGTTAAG 165 01-029R TGGTTCACCCCATCAAGAACATC 01-030F SSO_2685 TGAGCCCGGACAGTTTCAC 94 01-030R TTGTATGTTACGTCGCTGAACAC 01-031F SSO_3247 CCCAACCATATTGACGTGTTCTTC 101 01-031R GCGTAGTTGCTGCCGTTAAC 01-032F SSO_0721 TTTATGACAGTTGCTGATTTCAAAC 180 01-032R ATACTCTTTC TGAGGATGAATGTTC analysis in 20 l reaction mixtures consis ting of: 10 l IQ™ Supermix with SYBR green (Bio-Rad, Hercules, CA), 200 nM (final c oncentration) each of forward and reverse primer, 2.0 l DNA sample and purified water. The PCR cycling conditions were 95C

PAGE 75

62 Table 3-3. Shigella and nonShigella strains tested for specificity. Designation Culture Origin KRS101 Shigella boydii serotype 18 ATCC 35966 KRS102 Shigella boydii serotype 18 Outbreak isolate KRS103 Shigella sonnei Patient isolate KRS104 Shigella sonnei Patient isolate KRS105 Shigella sonnei Outbreak isolate KRS107 Shigella flexneri Dr. K.A. Lampel, FDA KRS108 Shigella dysenteriae serotype 1 ATCC 9361 KRS109 Shigella sonnei ATCC 25931 KRS110 Shigella sonnei ATCC 29930 KRS111 Shigella sonnei ATCC 29030 KRS112 Shigella sonnei ATCC 9290 KRS113 Shigella sonnei ATCC 29031 KRS201 Salmonella Typhimurium ATCC 15277 KRS202 Salmonella Agona ATCC BAA-707 KRS203 Salmonella Gaminara ATCC BAA-711 KRS204 Salmonella Poona ATCC BAA-709 KRS205 Salmonella Montevideo ATCC BAA-710 KRS206 Salmonella Enteritidis Dr. G.E. Rodrick, University of Florida KRS207 Salmonella Agona LJH617 Dr. L.J. Harris, University of California, Davis KRS208 Salmonella Gaminara LJH618 Dr. L.J. Harris, University of California, Davis KRS209 Salmonella Michigan LJH621 Dr. L.J. Harris, University of California, Davis KRS210 Salmonella Montevideo LJH619 Dr. L.J. Harris, University of California, Davis KRS211 Salmonella Poona LJH630 Dr. L.J. Harris, University of California, Davis KRS212 Salmonella Enteritidis Environmental isolate KRS213 Salmonella Miami Environmental isolate KRS214 Salmonella spp. Ground beef isolate KRS215 Salmonella spp. Ground beef isolate KRS216 Salmonella spp. Ground beef isolate KRS217 Salmonella spp. Ground beef isolate KRS301 Escherichia coli K-12 Dr. A.C. Wright, University of Florida KRS302 Escherichia coli JM104 Dr. A.C. Wright, University of Florida KRS303 Escherichia coli ATCC 25922 KRS304 Escherichia coli O157:H7 FSIS 063-93 KRS305 Escherichia coli O157:H7 FSIS 413-95 KRS306 Escherichia coli O157:H7 GFP-85 Deibel Laboratories, Gainesville, FL

PAGE 76

63 Table 3-3. Continued. Designation Culture Origin KRS307 Escherichia coli O157:H7 ATCC 700599 KRS308 Escherichia coli Ground beef isolate KRS309 Escherichia coli Ground beef isolate KRS310 Escherichia coli Ground beef isolate KRS311 Escherichia coli Ground beef isolate KRS312 Escherichia coli Ground beef isolate KRS313 Escherichia coli Ground beef isolate KRS314 Escherichia coli Ground beef isolate KRS315 Escherichia coli Ground beef isolate KRS316 Escherichia coli Ground beef isolate KRS402 Acinetobacter lwoffii Tomato isolate KRS403 Acinetobacter anitratus Tomato isolate KRS404 Citrobacter freundii Cantaloupe isolate KRS405 Hafnia alvei Cantaloupe isolate KRS406 Klebsiella ozonae Cantaloupe isolate KRS407 Klebsiella pneumoniae Environmental isolate KRS408 Enterobacter spp. Tomato isolate KRS409 Enterobacter cloacae Cantaloupe isolate KRS410 Enterobacter agglomerans Tomato isolate KRS411 Serratia mercesens Cantaloupe isolate KRS412 Pseudomonas cepacia Tomato isolate KRS413 Pseudomonas maltophila Tomato isolate KRS414 Staphylococcus aureus 12293 Deibel Laboratories, Gainesville, FL KRS415 Staphylococcus aureus 25293 Deibel Laboratories, Gainesville, FL KRS416 Citrobacter freundii Ground beef isolate KRS417 Citrobacter freundii Ground beef isolate KRS418 Citrobacter freundii Ground beef isolate KRS419 Enterobacter spp. Ground beef isolate KRS420 Citrobacter freundii Ground beef isolate KRS421 Hafnia alvei Ground beef isolate KRS422 Serratia mercesens Ground beef isolate KRS423 Citrobacter freundii Ground beef isolate KRS424 Klebsiella pneumoniae Tomato isolate KRS425 Hafnia alvei Ground beef isolate KRS426 Hafnia alvei Ground beef isolate KRS427 Enterobacter cloacae Ground beef isolate for 1 min followed by 45 cycles of 95.0C fo r 10 sec then 60.0C for 30 sec. Following PCR amplification a melt curve was pe rformed from 60.0C to 95C with the temperature increasing at a rate of 0.5 C/10 sec. The annealing temperature was

PAGE 77

64 optimized for primer pairs showing initial specificity and the specificity testing was repeated using reaction mixtures similar to that described above except that the hydrolysis probe was added at 200 nM (final concentration) and IQ™ Supermix was used in place of IQ™ Supermix with SYBR green. All real-time PCR was performed using the iCycler (BioRad). Binding of Biotinylated Capture Probes to Streptavidin-Coated Paramagnetic Beads The capture probe was designed usin g the conserved sequence of the chromosomally-located ipaH genes of S. sonnei Ss046 and was purchased from SigmaGenosys (The Woodlands, TX). The sequen ce of the capture probe was tested for specificity in silico using BLASTn. To reduce steric hindrance during hybridization, the capture probe was constructed with a 12-car bon spacer between the 5' nucleotide and the biotin molecule as shown in Figure 3-1. A 200-l aliquot of Dynabeads M-280 Stre ptavidin (Dynal Biot ech; Oslo, Norway) was transferred to a clean, ster ile 1.5 ml microcentrifuge tube. The beads were washed as follows: two times in 200 l 2X Binding/Wash ing buffer (B/W buffer; see Appendix A), two times (minimum of 1-3 min each) in 200 l Dynabeads Solution A (see Appendix A), two times in Dynabeads Solution B (s ee Appendix A), one time in 200 l B/W buffer and finally resuspended in 400 l B/W buffer. A 40-l aliquo t of a 10 M capture probe solution and 360 l sterile water were added to the beads and the resulting mixture was incubated at room temperature with gentle end-over-end rotation for 10 min. The capture probe-bead complexes (hereafter referred to as CPShigella beads) were then washed three times in 400 l 1X B/W buffer a nd resuspended in 400 l 1X B/W buffer containing 1.5 l et hanolamine. The CPShigella beads were incubated at room

PAGE 78

65 Figure 3-1. Design of the capture probe and CPShigella beads. Biotinylated DNA probes were used to label streptavidin (SA) -coated paramagnetic beads for use in DNA sequence capture experiments. A 12-carbon spacer was inserted between the biotin molecule and the DNA probe sequence to alleviate steric hindrance during hybridization. temperature with gentle endover-end rotation for 1 hr, washed 3 times in 400l low-salt wash buffer (see Appendix A) and finally re suspended in 400 l low-salt wash buffer. CPShigella beads were stored at 4C prior to use. Inoculum Preparation Three days prior to each experiment, S. sonnei cultures were individually cultivated (37C, static incubation) in 10 ml tubes of TSCR and ove rnight transfers were performed using 10 l sterile, disposable l oops (BD Diagnostics) each day. On the day of the experiment, a five-strain S. sonnei cocktail was compiled by transferring 2.0 ml from each of the five 18-hr S. sonnei cultures (late stationary phase ) to a clean, sterile 15 ml centrifuge tube. The cocktail was centrifuged (3,220 x g for 10 min at 4C) and the resulting pellet was washed twice in 10 ml of 0.1% peptone The final cell titer of the cocktail inoculum was determined by pour plate using TSA. Calculation of Generation Time of S. sonnei in Shigella Broth Each S. sonnei strain was cultured and transferred in TSB (37 C) overnight for three days. On the third day, 10 l of an 18-hr culture was used to inoculate 100 ml sterile SB in a 250 ml Erlenmeyer flask. The SB flask was then swirled to disperse the

PAGE 79

66 inoculum and a 1.0 ml aliquot was used to prepare dilutions in 0.1% peptone water. Appropriate dilutions were an alyzed by pour plate using TSB to estimate the initial bacterial population. The SB flas k was incubated aerobically at 44 C without shaking and 1.0 ml samples were analyzed as described above every 30 min. Using a Microsoft Excel spreadsheet, the log bacteria l population vs. time was plotted to determine the length of lag phase and the generation ti me during exponential growth. Preliminary Experiments with AntiShigella Beads Two whole tomatoes and two 50 g ground beef samples were inoculated with the five-strain S. sonnei cocktail at approximately 4.3 x 105 CFU/tomato and 2.2 x 105 CFU/25 g, respectively. Two 25-g aliquots of ground beef were transferred to clean, sterile filtered stomacher bags each containi ng 225 ml SB and homogenized for 30 sec. The two tomato samples were transferred to clean, sterile stomacher bags each containing 250 ml SB and subjected to a 30 sec vigorous shake followed a 1 min hand manipulation. The filtered SB from the ground beef samples and the SB rinse from the tomato samples were transferred to clean, sterile stomacher bags and incubated at 44C for 18 hr. After enrichment, the samples were analyzed by FTI using antiShigella beads prepared using either AB02 or AB04. The recovered antiShigella beads were analyzed by spread plate using MAC (44C for 24 hr). The resulti ng colonies were identified based on biochemical reactions using the BBL Enterotube™ II (Becton Dickinson, Sparks, MD). Separation of S. sonnei from Food Matrices Using Low-Speed Centrifugation A five-strain cocktail of the rifampicin resistant S. sonnei was prepared as described above and diluted in 0.1% peptone water to a fi nal concentration of 3.3 x 105 CFU/ml. To test low-speed centr ifugation for the separation of S. sonnei from food matrices, five replicate 25-g samples of potato salad and ground beef were homogenized

PAGE 80

67 with 25 ml SB and transferred to clean, sterile 50 ml centrifuge tubes. The SB homogenates were then spiked with 1.0 ml of the diluted S. sonnei cocktail and the centrifuge tubes were mixed by vigorous shak ing for 30 sec. Prior to centrifugation, 1.0 ml aliquots of the homogenate were serially diluted using 9.0 ml 0.1% peptone tubes and 1.0 ml aliquots from appropriate dilutions we re analyzed by pour-p late using TSA rif+. The remaining SB homogenates were subjec ted to low-speed centrifugation (LSC; 100 x g for 5 min). After centrifugati on, 1.0 ml aliquots from the supernatant were serially diluted using 9.0 ml 0.1% peptone tubes and 1.0 ml aliquots from a ppropriate dilutions were analyzed by pour-plate using TSA rif+. Survival Studies Sample Inoculation and Subsequent Recovery Tomatoes were placed in sterile fibergla ss trays with the blossom scar faced up. Smooth surfaces around the blossom scar of tomato es were spot inoculated at 10 sites per fruit with 10 l per site using appropriate d ilutions to obtain final inoculation levels of approximately 5.0 x 105 CFU/tomato with five replicat e tomatoes at each level. The inoculated tomatoes were dried for at le ast 1 hr in a laminar flow hood at room temperature. After drying, tomatoes were st ored in humidity chamber at 13C with 85% relative humidity (RH) and five inoculated tomatoes were tran sferred to sterile Stomacher bags (Seward, Norfolk, UK) containing 100 ml of sterile PBS at each observation day. For recovery of inocula, tomatoes were sh aken vigorously for 30 sec, then massaged by hand for 1 min in the stomach bag similar to that described by Zhuang et al (1995). For trials involving potato salad and gr ound beef, 50-g samples were aseptically weighed into sterile 4-oz specimen cups and inoculated with 1.0 ml of an appropriate dilution to obtain final inoc ulation levels of ca 1.0 x 106 CFU/g. Using sterile tongue

PAGE 81

68 depressors, the inoculum was homogenized in each sample for a minimum of 30 sec, after which the samples were allowed to sit at room temperature for 1 hr for bacterial attachment. After attachment, each set of sp ecimen cups with samples were stored at 2.5C and 8C, respectively. For recovery of i nocula, sterile tongue depressors were used to weigh 25 g aliquots from each sample to st erile, clean Stomacher bags after which 225 ml of PBS was added and the samples were placed in the stomacher (Tekmar Company, Cincinnati, OH) for 30 sec. Three-Tube Most Probable Numb er Estimation of Survivors Recovered inocula in 100 ml PBS rinses of tomatoes represented a 0.01 dilution of the tomatoes surface populat ion. Recovered inocula in 225 ml PBS homogenates of potato salad or ground beef samples repres ented a 0.1/g dilution of the surviving population. Recovered inocula were serially diluted using 9.0 ml 0.1% peptone tubes and 1.0 ml aliquots from appropriate dilutions were used to inoc ulate each of three TSB rif+ tubes. To enumerate survivors at the 0.1/toma to level, three 10 ml aliquots from the 100 ml PBS tomato rinse were used to inocul ate 10 ml double-strength TSB rif+ tubes. All TSB rif+ tubes were incubated overnight (37 C; static incubation) and scored as either “positive” or “negative” for Salmonella or S. sonnei based on the presence/absence of visible growth. Surviving populations were es timated using the thre e-tube most probable number (MPN) table located in Appendix 2 of the FDA BAM (2003). Non-inoculated samples of each food were analyzed to c onfirm the absence of indigenous microflora with resistance to rifampicin at 100 ppm. Evaluation of Detection Methods A schematic representation of the experime ntal design is presented in Figure 3-2. Tomato smooth surfaces, potato salad and ground beef samples were inoculated with a

PAGE 82

69 Figure 3-2. Flow diagram of the experiments involving inoculated food samples. Each food samples was analyzed by the Shigella culture method of the FDA Bacteriological Analytical Manual (BAM), by flow-through immunocapture followed by analysis of recovered beads by spread plate using MacConkey agar (MAC) or using real-time PCR a nd by DNA sequence capture. Suspected isolates on MAC were further analyzed using Triple Sugar Iron (TSI) agar slants, Lysine Iron Agar (LIA) slan ts and Motility Medium (MM). five-strain S. sonnei cocktail and recovery of the inoc ula was investigated using the BAM Shigella culture method, FTI-MAC, FTI-PCR and DSC. Inoculated Tomatoes, Potato Salad and Ground Beef Flow-Through Immunocapture Method DNA Sequence Capture Method BAM Shigella Culture Method Real-time PCR Pathatrix (37C, 20 min) Shigella Broth (44C, 18 hr) MAC (37C, 20 min) TSI, LIA, MM (37C, 24 hr) Shigella Broth (44C, 18 hr) Shigella Broth (44C, 18 hr, anaerobic) DNA Extraction Centrifugation and Cell Lysis Real-time PCR Biochemical Identification (37C, 24 hr) Hybridization (40C, 1 hr)

PAGE 83

70 Inoculation of Samples and Subsequent Recovery For tomato samples, tomatoes were placed in sterile fiberglass trays with the blossom scar faced up. Smooth surfaces around th e blossom scar of tomatoes were spot inoculated at 10 sites per fruit with 10 l per site using appropriate dilutions of the S. sonnei cocktail to obtain final inoculation levels from 104 to 100 CFU/tomato with ten replicate tomatoes at each level. The inocul a were allowed to dry completely at room temperature. After drying, inoculated tomatoes were transferred to sterile Stomacher bags (Seward) containing 250 ml of sterile SB pre-warmed to 44C. For recovery of S. sonnei the stomacher bags were sealed using stomacher bag clips and the tomatoes were shaken vigorously for 30 sec then massaged by hand fo r 1 min, similar to the method described by Zhuang et al (1995). For potato salad and ground beef samples, 50 g of sample was weighed into a sterile 4-oz specimen cup and inoculated with appropriate dilutions of the S. sonnei cocktail to obtain final inoculation levels from 102 to 100 CFU/25 g with ten replicate samples at each level. Sterile wooden tongue depressors were used to homogenize the inoculum in each specimen cup. The inocul ated potato salad and ground beef samples were allowed to sit for 1 hr at room temper ature for bacterial attachment. A 25-g aliquot from each sample was weighed into a steril e dual-chambered stomacher bag containing 225 ml of SB and homogenized for 30 sec. Modified BAM Culture Method for S. sonnei For tomato samples, the recovered inocula in SB was transferred to sterile 18 oz Whirl-Pak bags (Nasco, Modesto, CA) and the top of the bag was folded over twice and secured. For potato salad and ground beef samp les, the internal filter of the dualchambered stomacher bag was used to transfer the SB supernatant on ly to sterile 18 oz

PAGE 84

71 Whirl-Pak bags and the top of the bag was folded over twice and secured. Samples were incubated at 44C for 18-24 hr under anaerobic conditions using 7.0 liter rectangular jars (Mitsubishi Gas Chemical Company, Inc., Ja pan) with the Pack-Anaero anaerobic gas generating system (Mitsubishi Gas Chemi cal Company). SB enrichments were homogenized by hand and streaked for isol ation on MAC and incubated at 37C for 1824 hr. For confirmation, three isolates colonies demonstrating typical Shigella morphology were selected from each MAC plate. When three typical isolates were not present, atypical isolates we re selected for confirmation. Each selected isolate was used to inoculate a TSI slant, a LIA slant and MM and incubated overnight (37C). From each isolate that resulted in typical reactions for Shigella on TSI slants, LIA slants and MM, growth from the TSI slant was cultivated overnight (37C) in TSB. Growth from the TSB tubes was then streaked for isolation on MAC and incubated overnight (37C). Biochemical reactions were tested using the BBL Enterotube™ II (Becton Dickinson). Enterotubes were incubated overnight (37C) and positive reactions were read according to manufacturer’s instructions. Flow-Through Immunocapture (FTI) Using the Pathatrix All samples were inoculated and enriched as described above for the modified BAM method with the following exceptions. SB tomato rinses and the filtered SB supernatant of homogenized potato salad and ground beef samples were transferred to clean, sterile stomacher bags prior to overn ight enrichment. Anaerobic conditions were not generated for any samples analyzed by FTI. Instead Stomacher clips were used to seal the Stomacher bags and samples we re incubated at 44C (static).

PAGE 85

72 All FTI experiments were performed using the Pathatrix system (Matrix MicroScience, Golden, CO). The assembly/opera tion of the Pathatrix and the recovery of antiShigella beads were performed as per manufactur er instructions. Briefly, Stomacher bags containing sample enrichments were plac ed into a Pathatrix incubation pot. After the tubing was properly placed, a 50-l aliquot of antiShigella beads was injected to the system per manufacturer’s in structions. After 20 min circ ulation at 37C, the tubing assembly was disconnected from the sample and the antiShigella beads were washed with 100 ml of PBS and recove red in the collection vessel suspended in 5 to 10 ml PBS. Using a magnet, the antiShigella beads were drawn to the side of the collection vessel and the volume of PBS was reduced to appr oximately 250 l using a sterile transfer pipet. A 50-l aliquot of the resuspended antiShigella beads was analyzed by spread plate using MAC. MAC plates were incuba ted at 37C for 24 hr. In addition, DNA was extracted from a 100-l aliquot of the resuspended antiShigella beads as described above and analyzed by real-time PCR. Sequence Capture of Shigella DNA A schematic representation of the DS C method is presented in Figure 3-3. Preliminary experiments were performed to evaluate hybridization buffers, hybridization temperatures, type of streptavidin-coa ted paramagnetic beads and sensitivity. All procedures for inoculation and rec overy of food samples was followed as described above for FTI samples. After 1824 hr incubation, the samples were shaken briefly to mix contents. For potato salad a nd ground beef samples, a 10 ml aliquot was aseptically transferred to a clean, sterile 15 ml centrifuge tube and solid food material was sedimented using lowspeed centrifugation (100 x g for 5 min). For tomato samples, no solid food material was pres ent in the overnight enrichments, therefore no low-speed

PAGE 86

73 Figure 3-3. Flow diagram of the DNA sequence capture method. CPShigella beads were prepared using Dynabeads M-280 Streptavid in coated with a 5' biotin-labeled DNA probe with specificity for the ipaH gene of Shigella and enteroinvasive E. coli Inoculated Potato Salad or Ground Beef Shigella Broth (44C, 18 hr, Filtered stomacher bag) Inoculated Tomatoes Low-Speed Centrifugation (100 x g 5 min) High-Speed Centrifugation (6,000 x g, 5 min) Resuspend Pellet in Hybridization Buffer 1 Heat (100C, 10 min) Cool on ice (10 min) Discard Supernatant Shigella Broth (44C, 18 hr) 1.0 ml 10 ml Filtrate High-Speed Centrifugation (6,000 x g, 5 min) Discard Pellet Discard Pellet Hybridization (40C, 1 hr) Wash CPShigella Beads Real-time PCR

PAGE 87

74 centrifugation was performed. A 1.0-ml aliquot of the resulting supern atant (potato salad and ground beef samples) or a 1.0-ml aliquot of the overnight enrichment (tomato samples) was transferred to a clean, sterile 1.5 ml microcentrifuge tube. The bacterial cells were then sedimented using centrifugation (6,000 x g for 5 min) and the supernatant was discarded. The pellet was resuspende d in 530 l hybridization buffer 1 (see Appendix A), heated at 100C for 10 min and finally cooled on ice for 10 min. Cellular and solid material were sedimented by centrifugation (6,000 x g for 5 min) and the supernatant was transferred to a clean, sterile 1.5 ml microcentrifuge tube containing 200 l 3.75 M NaCl and 20 l CPShigella beads. The contents we re mixed using the action of the pipette and heated at 40C for 5 min on a dry-bath incubator followed by end-overend rotation (hybridization) at 40C for 1 hr. Heated end-over-end rotation was achieved using a Rugged Rotator (Glas-Col, Terre Haute, IN) inside of an environmental chamber (Lab-Line, E2 series, Barnstead Intern ational, Melrose Park, IL). Following hybridization, the beads were r ecovered using a magnetic rack and washed 2 times in 200 l wash buffer (see Appendix A), 2 times in 200 l low-salt wash buffer (see Appendix A) and resuspended in 50 l TE buffer. The re suspended beads were then heated at 75C for 10 min to release the captured DNA from the probe and, using the magnetic rack, the DNA-containing supernatant was transferred to a clean, sterile 1.5 ml microcentrifuge tube. The DNA samples were placed on ice or frozen (-20C) prior to analysis by PCR. Real-Time PCR and Reverse Transcriptase (RT) PCR All real-time PCR and real-time RT-PCR analyses were performed using the iCycler (Bio-Rad, Hercules, CA). For the analys is of FTI samples, real-time PCR was performed using 50 l reaction mixtures cons isting of: 25 l IQ™ Supermix (Bio-Rad), 200 nM (final concentration) each of fo rward and reverse primer, 200 nm (final

PAGE 88

75 concentration) hydrolysis probe, 20 l DNA sample and purified water. The PCR cycling conditions were 95C for 1 min follo wed by 40 cycles of 95C for 10 sec then 60C for 30 sec. For the analysis for DNA se quence capture samples, real-time PCR was performed using 20 l reaction mixtures cons isting of: 10 l IQ™ Supermix (Bio-Rad), 200 nM (final concentration) each of fo rward and reverse primer, 200 nm (final concentration) hydrolysis probe, 5 l DNA sa mple and purified water. The PCR cycling conditions were the same as for FTI samples. Real-time RT-PCR was performed using 20 l reaction mixtures consisting of: 10 l 2X RT-PCR Reaction Mix for Probes (Bio-Rad ), 200 nM (final con centration) each of forward and reverse primer, 200 nM (final con centration) hydrolysis probe, 0.4 l reverse transcriptase (Bio-Rad), 0.3 l RNasin (Pro mega, Madison, WI), 5 l RNA sample and purified water. The PCR cycling conditions were 50C for 10 min, 95C for 5 min followed by 40 cycles of 95C for 10 sec then 60C for 30 sec. For all RNA samples analyzed by RT-PCR, parallel reaction mi xtures without the addition of reverse transcriptase were prepared to ve rify complete digestion of DNA. All PCR and RT-PCR reaction mixtures we re prepared in a Labconco Purifier Class II safety cabinet (Labconco Corporation, Kansas City, MO). Recording of Data and Statistical Analysis All statistical analyses of survival st udies were performed using the Statview statistical software package (SAS) version 9.1 (SAS Institute Inc., Cary, NC) using a mixed model. Sample replications were treated as random variables within time. Statistical analysis of population means in experiments involving LSC were performed by hand using a two sample t-test. P values < 0.05 were considered significant.

PAGE 89

76 All results from the evaluation of detection methods were recorded as “positive” or “negative” for the detection of S. sonnei Positive isolation on plating media resulted from typical reactions for S. sonnei in all confirmation steps and positive biochemical BBL Enterotube™ II identification. Bias-reduce d logistic regression (BRLR) models were constructed using the R software (T he R Foundation for Statistical Computing, Version 2.2.0, http://cran.us.r-project.org/ ) to identify significant differences ( P < 0.05) among the detection methods tested.

PAGE 90

77 CHAPTER 4 RESULTS This study consisted of three phases of research. The first phase consisted of preliminary trials involving the preparation of growth curves to cal culate the generation time of S. sonnei in SB, the development of primers/probes for the detection of S. sonnei and testing the specificity of each set of pr imers/probes against a DNA library of positive and negative controls. The second phase cons isted of experiments designed to test the ability of S. sonnei to survive in/on selected foods. Th e third phase of this study involved the evaluation of newly developed sampling me thods for the detection of artificially inoculated S. sonnei on smooth tomato surfaces and in potato salad and ground beef. Recovery/detection of the inoc ula was tested using the BAM Shigella culture method and the newly developed methods: flow-through i mmunocapture (FTI) followed by analysis of recovered antiShigella beads by spread-plate using MAC (FTI-MAC), FTI followed by analysis of recovered antiShigella beads by real-time PCR (FTI-PCR) and DNA sequence capture (DSC). Preliminary Studies Calculation of Generation Time of S. sonnei in Shigella Broth Growth curves were prepared for S. sonnei ATCC 9290, S. sonnei ATCC 29031, S. sonnei ATCC 29030, S. sonnei ATCC 25931 and S. sonnei ATCC 29930 (Figures 4-1, 42, 4-3, 4-4 and 4-5, respectively) in Shigella broth (SB) incubate d aerobically without shaking at 44C. The length of the lag phase was identified and the exponential phase was used to calculate the generation time fo r each strain. The average growth kinetics

PAGE 91

78 among the five S. sonnei strains was a lag phase of approximately 2 hr and a generation time of 18.8 0.6 min. Growth curve of S. sonnei ATCC 9290 When the growth of S. sonnei ATCC 9290 was investigated in SB (44C), an initial lag phase of 2 hr was observed prior to e xponential growth (Figure 4-1). Logarithmic regression used to analyze the exponentia l phase of growth showed linearity (R2 = 0.997) and the equation of the line was used to calculate a generation time of 19.0 min. The initial population of S. sonnei ATCC 9290 was 4.3 x 104 CFU/ml and the final population was 3.8 x 108 CFU/ml. Growth curve of S. sonnei ATCC 29031 When the growth of S. sonnei ATCC 29031 was investigated in SB (44C), an initial lag phase of 2 hr was observed prior to exponen tial growth (Figure 4-2). Logarithmic regression used to analyze the e xponential phase of grow th showed linearity (R2 = 0.994) and the equation of the line was us ed to calculate a ge neration time of 19.5 min. The initial population of S. sonnei ATCC 29031 was 3.8 x 104 CFU/ml and the final population was 2.4 x 108 CFU/ml. Growth curve of S. sonnei ATCC 29030 When the growth of S. sonnei ATCC 29030 was investigated in SB (44C), an initial lag phase of 2 hr was observed prior to exponen tial growth (Figure 4-3). Logarithmic regression used to analyze the e xponential phase of grow th showed linearity (R2 = 0.996) and the equation of the line was us ed to calculate a ge neration time of 18.6 min. The initial population of S. sonnei ATCC 29030 was 4.0 x 104 CFU/ml and the final population was 5.0 x 108 CFU/ml.

PAGE 92

79 y = 0.9511x + 2.8482 R2 = 0.997 4.00 5.00 6.00 7.00 8.00 9.00 012345678 Time (hr)Log10 CFU/m l Figure 4-1. Growth curve: S. sonnei ATCC 9290 in Shigella broth. A 100 ml microcosm was inoculated with a 10-l aliquot of an 18-hr S. sonnei ATCC 9290 culture and incubated (44C, static). At approp riate time intervals, a 1.0 ml aliquot was serially diluted in 0.1% peptone and the population was estimated by pour plate using tryptic soy agar. ( ) lag/stationary phase growth; ( ) exponential phase growth. Error bars repr esent one standard deviation. y = 0.9271x + 2.5332 R2 = 0.994 4 5 6 7 8 9 012345678 Time (hr)Log10 CFU/m l Figure 4-2. Growth curve: S. sonnei ATCC 29031 in Shigella broth. A 100 ml microcosm was inoculated with a 10-l aliquot of an 18-hr S. sonnei ATCC 29031 culture and incubated (44C, static). At approp riate time intervals, a 1.0 ml aliquot was serially diluted in 0.1% peptone and the population was estimated by pour plate in tryptic soy agar. ( ) lag/stationary phase growth; ( ) exponential phase growth. Error bars repr esent one standard deviation.

PAGE 93

80 y = 0.9734x + 2.5896 R2 = 0.996 4 5 6 7 8 9 012345678 Time (hr)Log10 CFU/m l Figure 4-3. Growth curve: S. sonnei ATCC 29030 in Shigella broth. A 100 ml microcosm was inoculated with a 10-l aliquot of an 18-hr S. sonnei ATCC 29030 culture and incubated (44C, static). At approp riate time intervals, a 1.0 ml aliquot was used to make serial dilutions in 0.1% peptone and the population was estimated by pour plate using tryptic soy agar. ( ) lag/stationary phase growth; ( ) exponential phase growth. Error bars represent one standard deviation. Growth curve of S. sonnei ATCC 25931 When the growth of S. sonnei ATCC 25931 was investigated in SB (44C), an initial lag phase of 2 hr was observed prior to exponen tial growth (Figure 4-4). Logarithmic regression used to analyze the e xponential phase of grow th showed linearity (R2 = 0.990) and the equation of the line was us ed to calculate a ge neration time of 18.0 min. The initial population of S. sonnei ATCC 25931 was 2.9 x 104 CFU/ml and the final population was 3.7 x 108 CFU/ml. Growth curve of S. sonnei ATCC 29930 When the growth of S. sonnei ATCC 29930 was investigated in SB (44C), an initial lag phase of 2 hr was observed prior to exponen tial growth (Figure 4-5). Logarithmic regression used to analyze the e xponential phase of growth showed linearity

PAGE 94

81 y = 1.0243x + 2.1497 R2 = 0.990 4 5 6 7 8 9 012345678 Time (hr)Log10 CFU/m l Figure 4-4. Growth curve: S. sonnei ATCC 25931 in Shigella broth. A 100 ml microcosm was inoculated with a 10-l aliquot of an 18-hour S. sonnei ATCC 25931 culture and incubated (44C, static). At appropriate time intervals, a 1.0 ml aliquot was used to make serial dilu tions in 0.1% peptone and the population was estimated by pour plate using tryptic soy agar. ( ) lag/stationary phase growth; ( ) exponential phase growth. Error bars represent one standard deviation. y = 0.9452x + 2.6999 R2 = 0.998 4.00 5.00 6.00 7.00 8.00 9.00 012345678 Time (hr)Log10 CFU/m l Figure 4-5. Growth curve: S. sonnei ATCC 29930 in Shigella broth. A 100 ml microcosm was inoculated with a 10-l aliquot of an 18-hour S. sonnei ATCC 29930 culture and incubated (44C, static). At appropriate time intervals, a 1.0 ml aliquot was used to make serial dilu tions in 0.1% peptone and the population was estimated by pour plate using tryptic soy agar. ( ) lag/stationary phase growth; ( ) exponential phase growth. Error bars represent one standard deviation.

PAGE 95

82 (R2 = 0.998) and the equation of the line was us ed to calculate a ge neration time of 19.1 min. The initial population of S. sonnei ATCC 29930 was 1.6 x 104 CFU/ml and the final population was 4.3 x 108 CFU/ml. Evaluation AntiShigella Antibodies for Use with Flow-Through Immunocapture Two polyclonal antibodies (AB01 a nd AB04) and one monoclonal antibody (AB02) were evaluated for use in FTI of S. sonnei using the Pathatrix (Table 4-1). AB01 was generated using a mixture of S. boydii, S. flexneri and S. dysenteriae as the immunogens. AB04 was generated using a membrane extract mixture of S. sonnei and S. flexneri In addition, a seconda ry capture in which S. sonnei pre-bound with AB01 or AB02 in solution were separated from food ma trices using FTI with paramagnetic beads coated with goat anti-rabbit antibodies or rabbit anti-mous e antibodies (AB03 or AB05, respectively). The number of S. sonnei colonies which resulted from FTI followed by analysis of recovered beads by spread pl ate using MAC using antiShigella beads prepared with the various antiShigella antibodies is listed in Table 4-1. When AB01 was used for the preparation of antiShigella beads, a population too nume rous to count (TNTC) of S. sonnei ATCC 25931, two colonies of S. sonnei ATCC 29930, 78 colonies of S. sonnei ATCC 9290 and no colonies of S. sonnei ATCC 29030 or S. sonnei ATCC 29031 were observed. When AB02 was used for the preparation of antiShigella beads, a population TNTC of S. sonnei ATCC 25931 and S. sonnei ATCC 9290, 147 colonies of S. sonnei ATCC 29930, 112 colonies of S. sonnei ATCC 29030 and 87 colonies of S. sonnei ATCC 29031 was observed. When AB04 was used for the preparation of antiShigella beads, 261 colonies of S. sonnei ATCC 25931, 182 colonies of S. sonnei ATCC 29930, 15 colonies of S. sonnei ATCC 29030, 53 colonies of S. sonnei ATCC 9290 and

PAGE 96

83 Table 4-1. Evaluation of antiShigella antibodies for flow-thr ough immunocapture (FTI) of S. sonnei Microcosms of individual S. sonnei strains were prepared at approximately 2.0 x 105 CFU/ml in 250 ml phosphate buffered saline (PBS). The microcosms were then analyzed by FTI for 30 min at 37C using various antiShigella beads. After circulation, 100 l aliquots of the recovered beads were analyzed by spread plate using MacConkey agar (MAC; 37C for 24 hr). The numbers of resulting colonies were enumerated in order to compare immunocapture by the various antibodi es. In the AB03-AB03 and AB05AB02 experiments, AB01 and AB02 were added directly to inoculated PBS and allowed to bind for 5 min with sh aking (60 rpm) followed by FTI using beads coated with AB03 and AB05, respectively. Plate counts of S. sonnei strains Antibodies ATCC 25931 ATCC 29930 ATCC 29030 ATCC 9290 ATCC 29031 AB01 TNTCa 2 NGb 78 NG AB02 TNTC 147 112 TNTC 87 AB03-AB01 148 263 TNTC 68 TNTC AB04 261 182 15 53 47 AB05-AB02 TNTC NG NG 10 NG a too numerous to count b no growth 47 colonies of S. sonnei ATCC 29031 was observed. These data suggest that antiShigella beads prepared with AB02 or AB04, but not AB01, may be used for the consistent immunocapture of S. sonnei by FTI. When AB01 was used to bind S. sonnei prior to analysis by FTI using paramagnetic beads coated with AB03 (secondary capture), populations TNTC of S. sonnei ATCC 29030 and S. sonnei ATCC 29031, 148 colonies of S. sonnei ATCC 25931, 263 colonies of S. sonnei ATCC 29930 and 68 colonies of S. sonnei ATCC 9290 were observed. When AB02 was used to bind S. sonnei prior to analysis by FTI using paramagnetic beads coated with AB05 (secondary cap ture), populations TNTC of S. sonnei ATCC 25931, 10 colonies of ATCC 9290 and no colonies of S. sonnei ATCC 29930, ATCC 29031 or ATCC 29030 were observed. These data suggest that secondary capture by FTI using

PAGE 97

84 AB03 of S. sonnei which have been prior labeled with AB01 may be used for the consistent immunocapture. Secondar y capture by FTI using AB05 of S. sonnei that had been prior labeled with AB02 did not pr ovide consistent immunocapture under the conditions investigated. Using the Pa thatrix for secondary capture of S. sonnei that had been prior labeled with antiShigella antibodies required addi tional steps and time over using the Pathatrix for primary capture. Preliminary Experiments with AntiShigella Beads Tomato and ground beef samples were inoculated with a five-strain S. sonnei cocktail and analyzed by FTI followed by analys is of the recovered beads by spread plate using MAC. When antiShigella beads were prepared with AB02, analysis of tomato and ground beef samples resulted in MAC plates w ith only colonies exhi biting morphologies typical for S. sonnei and each colony selected (three colonies from each plate) for confirmation was identified as S. sonnei When antiShigella beads were prepared with AB04, analysis of tomato and ground beef sa mples resulted in MAC plates with colony morphologies both typical and atypical for S. sonnei Atypical colonies from tomato samples were identified as Citrobacter freundii Enterobacter spp. and Klebsiella pneumoniae Atypical colonies from ground b eef samples were identified as Enterobacter cloacae and E. coli Typical colonies from toma to samples were identified as S. sonnei or E. coli These data suggest that antiShigella beads prepared with AB02, but not those prepared with AB04, may be used for the spec ific detection of S. sonnei in tomato and ground beef samples. AntiShigella beads prepared with AB02 were used for all subsequent experiments (her eafter referred to as antiShigella beads). The manufacturer of AB02 reported no reactivity with S. boydii S. flexneri or S. dysenteriae however when antiShigella beads were tested against solutions of these

PAGE 98

85 serogroups in PBS using the FT I-MAC method (as described for S. sonnei strains) the antiShigella beads reacted strongly with the S. flexneri strain (KRS106) and the S. dysenteriae type 1 strain (KRS108). When test ed against two st ock strains of S. boydii type 18, antiShigella beads reacted weakly with one strain (KRS101) but did not react with the other strain (KRS102). Optimization of AntiShigella Bead Concentration for Flow-Through Immunocapture of S. sonnei To optimize the concentration of antiShigella beads required for detection of S. sonnei by FTI, antiShigella beads were diluted in PBS containing 0.1% sodium azide and tested against various concentrations of S. sonnei ATCC 25931 or S. sonnei 29930 (Table 4-2). Twenty-five-l aliquots of the initial concentration of antiShigella beads (50 mg/ml) and the following diluted concentrations were tested for FTI: 25 mg/ml (1:1 dilution), 16.7 mg/ml (1:2 dilution), 12.5 mg /ml (1:3 dilution) and 10 mg/ml (1:4 dilution). The number of colonies on MAC plat es which resulted from the analysis of various concentrations of S. sonnei ATCC 25931 and S. sonnei ATCC 29930 by FTI using the various conc entrations of antiShigella beads are given in Table 4-2. When S. sonnei ATCC 25931 microcosms of 2.8 x 103 CFU/ml and 3.7 x 105 CFU/ml were analyzed by FTI, all tested dilutions of antiShigella beads resulted in MAC plates with populations TNTC. For S. sonnei ATCC 25931 microcosms of 3.2 x 101 CFU/ml analyzed by FTI, undiluted antiShigella beads and dilutions of 1:1, 1:2, 1:3 and 1:4 antiShigella beads resulted in MAC plates with 27, 18, 13, 6 and 12 colonies, respectively. These data suggest the sensitivity of FTI for S. sonnei ATCC 25931 is approximately 3.0 x 101 CFU/ml and that any of the tested concentration of antiShigella beads may be used for detection.

PAGE 99

86 Table 4-2. Optimization of antiShigella bead concentration for flow-through immunocapture (FTI) of S. sonnei Microcosms of S. sonnei ATCC 25931 or ATCC 29930 at various concentrations were prepared in phosphate buffered saline (PBS). The microcosms were then analyzed by FTI for 30 min at 37C using 25-l aliquots of various dilutions of antiShigella beads. After circulation, 100-l aliquots of the rec overed beads were analyzed by spread plate using MacConkey agar (MAC; 37 C for 24 hr). The numbers of resulting colonies were enumerated in order to determine the sensitivity of FTI using various conc entrations of antiShigella beads. Plate counts from dilution of antiShigella beads S. sonnei titer (CFU/ml) None 1:1 1:2 1:3 1:4 ATCC 25931 3.2 x 101 27 18 13 6 12 2.8 x 103 TNTCa TNTC TNTC TNTC TNTC 3.7 x 105 TNTC TNTC TNTC TNTC TNTC ATCC 29930 2.1 x 101 NGb NG NG NG NG 2.7 x 103 NG NG NG NG NG 2.8 x 105 11 7 13 12 12 a too numerous to count b no growth When S. sonnei ATCC 29930 microcosms of 2.1 x 101 CFU/ml and 2.7 x 103 CFU/ml were analyzed by FTI, all tested dilutions of antiShigella beads resulted in MAC plates with no colonies. For S. sonnei ATCC 29930 microcosms of 2.8 x 105 CFU/ml analyzed by FTI, undiluted antiShigella beads and dilutions of 1:1, 1:2, 1:3 and 1:4 antiShigella beads resulted in MAC plates with 11, 7, 13, 12 and 12 colonies, respectively. These data suggest the sensitivity of FTI for S. sonnei ATCC 29930 is approximately 3.0 x 105 CFU/ml and that any of the tested concentration of antiShigella beads may be used for detection. Although the tested con centrations of antiShigella beads did not affect detection of S. sonnei ATCC 29531 or ATCC 29930 in PBS, the addition of 25 l antiShigella bead aliquots per FTI analys is of ground beef samples resulted in poor visual bead recovery.

PAGE 100

87 To improve visual bead recovery, 50-l aliquots of the 10 mg/ml antiShigella bead dilution were used for FTI in all inoculated trials. Identification of Potentially Shigella -Specific Genetic Loci The MBGD and BLASTn were used to identif y genetic loci potentially specific for Shigella spp. or for S. sonnei alone. The genes identified using the MBGD are listed in Table 4-3. Using BLASTn, none of the genes were identified with spec ificity for all five of the Shigella genomes within the database (T able 4-4), however the genes SSO_0670, SSO_2685, SSO_3247 and SSO_0721 were identified as potentially specific for some species of Shigella When analyzed using BLAS Tn, the genes SSO_2067, SSO_2071, SSO_1019, SSO_2059 and SSO_2863 were identifi ed as potentially specific for S. sonnei (Table 4-4). The nucleotide sequences of the identified genes were used to develop PCR primers (Table 3-2). Specificity of Primers Developed for Potentially Shigella -Specific Genetic Loci The developed primers were evaluated by real-time PCR against DNA extracted from stock Shigella spp. (Table 4-4) and closelyrelated microorganisms (data not shown). Primer sets 01-025, 01-026, 01028 and 01-029 amplified DNA from eight strains of S. sonnei and one strain each of S. flexneri and S. dysenteriae however DNA from two strains of S. boydii were not amplified. The obs erved amplification of DNA from the stock strains of S. flexneri and S. dysenteriae were not in agreement with the in silico analysis using BLASTn. Primer set 01-027 amplified DNA from only three of five S. sonnei strains tested; therefore it was not investigated against DNA from other Shigella

PAGE 101

88 Table 4-3. Genetic targets identified with potential specificity for Shigella spp. or for S. sonnei alone. Genes were identified using the Microbial Genome Database for Comparative Analysis (MBGD) to creat e orthologous gene tables from the available sequences of S. flexneri S. sonnei and E. coli All Gene IDs listed below are from the genome of S. sonnei Ss046 (accession number CP000038). Gene ID Gene Name Gene Description Primer Set SSO_0670 ybgD Putative fimbrial-like protein 01-024 SSO_2067 pduL Putative propanediol utilization protein 01-025 SSO_2071 pduP Putative propanediol utilization protein: CoA01-026 dependent propionaldehyde dehydrogenase SSO_1019 n/a Putative membrane protein 01-027 SSO_2059 pduB Putative propanediol utilization protein: 01-028 polyhedral bodies SSO_2863 n/a Putative periplasmic or exported protein 01-029 SSO_2685 n/a Conserved hypothetical protein 01-030 SSO_3247 n/a Putative minor pilin and initiator 01-031 SSO_0721 n/a Conserved hypothetical protein 01-032 spp. These data demonstrate that prim er sets 01-025, 01-026, 01-028 and 01-029 are not specific for S. sonnei Primer sets 01-030, 01-031 and 01-032 were tested DNA extracted from each of the Shigella spp. and E. coli isolates listed in Table 3-3. Pr imer set 01-030 performed in agreement with the in silico analysis with respect to the Shigella spp. tested. Primer sets 01-031 and 01-032, although not specific for S. dysenteriae by in silico analysis, resulted in positive amplification when tested against extracted DNA from the stock S. dysenteriae type 1 culture (Table 4-4). When tested against E. coli DNA however, primer sets 01-030 and 01-031 amplified DNA from all of the E. coli while primer set 01-032 amplified DNA from only two strains of E. coli (KRS312 and KRS316). These data suggest that primer sets 01-030, 01-031 and 01-032 may not be used for the specific detection of Shigella spp.

PAGE 102

89Table 4-4. Evaluation of primer specificity among stock Shigella cultures and by comparative analys is against previously sequenced Shigella genomes. Extracted DNA from stock Shigella cultures were analyzed by real-time PCR. BLAST searches for significant alignments were performe d using the National Center for Bi otechnology Information (NCBI) website. Primer set Species Strain identification/ Accession number 01-02401-02501-02601-027 01-02801-02901-03001-03101-032 Specificity by PCR against extracted DNA S. boydii KRS101 ND + + S. boydii KRS102 ND + + S. sonnei KRS103 + + + ND + + + + + S. sonnei KRS104 + + + ND + + + + + S. sonnei KRS105 + + + ND + + + + + S. sonnei KRS109 + + + + + + + + + S. sonnei KRS110 + + + + + + + + S. sonnei KRS111 + + + + + + + + + S. sonnei KRS112 + + + + + + + + S. sonnei KRS113 + + + + + + + + + S. flexneri KRS106 + + + ND + + + + + S. dysenteriae KRS108 + + + ND + + + + + Specificity by in silico analysis (BLAST) S. boydii NC_007613 + + S. sonnei NC_007384 + + + + + + + + + S. flexneri NC_004741 + + + + S. flexneri NC_004337 + + + + S. dysenteriae NC_007606 + ND = not determined

PAGE 103

90 Primer set 01-024 was initially tested against the complete DNA library for microorganisms listed in Table 3-3, except for E. coli strains KRS308 through KRS316. Amplification was observed only fr om DNA from eight strains of S. sonnei and one strain each of S. flexneri and S. dysenteriae however DNA from two strains of S. boydii were not amplified. The observed amplifica tion of DNA from the stock strain of S. dysenteriae was not in agreement with the in silico analysis using BLASTn. The E. coli strains KRS308 through KRS316 were is olates obtained during the BAM Shigella culture trials involving ground beef. Desp ite previous testing, in wh ich primer set 01-024 tested negative against DNA from four strains of E. coli O157:H7 and three strains of nonpathogenic E. coli primer set 01-024 resulted in amp lification of DNA from eight out of nine of the ground beef E. coli isolates. These data demonstrat e that primer set 01-024 is not specific for Shigella spp. Further investigations of pr imer set 01-024 were terminated. Taken together, these data demonstrate the limitations of DNA databases and the importance of verifying primer specificity against DNA extracts from numerous strains of target and non-target microorganisms. Separation of S. sonnei from Food Matrices by Low-Speed Centrifugation Since low-speed centrifugation (LSC; 100 X g for 5 min) was to be used as a sample preparation step for potato salad a nd ground beef sample, the concentration of S. sonnei before and after LSC was investigated (Table 4-5). The S. sonnei population in potato salad was 1.1 x 104 CFU/ml in the homogenized sample before LSC and 1.9 x 104 CFU/ml in the supernatant after LSC. The S. sonnei population in ground beef was 1.1 x 105 CFU/ml in the homogenized sample before LSC and 1.6 x 105 CFU/ml in the supernatant after LSC. In both the pota to salad and ground beef samples, the S. sonnei

PAGE 104

91 Table 4-5. Effects of low-speed centrifugation on S. sonnei populations in sample supernatant. A 25 g sample of potato sala d or ground beef was transferred to a clean, sterile 50 ml microcentr ifuge tube containing 25 ml Shigella broth. The samples were then inoculated with 100 l of a five-strain S. sonnei cocktail that was resistant to the antibiotic rifa mpicin and the tube was capped tightly. The mixture was homogenized by shaki ng the tube vigorously for 30 sec. A 1.0 ml aliquot of the supernatant was seri ally diluted and ap propriate dilutions analyzed by pour-plate using tryptic soy agar (TSA) supplemented with rifampicin at 80 ppm (rif+). The remaining sample was centrifuged (100 x g for 5 min) and again a 1.0 ml aliquot of the supernatant was serially diluted and appropriate dilutions analyzed by pour-plate using TSA rif+. S. sonnei (log10 CFU/ml) Potato salad Ground beef Replicate Before After Before After 1 3.98 4.26 5.12 5.16 2 4.07 4.31 5.05 5.26 3 4.12 4.29 4.92 5.19 4 3.99 4.24 4.93 5.18 5 4.02 4.24 5.07 5.20 Avg 4.04 4.27 5.03 5.20 Stdev 0.06 0.03 0.09 0.04 populations in the supernatant followi ng LSC were significantly higher ( P < 0.05) than those in the homogenized samples prior to LS C. In addition, some solid food particulates were successfully sedimented by LSC. Take n together, these data suggest LSC, as performed in this study, may be used for the separation of some solid food particulates from potato salad and ground beef samples without sedimentation of S. sonnei from sample homogenates. Development of DNA Sequence Captur e (DSC) for the Detection of S. sonnei Two hybridization buffers (HB1 and HB2) were evaluated for use in DSC for the detection of S. sonnei in foods (Table 4-6). When a five-strain S. sonnei cocktail was analyzed by DSC followed by real-tim e PCR using HB1 (40C hybridization temperature) with CPShigella beads or beads prepared with out the addition of a specific probe (unlabeled beads), the cycle number at which the fluorescence exceeded the

PAGE 105

92 Table 4-6. Evaluation of hybridization bu ffers for DSC for the detection of S. sonnei One-ml aliquots of a five-strain S. sonnei cocktail or a five-serovar Salmonella cocktail were analyzed by the DSC method using two hybridization buffers (HB1 and HB2) and two beads preparations (CPShigella beads and unlabeled beads). In addition, HB1 was investigated with hybridization temperatures of 40C and 55C, while HB2 was only investigated at 40C. The recovered DNA was analyzed by real-time PCR. Each reported Ct value is the average of three replicates. Ct value HB1 (40 C) HB1 (55C) HB2 Test Avg Stdev Avg Stdev Avg Stdev S. sonnei CPShigella beads 20.4 0.5 19.3 0.6 18.7 0.8 Unlabeled beads 25.8 3.2 25.1 0.3 18.2 0.5 Salmonella spp. CPShigella beads 28.4 1.0 25.9 1.3 21.1 1.5 Unlabeled beads 26.8 2.1 28.3 0.8 21.7 1.0 threshold value (Ct value) was 20.4 and 25.8, respectively. The difference in Ct values (5.4 cycles) represented an approx. 100-fold increase in ipaH DNA capture when the CPShigella beads were used versus the unlab eled beads. When a five-serovar Salmonella cocktail was analyzed by the DSC method using HB1 with CPShigella beads or unlabeled beads, Ct values of 28.4 and 26.8 were observed, respectively. These data suggest that when HB1 is used with the DSC method, S. sonnei and Salmonella DNA bind in a non-specific manner to the unlabeled beads; however the CPShigella beads permit specific capture of ipaH DNA. A higher hybridization temperature (55C) was also investigated for DSC using HB1. When a five-strain S. sonnei cocktail was analyzed by the DSC method using HB1 (55C) with CPShigella beads or unlabeled beads, Ct values of 19.3 and 25.1 were observed, respectively. The difference in Ct va lues (5.8 cycles) represented an approx. 100-fold increase in ipaH DNA capture when the CPShigella beads were used versus the

PAGE 106

93 unlabeled beads. When a five-serovar Salmonella cocktail was analyzed by the DSC method using HB1 (55C) with CPShigella beads or unlabeled beads, Ct values of 25.9 and 28.3 were observed, respectively. These da ta suggest that the higher temperature for hybridization did not preven t non-specific adsorption of Salmonella DNA to CPShigella beads. When a five-strain Shigella cocktail was analyzed by the DSC method using HB2 with CPShigella beads or unlabeled beads, Ct values of 18.7 and 18.2 were observed, respectively. When a five-serovar Salmonella cocktail was analyzed by the DSC method using HB2 with CPShigella beads or unlabeled beads, Ct values of 21.1 and 21.7 were observed, respectively. These data suggest that when HB2 is used with the DSC method only non-specific capture of S. sonnei and Salmonella DNA occurs. HB2 was not investigated further in this study. Streptavidin-coated paramagnetic beads fr om two manufacturers were investigated for DSC of S. sonnei DNA (Table 4-7). A five-strain S. sonnei cocktail was analyzed with CPShigella beads prepared with ei ther Dynabeads M-280 Stre ptavidin or MagaBeads Streptavidin. As with prior experiments, unlabeled beads of each type were tested as controls. When Dynabeads M-280 Streptav idin were used for DSC, the CPShigella beads and unlabeled beads resulted in Ct values of 20.5 and 25.7, respectively. When MagaBeads Streptavidin were used for DSC, CPShigella beads and unlabeled beads resulted in Ct values of 21.5 and 24.0, resp ectively. Based on differences between Ct values of CPShigella and unlabeled beads, CPShigella beads prepared with Dynabeads M-280 Streptavidin performed slightly better for the specific capture of ipaH DNA.

PAGE 107

94 Table 4-7. Comparison of paramagnetic beads for use with CPShigella beads. One-ml aliquots of a five-strain S. sonnei cocktail was analyzed by the DSC method using CPShigella and unlabeled beads prepared with streptavidin-coated paramagnetic beads from two manu facturers. The recovered DNA was analyzed by real-time PCR. Each report ed Ct value is the average of three replicates. Ct value MagaBeads Dynabeads Beads Avg Stdev AvgStdev CPShigella beads 21.5 0.8 20.50.3 Un-labeled beads 24.0 1.4 25.71.2 The specific capture of the ipaH DNA by CPShigella beads in the presence of non-target DNA was also investigated (Table 4-8). A 1.0-ml aliquot from a five-strain S. sonnei cocktail (2.3 x 108 CFU/ml) and a 1.0-ml aliquot from a five-serovar Salmonella cocktail (5.5 x 108 CFU/ml) were mixed in a sterile te st tube containi ng 8.0 ml PBS and the resulting solution was analyzed using the DSC method using CPShigella beads or unlabeled beads (Table 4-8). Captured ipaH DNA analyzed by real-time PCR resulted in amplification at a Ct value of 23.6 when CPShigella beads were used and at a Ct value of 29.8 when unlabeled beads were used. This difference in Ct values (6.2 cycles) represented an approx. 100-fold in crease in DNA capture when the CPShigella beads were used versus the unlabeled beads. Captured Salmonella DNA analyzed by real-time PCR resulted in amplification at a Ct value of 30.7 when CPShigella beads were used and at a Ct value of 33.3 when unlabeled b eads were used. This difference (2.6 cycles) represented a less than 10-fold in crease in capture when the CPShigella beads were used versus the unlabeled beads. Based on differe nces of Ct values (7.1 cycles), the CPShigella beads captured ipaH DNA at an approximate 100-fold increase over Salmonella DNA. These data suggest that the CPShigella beads bind ipaH DNA in a specific

PAGE 108

95 Table 4-8. Specific capture of ipaH DNA in the presence of non-target DNA by CPShigella beads. One-ml aliquots of a five-strain S. sonnei and five-serovar Salmonella cocktail were analyzed by the DSC method using CPShigella and unlabeled beads. The recovered DNA was analyzed by real-time PCR. Each reported Ct value is the average of three replicates. Ct value Test Avg Stdev S. sonnei CPShigella beads 23.9 0.3 Un-labeled beads 30.9 1.2 Salmonella spp. CPShigella beads 30.7 0.6 Un-labeled beads 33.3 0.5 manner, however non-specific binding of DN A occurs with reduced efficiency. In order to determine the sensitivity of the DSC assay, serial dilutions of a fivestrain S. sonnei cocktail were prepared in PBS and analyzed using the DSC method (Table 4-9). After analysis of captured DNA by real-time PCR, the sensitivity was determined to be 2.3 x 102 CFU/ml. No amplification was observed from samples diluted to 2.3 x 101 or 2.3 x 100 CFU/ml. These data suggest that S. sonnei populations at levels as low as 102 CFU/ml may be detected by the DSC method. Expression of ipaH RNA in Log and Stationary Phase S. sonnei Using a five-strain S. sonnei cocktail in 1.0 ml SB microcosms supplemented with HeLa cell extracts, the dye Congo red or neither the extract nor the dye, ipaH RNA expression was investigated in stationary and ex ponential phase cells (Table 4-10). When RNA was extracted from stationary ph ase cells and analyzed by RT-PCR, no ipaH RNA was detected whether the inducers were pr esent or not. In contrast, when RNA was extracted from exponential phase cells and analyzed by RT-PCR, ipaH RNA was detected in all of the samples except when HeLa cell extract 3 was used. Visible growth

PAGE 109

96 Table 4-9. Sensitivity of DNA sequence capture method. One-ml aliquots of a five-strain S. sonnei were serially diluted in PBS and analyzed by the DSC method using CPShigella beads. The recovered DNA was analyzed by real-time PCR. Each reported Ct value is the aver age of three replicates. Ct value Cell titer (CFU/ml) Avg Stdev 2.3 x 106 26.2 0.7 2.3 x 105 29.8 0.6 2.3 x 104 32.4 0.3 2.3 x 103 37.6 0.8 2.3 x 102 40.6 1.4 2.3 x 101 n/a n/a 2.3 x 100 n/a n/a was achieved after 4 hr incubati on in all of the microcosms except those with HeLa cell extract 3, which indicated growth inhibiti on. Ct values from all microcosms were between 26.5 (HeLa cell extract 4) and 29.0 (HeLa cell extract 1), which did not indicate an increase in ipaH expression over the control culture in SB (Ct value of 28.2). These data suggest that for RT-PCR or other RNA am plification procedure to be effective for the detection of ipaH RNA, S. sonnei must be in exponential phase. Further, these data suggest that none of the HeLa cell extracts or the dye Congo red resulted in up-regulation of the ipaH gene to permit detection of very low populations of S. sonnei Survival Studies Survival of S. sonnei on Smooth Tomato Surfaces To evaluate the effect of drying each inoculum cocktail on the surface of the tomatoes, populations were enumerated immedi ately after inoculation and at 90 min after inoculation (time for inocula to be completely dry on all tomatoes). Compared with initial populations (5.68 log10 MPN/tomato), the level of S. sonnei (3.23 log10 MPN/tomato) was significantly reduced ( P < 0.05) after the inoculum on toma to surfaces was allowed to dry completely. These data suggest that S. sonnei is not resistant to drying on the smooth

PAGE 110

97 Table 4-10. Transcript ional induction of the ipaH gene using HeLa cell extracts and the dye Congo red. Protein extracts from HeLa cells (HeLa Extract 1-4) and Congo red were added to Shigella broth and the resulting solutions were inoculated with a five-strain S. sonnei cocktail. Total RNA was extracted from a 250 l aliquot of each test solution and analyzed by RT-PCR using ipaH gene-specific primers. The remaining portions of the test solutions were incubated at 37C for 4 hr, after which the total RNA was extracted from a second 250 l aliquot of each test so lution and analyzed by RT-PCR using ipaH gene-specific primers. The S. sonnei cocktail in SB alone was tested as a control. The Ct values given represen t the average of duplicate samples. Ct values Stationary phase Exponential phase Test solution Avg Stdev Avg Stdev HeLa Extract 1 n/aa n/a 29.0 0.8 HeLa Extract 2 n/a n/a 27.0 1.1 HeLa Extract 3 n/a n/a n/a n/a HeLa Extract 4 n/a n/a 26.5 0.1 Congo Red n/a n/a 28.0 0.4 Control n/a n/a 28.2 0.3 a no amplification surfaces of tomatoes. Significant decreases ( P < 0.05) in S. sonnei populations were observed on days 1 and 2, indicating that S. sonnei populations continued to decr ease rapidly after drying. No survivors (detection limit of th e assay was < 3 MPN/tomato) were detected on days 3 or 5 (Figure 4-6), therefore sampling was terminat ed at this point. No mold growth was observed on any tomato surface in studies involving S. sonnei during five-day observation. These data suggest that toma to surfaces do not support the survival of S. sonnei when held at the recommended temperat ure and relative humidity combination. Survival of S. sonnei in Potato Salad For studies involving S. sonnei survival in potato sa lad, there was no significant difference ( P > 0.05) between initial populations inocul ated in samples to be stored at

PAGE 111

98 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 012345 Time (day)Survivors log10 MPN/tomat o Figure 4-6. Survival of a five-strain S. sonnei cocktail on the smooth surfaces of tomatoes. Each individual S. sonnei strain was resistance to the bactericidal agent rifampicin by spontaneous adaptati on. Inoculated tomatoes were stored at 13C and 85% relative humidity. Inoc ula were recovered by placing each tomato into a sterile stomacher ba g containing 100 ml phosphate buffered saline and shaking vigorously for 30 sec followed by a 1 min hand massage. Five replicate tomatoes were sample d at each time point. Survivors were enumerated using a three-tube most probable number method in tryptic soy broth supplemented with 100 g/ml rifampicin. Error bars represent one standard deviation. 2.5 C or 8.0 C (6.03 log10 MPN/g and 5.94 log10 MPN/g, respectively) (Figure 4-7). The potato salad had an initial pH of 5.2, how ever after 28 days storage at both 2.5 C and 8.0 C, a pH of 4.5 was observed. When samples were stored at 2.5 C, S. sonnei populations were significantly lower ( P < 0.05) than the initial population on days 5, 14, and 28 (5.49 log10 MPN/g, 5.52 log10 MPN/g, and 5.54 log10 MPN/g, respectively). When samples were stored at 8.0 C, S. sonnei populations were significantly lower ( P < 0.05) than the initial population on days 7, 14, and 21 (5.51 log10 MPN/g, 4.82 log10 MPN/g, and 5.04 log10 MPN/g, respectively). S. sonnei populations observed on days 3, 5, 7, 14, and 21 were significantly lower ( P < 0.05) in samples stored at 8.0 C than in samples held at 2.5 C (Figure 4-7). At day 28, however, there was no significant difference ( P > 0.05) in S. sonnei populations in samples stored at 2.5 C or 8.0 C (5.54 log10 MPN/g and

PAGE 112

99 1.0 2.0 3.0 4.0 5.0 6.0 7.0 051015202530 Time (day)Survivors (log10 MPN/g ) 2.5C 8.0C Figure 4-7. Survival of five-strain S. sonnei cocktail in potato sa lad. Each individual S. sonnei strain was resistance to 100 g /ml rifampicin by spontaneous adaptation. Inoculated potato salad wa s stored at either 2.5C or 8.0C. Survivors were recovered by transferri ng 25 g of potato salad into a sterile stomacher bag containing 100 ml phos phate buffered saline and stomaching for 30 sec. Five replicate samples were analyzed at each time point. Survivors were enumerated using a three-tube most probable number method in tryptic soy broth supplemented with 100 g/ml rifampicin. Error bars represent one standard deviation. 5.67 log10 MPN/g, respectively). The increase in surviving S. sonnei observed on day 28 in samples stored at 8.0 C may be due to variation in potato salad ingredients among samples (i.e., onions, celery, egg, etc.). These data demonstrate the ab ility of potato salad to support long-term survival of S. sonnei at refrigerated temperatures. Survival of S. sonnei in Ground Beef For studies involving the survival of S. sonnei in ground beef, there was no significant difference ( P > 0.05) between initial populations inoculated in samples to be stored at 2.5 C or 8.0 C (6.06 log10 MPN/g and 6.06 log10 MPN/g, respectively) (Figure 4-8). The ground beef had an in itial pH of 5.6, howe ver after nine days storage at 2.5 C or 8.0 C, a pH of 5.8 and 6.1, respectively, was ob served. When samples were stored at

PAGE 113

100 1.0 2.0 3.0 4.0 5.0 6.0 7.0 012345678910 Time (day)Survivors (log10 MPN/g ) 2.5C 8.0C Figure 4-8. Survival of five-strain S. sonnei cocktail in ground b eef. Each individual S. sonnei strain was resistance to 100 g /ml rifampicin by spontaneous adaptation. Inoculated ground beef was stored at either 2.5C or 8.0C. Survivors were recovered by transferri ng 25 g of ground beef into a sterile stomacher bag containing 100 ml phos phate buffered saline and stomaching for 30 sec. Five replicate samples were analyzed at each time point. Survivors were enumerated using a three-tube most probable number method in tryptic soy broth supplemented with 100 g/ml rifampicin. Error bars represent one standard deviation. 2.5 C, S. sonnei populations were significantly lower ( P < 0.05) than the initial population on day 1 (5.68 log10 MPN/g). When samples were stored at 8.0 C, S. sonnei populations were significantly lower ( P < 0.05) than the initial population on days 1, 7, and 9 (5.65 log10 MPN/g, 5.66 log10 MPN/g, and 5.49 log10 MPN/g, respectively). S. sonnei populations observed on days 3 a nd 9 were significantly lower ( P < 0.05) in samples held at 8.0 C than in samples held at 2.5 C (Figure 4-8). Sampling was terminated on day 9 due to product spoilage determined by color and odor. These data demonstrate that ground beef supports survival of S. sonnei at refrigerated temperatures beyond the shelf-life of the product.

PAGE 114

101 Evaluation of Detection Methods Detection of S. sonnei in Selected Foods by a Modified FDA Bacteriological Analytical Manual (BAM) Shigella Culture Method A modified BAM Shigella culture method was investig ated for the detection of S. sonnei on tomato surfaces and in potato salad and ground beef samples (Table 4-11). For tomato samples inoculated at the 104 CFU/tomato, 103 CFU/tomato, 102 CFU/tomato, 101 CFU/tomato and 100 CFU/tomato levels, S. sonnei was detected in 9, 8, 6, 1 and 0 out of 10 samples, respectively. Typical colonies not identified as S. sonnei were identified as Klebsiella pnemoniae or Enterobacter spp. Atypical colonies tested for biochemical reactions were identified as Enterobacter spp. or Citrobacter freundii When potato salad samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by the BAM, S. sonnei was detected in 10, 10 and 9 out of 10 samples, respectively. All of the typi cal colonies selected for confirmation were identified as S. sonnei and there were no MAC plates fr om potato salad samples that contained atypical colonies. When ground beef samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by the BAM, S. sonnei was detected in 7, 2 and 2 out of 10 samples, respectively. The MAC plates fr om ground beef samples contained extensive growth of atypical colonies. Atypical coloni es tested for biochemical reactions were identified as Enterobacter cloacae Escherichia coli or Hafnia alvei Typical colonies not identified as S. sonnei were identified as E. cloacae or E. coli Taken together, these data suggest that more selective enrich ment and isolation media are needed for the conventiona l culture analysis of foods for S. sonnei

PAGE 115

102 Table 4-11. Number of samples positive for S. sonnei by various detection methods. Smooth tomato surfaces, potato salad, and ground beef were analyzed by the FDA, Bacteriological Analytical Manual (BAM) Shigella culture method, flow-through immunocapture (FTI) fo llowed by direct plating on MacConkey (MAC) agar (FTI-MAC), FTI followed by real-time PCR (FTI-PCR) and DNA sequence capture (DSC). Detection method Food sample Inoculation level BAM FTI-MAC FTI-PCR DSC Tomato 104 CFU/tomato 9 10 10 10 103 CFU/tomato 8 8 10 9 102 CFU/tomato 6 10 10 10 101 CFU/tomato 1 7 9 9 100 CFU/tomato 0 0 0 0 Potato salad 102 CFU/25 g 10 10 10 10 101 CFU/25 g 10 10 10 10 100 CFU/25 g 9 8 8 10 Ground beef 102 CFU/25 g 7 4 10 10 101 CFU/25 g 2 4 9 10 100 CFU/25 g 2 0 2 6 Detection of S. sonnei in Selected Foods by Flow -Through Immunocapture (FTI) FTI-MAC was investigated for the detection of S. sonnei on tomato surfaces and in potato salad and ground beef samples (Table 4-11 ). For tomato samples inoculated at the 104 CFU/tomato, 103 CFU/tomato, 102 CFU/tomato, 101 CFU/tomato and 100 CFU/tomato levels, S. sonnei was detected in 10, 8, 10, 7 and 0 out of 10 samples, respectively. All of the colonies selected for confirmation were identified as S. sonnei Only one MAC plate from a tomato inoculated at the 103 CFU/tomato level contained atypical colonies identified as Enterobacter spp. through biochemical characterization. BRLR models determined the FTI-MAC method achieved greater detection ( P < 0.05) of inoculated S. sonnei on tomato surfaces over the BAM Shigella culture method. These data suggest the FTI-MAC method may be used for the analysis of tomato surfaces for the presence of S. sonnei

PAGE 116

103 When potato salad samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by FTI-MAC, S. sonnei was detected in 10, 10 and 8 out of 10 samples, respectively. All of the coloni es selected for confirmation were identified as S. sonnei and there were no MAC plates contai ning atypical colonies. BRLR models revealed there was no significant difference ( P > 0.05) between the FTI-MAC method and the BAM Shigella culture method. These data suggest the FTI-MAC method may be used for the analysis of pot ato salad for the presence of S. sonnei When ground beef samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by FTI-MAC, S. sonnei was detected in 4, 4 and 0 out of 10 samples, respectively. The MAC plates fr om ground beef samples contained extensive growth of atypical colonies. Atypical coloni es tested for biochemical reactions were identified as E. cloacae E. coli or H. alvei Typical colonies not identified as S. sonnei were identified as C. freundii E. cloacae or E. coli BRLR models revealed there was no significant difference ( P > 0.05) between the FTI-MAC method and the BAM Shigella culture method for the detection of S. sonnei in ground beef. These data suggest the FTIMAC method may be used for the analysis of ground beef for the presence of S. sonnei however more selective enrich ment/isolation media and antiS. sonnei antibodies are required to effectively eliminate co-isolation of closely-related Enterobacteriaceae such as E. coli FTI-PCR was investigated for the detection of S. sonnei on tomato surfaces and in potato salad and ground beef samples (Table 4-11 ). For tomato samples inoculated at the 104 CFU/tomato, 103 CFU/tomato, 102 CFU/tomato, 101 CFU/tomato and 100 CFU/tomato levels, S. sonnei was detected in 10, 10, 10, 9 and 0 out of 10 samples,

PAGE 117

104 respectively. Bias-reduced l ogistic regression (BRLR) m odels determined the FTI-PCR method achieved greater detection ( P < 0.05) of inoculated S. sonnei on tomatoes over the BAM Shigella culture method. All FTI-PCR analyses were completed within 24 hr. These data suggest the FTI-PCR method may be used for the rapid detection of S. sonnei on tomato surfaces. When potato salad samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by FTI-PCR, S. sonnei was detected in 10, 10 and 8 out of 10 samples, respectively. BRLR models re vealed there was no significant difference ( P > 0.05) between the FTI-PCR method and the BAM Shigella culture method. All FTIPCR analyses were completed within 24 hr. These data suggest the FTI-PCR method may be used for the rapid detection of S. sonnei in potato salad. When ground beef samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by FTI-PCR, S. sonnei was detected in 10, 9 and 2 out of 10 samples, respectively. BRLR models determined the FTI-PCR method achieved greater detection ( P < 0.05) of inoculated S. sonnei in ground beef samples over the BAM Shigella culture method. All FTI-PCR analyses were completed within 24 hr. These data suggest the FTI-PCR method may be us ed for the rapid detection of S. sonnei in ground beef. Detection of S. sonnei in Selected Foods by DNA Sequence Capture (DSC) DSC was investigated for the detection of S. sonnei on tomato surfaces and in potato salad and ground beef samples (Table 4-11 ). For tomato samples inoculated at the 104 CFU/tomato, 103 CFU/tomato, 102 CFU/tomato, 101 CFU/tomato and 100 CFU/tomato levels, S. sonnei was detected in 10, 10, 10, 9 and 0 out of 10 samples, respectively. BRLR models determined the DSC method achieved greater detection ( P <

PAGE 118

105 0.05) of inoculated S. sonnei on tomatoes over the BAM Shigella culture method. All DSC analyses were completed within 24 hr. These data suggest the DSC method may be used for the rapid detection of S. sonnei on tomato surfaces. When potato salad samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by DSC, S. sonnei was detected in 10, 10 and 10 out of 10 samples, respectively. BRLR models rev ealed there was no significant difference ( P > 0.05) between the DSC method and the BAM Shigella culture method. All DSC analyses were completed within 24 hr. These data s uggest the DSC method may be used for the rapid detection of S. sonnei in potato salad. When ground beef samples inoculated at the 102 CFU/25 g, 101 CFU/25 g and 100 CFU/25 g levels were analyzed by DSC, S. sonnei was detected in 10, 10 and 6 out of 10 samples, respectively. BRLR models dete rmined the DSC method achieved greater detection ( P < 0.05) of inoculated S. sonnei in ground beef samples over the BAM Shigella culture method. All DSC analyses were completed within 24 hr. These data suggest the DSC method may be us ed for the rapid detection of S. sonnei in ground beef. BRLR models were also used to compare the DSC method to the FTI-PCR methods for the detection of S. sonnei on tomato surfaces and in potato salad and ground beef samples. There was no significant difference ( P > 0.05) between the DSC method and the FTI-PCR method for the analysis of to mato or potato salad samples; however the DSC method achieved greater detection ( P < 0.05) of inoculated S. sonnei in ground beef samples over the FTI-PCR method. These data suggest that the DS C method may provide superior detection of S. sonnei over the FTI-PCR method when closely related Enterobacteriaceae such as E. coli are present in food.

PAGE 119

106 CHAPTER 5 DISCUSSION AND CONCLUSIONS Conventional culture methods for the detection of Shigella in foods are problematic, in that appropriate selectiv e media are not currently available and Shigella spp. are often overgrown by competitive microorganisms during enrichment. This study investigated two alternative sample prepara tion methods for the detection of inoculated Shigella sonnei from tomato surfaces and in pota to salad and ground beef. Flow-through immunocapture (FTI) followed by analysis of recovered beads by spread-plate using MacConkey agar (MAC) and by real-time PCR and DNA sequence capture (DSC) were developed. Food samples were inoculated at decreasing levels to determine the lowest detection level for each assay. The FDA Bacteriological Analytical Manual (BAM) Shigella culture method was performed for compar ison to the newly developed methods. Preliminary Studies Five S. sonnei strains were investigat ed in this study to account for variation among strains. All S. sonnei strains were purchased from the American Type Culture Collection (ATCC 9290, ATCC 29031, ATCC 29030, ATCC 25931 and ATCC 29930). S. sonnei represents the serogroup of Shigella most commonly associated with foodborne outbreaks in North America. According to the Center s for Disease Control and Prevention’s (CDC) Public Health Laboratory Information System (PHLIS), S. sonnei accounts for >80% of the Shigella isolates in the U.S. in recent years. The most current PHLIS data, however reports that S. sonnei isolates in 2004 accounted for only 68.9% of the Shigella isolates in

PAGE 120

107 the U.S. It remains to be determined if this decreased incidence of S. sonnei will continue in future years. Three distinctly different food matrices we re chosen for analysis by the developed methods: tomatoes, potato salad and ground beef. These foods were chosen due to their association with previous sh igellosis outbreaks and/or th eir chemical/microbiological composition. The tomato can be used as a re presentative for produce and is the only food matrix chosen in which the inoculum is dried completely prior to analysis. In addition, previous research on detection methods for Shigella spp. on tomato surfaces provides a means for direct method comparison (Warren, 2003; Warren et al ., 2005b). Potato salad is representative of the food category ‘prepa red salads’ that have been implicated in previous shigellosis outbreaks (Lew et al ., 1991; TPH, 2002). Potato salad, composed of carbohydrates (potatoes), proteins (eggs) and lipids (mayonnaise ), contains low levels of microorganisms if prepared under sanitary c onditions and not subjected to temperature abuse. Ground beef, while high in protein and fat, typically contains higher microbial loads than potato salad and is often contaminated with E. coli a microorganism closelyrelated to Shigella Growth Characteristics of S. sonnei in Shigella Broth (SB) Growth characteristics were dete rmined for each of the five S. sonnei strains by performing growth curves in SB incubated at 44C without shaking. All five strains experienced an initial lag phase of 2 hr prio r to entering the exponen tial phase of growth. The average doubling time of the five S. sonnei strains during exponential growth was 18.8 min. The growth data was used to estimate cell populations after S. sonnei on inoculated tomatoes, potato salad and ground beef samples were recovered in SB and incubated for short (4-5 hr) periods of tim e. From these estimates, it was determined

PAGE 121

108 whether S. sonnei populations could meet or exceed the detection limits of the newly developed assays in order to facilitate sa me-day (within 8 hr) de tection. Such growth estimates make the assumptions that all of the S. sonnei inoculated to the food matrices would be recovered into the SB enrichment and that growth of S. sonnei in the presence of food components and indigenous microflora would be similar to that in pure culture. In practice, it is more likely that only a portion of the inoculated S. sonnei would be recovered in the analysis of food samples and that these S. sonnei would not achieve optimal growth rates in the presence of competitive microorganisms. The best-case estimates of S. sonnei populations in SB enrichments after short incubation times concluded that cell titers requ ired for reliable detection by FTI or DSC in same-day formats could not be achieved with low initial inoculati on levels. Instead of determining the minimum enrichment time n ecessary for reliable detection, an 18-hr enrichment was used to simulate the wa y the developed methods would be used by commercial diagnostic laboratorie s. In practice, food samples to be analyzed would be received in the morning or early afternoon a nd set-up in the late afternoon. The following morning, the detection assay would be performed on the enriched samples. Expression and Induction of the ipaH Gene of S. sonnei In preliminary experiments, ipaH RNA was extracted from stationary and exponential phase S. sonnei in SB and analyzed by RT-P CR. The addition of HeLa cell extracts and Congo red dye was test ed for their ability to induce ipaH expression. Results demonstrated no ipaH RNA was detected in stat ionary phase cells, however ipaH RNA was detected in exponential phase cells. The addition of HeLa cell extracts and the dye Congo red was found not to induce ipaH expression above that observed in control cultures, with the exception of HeLa cell extract 3, which resulted in no ipaH RNA

PAGE 122

109 expression. The SB supplemented with HeLa cell extract 3 showed no visible growth after 4 hr incubation, indi cating the extract inhib ited the growth of the S. sonnei cocktail. Initial S. sonnei populations were approx. 2.0 x 106 CFU/ml and based on growth curve data presented in this study, S. sonnei populations after 4 hr incubation were >107 CFU/ml. From these results it wa s determined that detection of S. sonnei by RT-PCR or other RNA amplification methods, such as nucleic acid sequence-based amplification (NASBA) would require a gene tic target other than the ipaH gene unless significant preenrichment was performed prior to RNA extraction. Identification of Potentially Shigella -Specific Genetic Loci Attempts were made to identify a Shigella -specific chromosomally-located gene whose RNA might be useful fo r detection by RT-PCR or NASBA. In total, nine genetic targets were identified for further anal ysis using the MBGD and BLASTn software (Table 4-3). Of these, the ybgD gene was the only identified target for which primers were developed with specificity for S. sonnei S. flexneri and S. dysenteriae when tested against the DNA library available at the time (all of those listed in Table 3-3 except KRS308 through KRS316). The ybgD gene has been described as a putative fimbrial-like protein, therefore the expre ssion of YbgD is questionabl e given that none of the sequenced Shigella genomes contain intact loci for fimbrial biogenesis (Yang et al ., 2005). The ybgD gene was found to be expressed in actively growing cells (in SB) when tested by RT-PCR (data not shown). Un fortunately, the addition of nine E. coli strains isolated from ground beef dur ing evaluation of the BAM Shigella culture method to the DNA library resulted in eight of the nine testing positive for the ybgD sequence. Since the ybgD sequence was no longer specifi c to species within the Shigella all further testing of the ybgD for use in detection assays was terminated.

PAGE 123

110 Using the BLASTn software for testing the specificity of a given DNA sequence is limited by the number of DNA sequences available in the database. Results from this study are reflective of this; many of the mol ecular targets when an alyzed using BLASTn returned significant hits only with the sequences of Shigella spp., however when tested against DNA extracted from the microorganism s in the DNA library these targets were not specific. Furthermore, the BLASTn resu lts often did not matc h the in house testing for species of Shigella For example, the primers developed for the ybgD gene (primer set 01-024) were not homologous to sequences within the S. dysenteriae or S. boydii genomes when tested using BLASTn, however primer set 01-024 amplified DNA from S. dysenteriae ATCC 9361 when tested in house. This is significant as S. dysenteriae ATCC 9361 and the S. dysenteriae whose genome has been sequenced are both serotype 1 strains. E. coli DNA sequences recently made publicly available (April 2006) would have eliminated most of the identified gene tic targets as potentially specific for Shigella At onset of this project, the only Shigella genomes available were that of S. flexneri 2a strain 301 and S. flexneri 2a strain 2457T. Over the course of this work, the genomes of S. boydii serotype 4 (strain 227), S. dysenteriae serotype 1 (strain 197) and S. sonnei (strain 046) have been made pu blicly available (Yang et al ., 2005). As genome sequences from additional serotypes/strains of Shigella and E. coli become available, a more complete and representative in silico analysis of gene distribution among the E. coli / Shigella species can be performed. Survival Studies In previous studies where the survival of Shigella was investigated on smooth tomato surfaces, samples were inoc ulated with rifampicin-resistant Shigella and rinsed

PAGE 124

111 with 100 ml phosphate buffered saline (PBS) a nd subjected to the same shake/rub method as employed in the present study (Schneid er and Warren, unpublished data). A 1.0 ml aliquot of the PBS rinse was then analyzed by pour-plate using TSA rif+ to determine the number of survivors. The detection li mit for this type of method is 1.0 x 102 CFU/ml tomato rinse, therefore as Shigella populations declin ed below this level it could not be determined if a small, but viable, populati on had survived. To overcome this problem, the MPN method described in the present study was developed and used to investigate S. sonnei survival on tomatoes, in potato salad a nd in ground beef. The de tection limit of the MPN method was > 3 MPN/ml tomato rinse. Rapid S. sonnei Inactivation on Tomato Surfaces The waxy surface of the tomato is very smooth, unlike the corky surfaces of cantaloupes or potatoes, and without invaginations (as f ound on oranges) or epidermal and peridermal pores, such as stoma or le nticels, respectively. Survival experiments revealed that S. sonnei inactivate rapidly when dried on the waxy surface of a tomato. The inocula took approx. 90 min to dry comple tely on the tomato surface, during which a 2.45 log10 CFU/tomato decrease in S. sonnei was observed. After drying, S. sonnei continued to die rapidly and no survivors we re observed after three days of storage (Figure 4-6). These results were consistent with Islam et al (2001), who reported that no S. dysenteriae serotype 1 inoculated on cloth, wood, plastic, aluminum, and glass objects could be recovered after five days by conven tional culture methods. In contrast, Spicer (1959) reported S. sonnei survival for up to 12 days on cotton threads held at 5-10C and Nakamura (1962) reported S. sonnei survival for up to 14 days on cotton, glass, wood, paper, and metal at various temperatur es. Strain variati on, holding conditions

PAGE 125

112 (temperature/relative humidity) and surface pH may partially explain the differences in survival observed among these reports. S. sonnei Survives in Potato Salad and Ground Beef Survival experiments revealed that the potat o salad (as prepared in this study) and ground beef support survival of S. sonnei over the shelf-life of these products. These observations were in agreem ent with other reports of Shigella spp. survival in similar foods. Rafii and Lunsford (1997) observed S. flexneri survival in carrot salad, potato salad, coleslaw and cr ab salad held at 4 C. While S. flexneri populations d eclined in the carrot and potato salads, significant populations remained at day 11. The pH of the carrot and potato salads used by Rafii and Lunsfo rd (1997) were pH 2.7 to 2.9 and pH 3.3 to 4.4, respectively, whereas the potato salad us ed in the present study was pH 5.2 to 5.9. Rafii and Lunsford (1997) reported S. flexneri survival at higher levels in the coleslaw and crab salads, pH 4.1 to 4.2 and pH 4.4 to 4.5, respectively. No previous reports were found documenting the survival of Shigella in raw ground beef, however Islam et al (1993b) reported that cooked b eef supported the growth of S. flexneri under temperature abuse conditions. Evaluation of Detection Methods Recovery of S. sonnei by the BAM Shigella Culture Method The BAM Shigella culture method was used to analyze tomatoes, potato salad and ground beef inoculated with d ecreasing concentrations of S. sonnei Closely related members of the Family Enterobacteriaceae such as E. coli Enterobacter spp., Citrobacter spp. and Klebsiella spp., were observed to grow in SB enriched aerobically at 44 C and resulted in non-specific colonies when SB enrichments were streaked to MAC. This is in agreement with previous studies in which the enrichment media of the BAM

PAGE 126

113 Shigella culture method were investigated for the isolation of S. sonnei from foods (Uyttendaele et al ., 2000; Warren et al ., 2005a). Excessive presence of background microorganisms on MAC plates from to mato and ground beef samples made identification of S. sonnei colonies difficult and contribute d to decreased detection. This highlights the need for more sel ective media for the isolation of S. sonnei from food. Analysis of lowest detection levels of the BAM Shigella culture method The lowest detection level (LDL) and the lowest detection level in which S. sonnei was detected in all 10 of the replicates (LDL100) were determined for the BAM Shigella culture method (Table 5-1). The LDL and LDL100 when inoculated tomatoes were analyzed by the BAM Shigella culture method were 4.9 x 101 CFU/tomato and >4.9 x 104 CFU/tomato, respectively. Th e lack of detection of S. sonnei from all 10 replicate tomatoes at any of the inoculat ion levels was likely influenced by the effect of drying the inoculum on the tomato surface and/or the pr esence of competitive microorganisms. In support of this possibility, Flessa et al. (2003) reported the loss of 1.0 log10 CFU when S. sonnei populations in 0.1% peptone were inoculated on glass cover slips and allowed to dry completely (approx. 40 min). The LDL and LDL100 when inoculated potato salad was analyzed by the BAM Shigella culture method were 2.6 x 100 CFU/25 g and 2.6 x 101 CFU/25 g, respectively. When potato salad sample s were analyzed for total aerobic plate count using TSA (data not shown), it was determined that very low levels (<103 CFU/g) of background microorganisms were present prio r to inoculation. It was considered that this contributed to the increased recovery of S. sonnei from potato salad by the BAM Shigella culture method over that observed with ground beef samples. The LDL and LDL100 when inoculated ground beef samples were analyzed by the BAM Shigella culture method were 1.1 x 100 CFU/25 g and >1.1 x 102 CFU/25 g, respectively. The

PAGE 127

114 Table 5-1. Lowest detection levels of the BAM Shigella culture method. Tomatoes, potato salad and ground beef samples we re inoculated with a five-strain S. sonnei cocktail at various levels and analyzed using BAM Shigella culture method The lowest detection level (LDL) and lowest detection level in which S. sonnei was detected in all 10 of the replicates (LDL100) were determined based on initial inoculation levels. Food matrix LDL LDL100 Tomato 4.9 x 101 >4.9 x 104 (CFU/tomato) Potato salad 2.6 x 100 2.6 x 101 (CFU/25 g) Ground beef 1.1 x 100 >1.1 x 102 (CFU/25 g) presence of high competitive background microo rganisms likely contributed to the lack of detection of S. sonnei in all 10 replicates at any of the inoculation levels tested in this study. In order to determine the true LDL100 when the BAM Shigella culture method is used to analyze tomatoes or ground beef, this experiment would have to be repeated using higher inoculation levels than were investigated in this study. In a similar study, Jacobson et al (2002) evaluated the BAM Shigella culture method using two strains of S. sonnei (strains 357 and 20143) on selected types of produce. LDLs were determined using unstres sed, chill-stressed, a nd/or freeze-stressed cells. LDLs with unstressed cel ls were less than 1.0 x 101 CFU/25 g for all produce types, while LDLs with chill-stressed and freezestressed cells were less than 5.2 x 101 CFU/25 g for all produce types tested (Jacobson et al. 2002). In another similar study, Warren (2003) reported the LDL of inoculated S. sonnei on tomato surfaces when analyzed by the BAM Shigella culture method to be 1.9 x 102 CFU/tomato. Variation among strains of S. sonnei could explain the difference in reported LDLs with the BAM Shigella culture method. Taken together, these a nd the current study demonstrat e the importance of using more than one strain when evaluating detection methods for foodborne pathogens.

PAGE 128

115 Evaluation of Flow-Through Immunocapture for the Detection of S. sonnei in Food Initial attempts were to develop the FTI methods into a same-day format, as was previously reported for Salmonella (Yuk et al ., 2006). For same-day enrichments, previous studies used a combination of enrichment media supplemented with yeast extract and shaking incubation to decrease lag phase and increase the growth rate of Salmonella prior to FTI in order to detect very low (100 CFU/25 g) inoculation levels (Schneider and Warren, unpublished data ). The minimum concentration of S. sonnei ATCC 29930 required for dete ction by FTI-MAC (2.8 x 105 CFU/ml), however was too high to achieve using the shor t enrichment times availabl e for same-day testing. In contrast, S. sonnei ATCC 25931 was only required at 3.1 x 101 CFU/ml for detection by FTI-MAC, therefore short enrichment times ma y have generated a sufficient cell titer for a same-day test. This demonstrates the importa nce of using several different strains when evaluating detection methods for foodborne bacteria. Operational issues with antiShigella beads in flow-through immunocapture Un-coated paramagnetic beads are not av ailable from Matrix MicroScience, therefore other commercially available beads we re investigated for use with the Pathatrix system. The beads from Matrix MicroScien ce, although of proprietary composition, are 0.8 microns in diameter with high (>50 %) iron content. The Dynabeads M-270 Epoxy (Invitrogen), 2.8 microns in diam eter with ~15% iron content, were investigated for use with the Pathatrix system with poor results. When un-labeled beads were added to the Pathatrix system and circulated (30 min at 37 C) with 250 ml PBS, no visible recovery of beads in the capture phase was observed. It was considered that the in creased size and/or lower iron content as compared to the Matr ix MicroScience beads prevented the magnetic draw of the Dynabeads M-270 out of the ci rculating sample. Since both the Matrix

PAGE 129

116 MicroScience and the Dynabeads M-270 bead s are polystyrene, it was not considered that the Dynabeads were sticking to components of the tubing system and not returning to the capture phase. Lowering the flow rate may have increased recovery, however the Pathatrix has only three possible flow rates: 280 ml/min (speed 1), 400 ml/min (speed 2) or 500 ml/min (speed 3). The normal flow rate used for sample circulation is speed 2, while washing of recovered bead s is performed at speed 1. None of the currently reported Pathatrix protocols utilized speed 3. The us e of speed 1 for sample circulation would limit the number of times the sample is pass ed through the capture phase in a 30 min run time, therefore paramagnetic beads with hi gher iron content were investigated. The MagaCell beads, approx. 3.0 micr ons in diameter with 50% ir on content, were tested in the Pathatrix system using the normal circul ation flow rate. Although the MagaCell beads were similar in diameter to the Dynabead s M-270 beads, the increased iron content facilitated recovery in the Pathatrix system. Recovery of the antiShigella beads during FTI experiments was observed to be variable among replicates of the same type of food. For example, when tomato samples were analyzed on the Pathatrix, some sample s resulted in visible beads however a few samples did not. A portion of the beads may be immobilized within the capture phase without being visible to the na ked eye (Dr. J.P. Coombs, Ma trix MicroScience, personal communication). Tomato samples without vi sibly recovered beads tested positive for S. sonnei by the FTI-MAC and FTI-PCR methods, which verified the presence of beads in the capture phase without being visible to the naked eye. In some of the food samples, antiShigella beads were observed to be immobilized in an unconventional ma nner within the capture phase. Normally, the antiShigella beads

PAGE 130

117 are immobilized within the capture phase at the surface bordering th e top of the magnet (Figure 5-1, arrow 1). In some samples, how ever, some or all of the recovered antiShigella beads were immobilized within the cap ture phase at the surface bordering the side of the magnet (Figure 5-1, arrow 2). It was not determined how this unconventional placement of antiShigella beads affected immunocapture. The design of the capture phase creates a vortex in the area immediat ely beyond the magnet and the second baffle, therefore it is likely that target microorganisms may contact antiShigella beads immobilized at the side of the magnet. Detection of S. sonnei was observed in samples, which contained recovered beads only at the si de of the magnet, how ever it could not be determined if the S. sonnei cells were bound by the antiShigella beads before or after they were pushed to the side of the magnet by the sample flow. It should be noted that some immunocapture of S. sonnei may have occurred as antiShigella beads circulated through the Pathatrix system prior to immobiliz ation in the capture phase. This could be tested by immobilizing the antiShigella beads in the capture phase before adding the sample, however this was not performed in this study. Non-specific immunocapture of Enterobacter cloacae and Escherichia coli In FTI experiments involving gr ound beef, non-specific binding of E. coli and E. cloacae was observed and most likel y affected detection of S. sonnei at all of the inoculation levels. This n on-specific immunocapture was not observed when ground beef was analyzed by FTI-MAC in preliminary e xperiments. Initially, degradation of the blocking agent (ethanolamine) used during the preparation of antiShigella beads was suspected of contributing to non-specific binding since preliminary experiments involving FTI with ground beef resu lted in specific isolation of S. sonnei The blocking solution (sodium bicarbonate buffer, pH 8.6, containing 0.3% ethanolamine) had been

PAGE 131

118 Figure 5-1. Location of antiShigella beads in the capture phase of the Pathatrix during FTI. The direction of flow is indi cated by the white dashed arrow. The Pathatrix is designed to immobilize IMS beads at the location shown by arrow 1, however in FTI experiments, beads were often observed to accumulate at the location shown by arrow 2. prepared in a 50 ml centrifuge tube and held at room temperature unprotected from light. The antiShigella beads were prepared again with freshly made blocking solution and 10 replicates of 25 g ground beef sa mples were inoculated at 102 CFU/25 g and re-analyzed by FTI-MAC and FTI-PCR. S. sonnei was detected in all 10 samples by both methods; however MAC plates continue d to contain colonies of E. coli and/or E. cloacae Based on these findings, the non-specific immunocapture was most likely due to a shared or very similar epitope among S. sonnei and at least some strains of E. coli and E. cloacae Competitive inhibition may have also contributed to poor recovery of S. sonnei in ground beef by FTI-MAC. If growth of S. sonnei was inhibited during enrichment in SB, the ratio of S. sonnei to the competitive microflora after 18 hr would be low. Previous studies involving FTI of Salmonella in ground beef determined that when Salmonella populations were present in a low ratio to that of competitive background, increased non1 2

PAGE 132

119 specific capture was observed (Schneider a nd Warren, unpublished data). However, when the ratio of Salmonella to competitive background was higher, non-specific capture was decreased or absent completely (Sch neider and Warren, unpublished data). S. sonnei was detected in ground beef samp les inoculated at the 100 CFU/25 g level by FTI-PCR and DSC, which provided eviden ce that the growth of S. sonnei was not completely inhibited, since some growth would be necessary to reach the detection limit of both assays (105 CFU/ml and 102 CFU/ml, respectively). Anot her possibility is that E. coli and E. cloacae may have achieved higher cell titers than S. sonnei when enriched in SB, which would have naturally created the low target-to-non-target ratio that resulted in non-specific immunocapture in previous studies. Competitive inhibition of S. sonnei by ground beef microflora could be investigated fu rther using the rifampicin resistant S. sonnei subcultures used in the survival studies. Gr ound beef samples would be inoculated with a cocktail of the rifampicin resistant S. sonnei enriched in SB and then dilutions of the SB enrichment analyzed by pour-plate using TS A and TSA rif+. By comparing the counts from TSA (total plate count in ground beef enrichment ) with TSA rif+ (total S. sonnei count in ground beef enrichment) it c ould be determined if growth of S. sonnei was affected by the competing microorganisms. Analysis of lowest detection levels of the FTI-MAC and FTI-PCR methods The LDL and LDL100 were determined for the FTI-MAC and FTI-PCR methods (Table 5-2). For tomato samples, the LD L of the FTI-MAC and FTI-PCR methods was 4.9 x 101 CFU/tomato and the LDL100 was 4.9 x 102 CFU/tomato. Notably, S. sonnei was only detected in eight of ten re plications inoculated at 4.9 x 103 CFU/tomato (Table 411). This failure to detect S. sonnei in all replications could ha ve been due to variation of indigenous microflora among tomatoes. For pot ato salad samples, the LDL of the FTI

PAGE 133

120 Table 5-2. Lowest detection levels of th e FTI-MAC and FTI-PCR methods. Tomatoes, potato salad and ground beef samples we re inoculated with a five-strain S. sonnei cocktail at various levels and an alyzed using FTI-MAC or FTI-PCR. The lowest detection level (LDL) a nd lowest detection level in which S. sonnei was detected in all 10 of the replicates (LDL100) were determined based on initial inoculation levels. FTI-MAC FTI-PCR Food matrix LDL LDL100 LDL LDL100 Tomato 4.9 x 101 4.9 x 102 4.9 x 101 4.9 x 102 (CFU/tomato) Potato salad 1.5 x 100 1.5 x 101 1.5 x 100 1.5 x 101 (CFU/25 g) Ground beef 1.9 x 101 >1.9 x 102 1.9 x 100 1.9 x 102 (CFU/25 g) MAC and FTI-PCR methods was 2.6 x 100 CFU/25 g and the LDL100 was 2.6 x 101 CFU/25 g. Low indigenous microbial levels c ould have contributed to the low LDL and LDL100 values observed in potato salad samples. It should be noted that in practical application, potato salad-type products may have been subjected to temperature abuse and therefore may contain far higher microbial counts, which may affect the sensitivity of the FTI-MAC and/or FTI-PCR methods. For gr ound beef samples, the LDL of the FTIMAC and FTI-PCR methods were 1.5 x 101 CFU/25 g and 1.5 x 100 CFU/25 g, respectively. Since S. sonnei was not detected in all of the ground beef replic ations at any of the inoculation levels by FTI-MAC, the LDL100 was reported as >1.5 x 102 CFU/25 g. In order to determine the true LDL100 when FTI-MAC is used to analyze ground beef, this experiment would have to be repeated using higher inoculation levels than were investigated in this study. The LDL100 of the FTI-PCR method was 1.5 x 102 CFU/25 g. Higher LDL100 values for ground beef samples likely reflect the type and population of indigenous microflora as compared to that of tomatoes and potato salad.

PAGE 134

121 The LDLs of the developed FTI-PCR me thod meet or exceed many previous reports of PCR detection of Shigella spp. in food. Vantarakis et al (2000) reported the detection of S. dysenteriae serotype 1 in mussels at 1.0 x 103 CFU/ml with no preenrichment and 1.0 x 101-2 CFU/ml following 22 hr incubation in buffered peptone water using a chemical DNA extraction/ethanol precipitation method. Villalobo and Torres (1998) reported the detection of S. dysenteriae serotype 1 in mayonnaise at 1.0 x 102-3 CFU/ml following phenol-chloroform DNA extr action/ethanol precipitation. In a study by Lindqvist (1999), nested PCR in combina tion with buoyant dens ity centrifugation was able to detect S. flexneri at 1.0 x 101 CFU/g in lettuce, shrimp, milk, and blue cheese samples. Without the use of buoyant density centrifugation prior to nested PCR, the detection limit was 1.0 x 103 CFU/ml in aqueous solution. Theron et al (2001) investigated a semi-nested PCR for the detection of S. flexneri in spiked environmental water samples with detection limits of 2.0 x 103 CFU/ml for well water, 1.4 x 101 CFU/ml for lake water, 5.8 x 102 CFU/ml for river water, 6.1 x 102 CFU/ml for treated sewage water, and 1.1 x 101 CFU/ml for tap water following a six hr pre-enrichment in GN broth. Variability in results among the wate r samples was attributed to the presence of humic substances that inhibited PCR. The FTI-MAC and FTI-PCR methods did not meet the LDL of a previously reported FTA filtration-nested PCR (F TA-PCR) method for the detection of S. sonnei on tomatoes (7.4 x 100 CFU/tomato) (Warren, 2003; Warren et al ., 2005b). For a description of the FTA-PCR method, the read er is directed to Chapter 2, Literature Review. Survival studies determined that significant numbers of S. sonnei die or become sublethally injured when dried completely on the tomato su rface. Therefore it is likely that all of the

PAGE 135

122 S. sonnei inoculated at the 100 CFU/tomato level in the prev ious and present studies may have been inactivated or subl ethally injured prior to analysis. Due to the nature of FTA filters, naked DNA or sublethally injured S. sonnei could still be detected and the sensitivity of the FTA-PCR me thod did not depend on pre-enrichment of the sample. In contrast, the FTI-MAC and FTI-PCR methods required a lengthy pre-enrichment in which sublethally injured S. sonnei may not have recovered. Unlike the FTA-PCR method, however the FTI-MAC method allows fo r the isolation of viable colonies, a feature that is desirable if furthe r characterization of the detected Shigella is required. Future research involving flow-throu gh immunocapture for the detection of S. sonnei There are several areas that warrant fu ture research for FTI detection of S. sonnei in food. As more specific antiShigella antibodies become available, the non-specific binding of E. coli and E. cloacae by the antiShigella beads observed in the present study may be reduced or eliminated. Only commercia lly available antibodies were investigated in this report, however there are several reports in the literature that describe specific antiShigella antibodies (Rahman and Stimson, 2001). Unsuccessful attempts were made to obtain antibodies report ed to be specific for S. sonnei from other researchers. Another area of future research that could improve the detection of S. sonnei by FTI methods is the development of more selective media for the enrichment or isolation of Shigella spp. Previously, a chromogenic me dia for the isolation of Shigella (CSPM) was compared to MAC and Salmonella Shigella agar (SSA) for the isolation of S. sonnei and S. boydii from tomatoes with no significant differen ce noted among these isolation media (Warren, 2003; Warren et al ., 2005a). In the present study, a preliminary experiment compared CSPM and MAC for the isolation of S. sonnei from potato salad and ground beef (data not shown). The use of MAC over CSPM re sulted in greater isolation rates of S. sonnei

PAGE 136

123 from potato salad and ground beef; therefore CSPM was not further evaluated in this study. Finally, the addition of nutrients and/or a dditives to SB, such as yeast extract, may be investigated for improved r ecovery of sublethally injured S. sonnei from tomato surfaces. In a preliminary study, SB was suppl emented with 0.5% yeast extract (SBYE) in an attempt to shorten lag pha se and boost the growth rate of S. sonnei to generate enough cells to allow detection in a same-day format. It was observed that one tomato inoculated at the 100 CFU/tomato level tested positive by FTI-MAC and FTI-PCR. Further research is necessary to determine if this observati on is repeatable and if SBYE may improve the detection of S. sonnei from tomato surfaces. Evaluation of DNA Sequence Captur e (DSC) for the Detection of S. sonnei in Food Evaluation of hybridization buffers for DSC Two hybridization buffers were evaluated fo r use with DSC. Hybridization buffer 1 (HB1), a Tris-HCl based buffer, was described by Mangiapan et al (1996) for the detection of mycobacterial DNA from clinical samples. HB1 was chosen for investigation in this study since, unlike ot her reports of DSC, DNA prepar ations were crude tissue and cell lysates were prepared directly in HB1, therefore the entire sample was analyzed. Hybridization buffer 2 (HB2), a sodium-sodium citrate (SSC) based buffer was similar to that described by Tsai et al. (2003). In most reports where variations of HB2 were used, DNA was extracted from samples and an ali quot of the extracted DNA was used in DSC methods. In preliminary studies, both HB1 and HB2 resulted in the capture and PCR amplification of S. sonnei DNA from SB cultures, potato salad and ground beef samples (data not shown).

PAGE 137

124 Non-specific adsorption of DNA to CPShigella beads When HB1 was used in DSC methods, a consistent and reproducible increase (approx. 100-fold) in detection of S. sonnei DNA was observed when CPShigella beads were used instead of unlabeled beads. Th ere was, however, non-specific adsorption of DNA to the unlabeled beads. Se veral variations to the DSC method were investigated for the reduction of non-specific DNA adso rption, including the temperature of hybridization, the type of paramagnetic beads an d the bacterial cell titer in the sample to be analyzed. Hybridizations should normally be performed at a temperature 20-25C below the melting temperature of the probe -target hybrid (Castora and Greene, 1998). The melting temperature of the CPShigella probe hybrid was calc ulated to be 62.6C (Sigma-Genosys website, 2006); therefore hybr idization in the DSC method was performed at 40C. To reduce non-specific bi nding, a higher hybridizat ion temperature of 55C was investigated, however non-specific binding was not re duced (Table 4-8). It was also considered that the type of paramagne tic beads used in the preparation of CPShigella beads may affect non-specific bindi ng. Two types of streptavidin-coated paramagnetic beads were investigated, th e Dynabeads M-280 Streptavidin and the MagaBeads Streptavidin, however neither re sulted in the elimination of non-specific DNA binding (Table 4-7). Based on Ct values from real-time PCR, the Dynabeads M280 Streptavidin resulted in the larges t difference between DNA capture by the CPShigella beads and the unlabeled beads, i ndicating more specific capture of S. sonnei DNA. For this reason the Dynabeads M-280 Strept avidin were used for the preparation of CPShigella beads in all inoculated trials. Unexpectedly, the use of HB2 in DSC met hods gave poor results with respect to specific capture. Based on Ct values from PCR amplification, there was no difference

PAGE 138

125 between capture of S. sonnei DNA by CPShigella beads and unlabeled beads when HB2 was used in the DSC method (T able 4-6). It was considered that components of HB2, such as sodium dodecyl sulfate (SDS) a nd/or Sarkosyl, may have disrupted bonds between the streptavidin molecules and the bead surface, the streptavidin and biotinlabeled DNA probes, or both, thereby elim inating the specificity of the CPShigella beads over that of the unlabeled beads. In contrast, Tsai et al (2003) reported sequence capture of enterotoxigenic E. coli using a hybridization buffer similar to HB2, except that 1X SSC was used rather than 5X SSC to avoi d stringency problems in subsequent PCR. Non-specific adsorption of DNA to magne tic beads was not observed in this study, however, unlike the present st udy, hybridization was performe d at 85C. In preliminary studies it was observed that when HB2 wa s used for DSC, the cell membrane/food material pellet was solubilized during the he ating step and in many cases no visible pellet remained after subsequent centrifugation. It was not determined how the additional solutes may or may not have affected hybridization of S. sonnei DNA to the CPShigella beads in food samples. Several previous studies also report the specific capture of verotoxigenic E. coli (Chen et al ., 1998) and Salmonella and Shiga-like toxin producing E. coli (Chen and Griffiths, 2001) using sequence capture follo wed by PCR. In the first study, specific capture was determined using slot hybridizat ions, where target and non-target bacterial DNA were cross-linked to positively-charge d nylon membranes and hybridized to digoxigenin-labeled capture probes (Chen et al ., 1998). In the second study, specificity was determined through PCR amplification of DNA sequences, not by capture of target sequences to the capture probe/bead comple xes (Chen and Griffiths, 2001). Neither study

PAGE 139

126 investigated non-specific adsorp tion of DNA to streptavidin-c oated beads with or without the presence of the capture probe. Mangiapan et al (1996) reported the use of a tw o-step capture procedure where biotinylated capture probes were hybridized to target sequences in solution in the first step, then separated from the mixture in a second step by the addition of streptavidincoated paramagnetic beads to bind the biotin -labeled hybrids. The two-step approach increased the sensitivity of the assay >10-fold over the direct capture method (Mangiapan et al ., 1996). There was no mention that non-specific adsorption of DNA was investigated, therefore it is unc lear how the two-step capture compares to the non-specific DNA adsorption observed in the present study. Amagliani et al (2006) reported specif ic detection of the hlyA gene of Listeria monocytogenes using NH2-labeled probes immobilized on amino modified nanoparticles. Specificity was tested using fluorescein-la beled oligonucleotides with complementary and non-complementary sequences to that of the capture probe. Following hybridization, the nanoparticles were washed in buffer and th en heated to 80C for 4 min to dissociate the annealed sequences for spectr ophotometric analysis (Amagliani et al ., 2006). Supernatants from only hybridizations w ith complementary fluorescein-labeled oligonucleotides resulted in significant fluorescence; therefore non-complementary fluorescein-labeled oligonucleotides were not hybridized to the probe-labeled nanoparticles. Analysis of lowest detecti on levels of the DSC method The LDL and LDL100 were determined for the DSC method (Table 5-3). For tomato samples, the LDL and LDL100 were 2.0 x 101 CFU/tomato and 2.0 x 102 CFU/tomato, respectively. For potato salad samples, the LDL and LDL100 were both 5.6

PAGE 140

127 Table 5-3. Lowest detection levels of the DSC method. Tomatoes, potato salad and ground beef samples were inoc ulated with a five-strain S. sonnei cocktail at various levels and analyzed using DSC. The lowest detection level (LDL) and lowest detection level in which S. sonnei was detected in all 10 of the replicates (LDL100) were determined based on in itial inoculation levels. Food matrix LDL LDL100 Tomato 2.0 x 101 2.0 x 102 (CFU/tomato) Potato salad 5.6 x 100 5.6 x 100 (CFU/25 g) Ground beef 4.6 x 100 4.6 x 101 (CFU/25 g) x 100 CFU/25 g. As discussed above for the FTI methods, low indigenous microbial levels likely contributed to positive detection of S. sonnei in potato salad samples. For ground beef samples, the LDL and LDL100 were 1.5 x 100 CFU/25 g and 1.5 x 101 CFU/25 g, respectively. Based on the LDL and LDL100 values for ground beef, the DSC method was not affected by the presence of competitive background microorganisms as observed with the FTI methods. For this re ason, it is likely that the analysis of temperature abused potato salad containing high levels of background microorganisms by the DSC method would give similar resu lts as seen in the present study. The LDL values reported in the pres ent study are in agreement with those previously reported for DNA sequence cap ture of foodborne pathogens (Chen et al ., 1998; Chen and Griffiths, 2001; Tsai et al ., 2003; Amagliani et al ., 2006). Sequence capture followed by PCR was able to detect verotoxigenic E. coli at 100 CFU/g ground beef after 15 hr nonselective enrichment (Chen et al ., 1998) and Salmonella and Shigalike toxin producing E. coli at 100 CFU/ml after 10 hr nonselective enrichment (Chen and Griffiths, 2001). Tsai et al ., (2003) reported a detect ion limit of 1.0 CFU/liter environmental water after filtering the water sample through a 0.45 micron filter,

PAGE 141

128 incubating the filters face-down for approximately 24 hr on selective agar plates and extracting the DNA from suspect colonies prior to sequence capture-PCR. Finally, Amagliani et al (2006) reported an LDL for L. monocytogenes in fluid milk of 1.0 x 101 CFU/ml after recovering bacterial cells from 10 ml samples by centrifugation and extracting DNA from the resulting pellet. Cent rifugation of a 10-ml sample inoculated at 1.0 x 101 CFU/ml resulted in approx 1.0 x 102 CFU/sequence capture hybridization, which is in agreement with the sensitivity of the DSC method observed in the present study. Future research to improv e DSC for the detection of S. sonnei in food Several areas in which future research may improve DSC for the detection of S. sonnei in food include automation of bead handl ing steps and elimination of non-specific binding. Excessive time and labor was necessary to perform the beads washing steps as described for the DSC method, especially for th e number of samples analyzed in one day. The Dynal BeadRetriever (Invitrogen) is an instrument designed for automated IMS enrichment, using magnetic rods to transfer IMS beads from tube to tube (Invitrogen, 2006). Using the BeadRetriever, samples may be analyzed in a little as 20 min, which would greatly improve the number of analys es possible on one day. In addition to automation, the elimination of non-sp ecific adsorption of DNA to CPShigella beads may improve sensitivity of the DSC method. The presence of large amounts of non-target DNA have been demonstrated to reduce the sensitivity of PCR (Kramer and Coen, 1994), therefore reduction of nontarget DNA binding during DSC would likely improve the sensitivity of the overall method. Finally, a longer capture probe (>100 bases) for use with the CPShigella beads would allow hybridization at temperatures higher than 55C, which may impact non-specific DNA binding to the beads.

PAGE 142

129 Sources of Variation Among Inoculated Studies The BAM Shigella culture method, FTI and DSC e xperiments were not performed on the same day; therefore several preparations of potato salad and several different lots of ground beef were used in this study. Fo r this reason, total populations of background microflora and the types of mi croflora present varied among trials. The effects of this could be seen with the preliminary evaluations of antiShigella beads in ground beef versus the inoculated trials as described above. Nevertheless, E. coli and E. cloacae were isolated from all inoculated ground beef tria ls. For potato salad samples, variations in sugar content among potatoes may have resulte d in variable amounts of simple sugars among potato salad samples. Increased sugar content in potato salad samples may result in increased sugar concentrations in SB enrich ments. It has previously been reported that the acids produced by microbial fermentation of sugars may inhibit Shigella in broth cultures, thus the enrichment media reco mmended in the BAM contain very little carbohydrate. These sources of variation may be eliminated by performing all of the methods in parallel using potato salad or ground beef samples from the same preparation/purchase. In addition, potato salad and ground beef samp les were inoculated in 50-g aliquots, however only 25 g of the sample was transfer red to enrichment after the attachment period. Homogenation of the inocula was by ha nd using a sterile tongue depressor. For ground beef samples, it was difficult to ensure complete homogenation due to the tendency of ground beef to form together. Furt hermore, it was possible that for some of the samples at the low inoculation levels, none of the inoculum was in the 25-g aliquot transferred for enrichment but was rather discarded with the remaining sample. Initially, 50-g samples were inoculated so that each sample could be divided and used for two

PAGE 143

130 methods (FTI and DSC). As the experime nts were performed, only one method was investigated on a given day, however the inoculation method was maintained for consistency among trials. Variation due to this type of inoculation method could easily be eliminated by the inoculation of 25 g samples and then using the entire sample for each method. As with previous studies, tomatoes used in this study were direc tly out of the field and variations in filth, dirt or sand pres ent on the tomato fruits was observed among shipments (Warren, 2003). The e ffects of filth, dirt or sand on tomato surfaces on the attachment, recovery and survival of S. sonnei has yet to be determined. As previously suggested, an experiment in which recovered S. sonnei from both clean and filthy tomatoes are enumerated could provide insight as to the effect of filth, dirt or sand on attachment, recovery and survival (Warren, 2003). Practical Applications of Flow-Through Immunocapture and DNA Sequence Capture Recently proposed approaches for the detection of Shigella in food have combined conventional culture methods with PCR methods to facilitate specifi c and rapid detection while still providing the isolati on of viable colonies. The FT I methods developed in this study fit this type of appro ach, in that once the antiShigella beads are recovered, rapid and specific PCR analysis may be performed while at the same time an aliquot of the beads may be analyzed using culture tec hniques (spread-plate) which would provide isolated colonies. The FTI methods are some what limited by the numbe r of samples that could be processed in one day using a singl e Pathatrix unit, ther efore the FTI methods would be most useful where the analysis of a small number of samples (<30) each day is required. In contrast, the DSC method can be automated; therefore rapid and specific analysis of a large number of samples (>50) in a single day is possibl e. For this reason,

PAGE 144

131 the DSC method would be most useful in la boratories screening large numbers of food samples for the presence of S. sonnei Since the DSC method does not isolate viable colonies, it would not be useful when bioche mical or serological characterizations of detected strains are necessary. Conclusions The results if this study did not suppor t the hypothesis that specific antibodies and/or specific DNA probes may be attached to paramagnetic beads and used for the analysis of foods for the presence of S. sonnei with increased sensitivity over the BAM Shigella culture method. Potato salad and ground beef support the survival of S. sonnei beyond their shelf-life when held unde r normal refrigerated conditions, while S. sonnei declines rapidly on the smooth surfaces of tomatoes held under normal storage conditions. All of the methods developed in this study (FTI-MAC, FTI-PCR and DSC) performed as well or better than the BAM Shigella culture method for the detection of S. sonnei on tomato surfaces and in potato salad and in ground beef. The FTI-PCR and DSC methods may be used as rapid methods for the detection of S. sonnei since final results were obtained within 24 hr. The FTI-MAC me thod resulted in the isolation of viable S. sonnei colonies, however the lack of speci ficity of commercially available antiShigella antibodies resulted in the co-isolation of competitive microorganisms, such as E. coli and E. cloacae The DSC method was most succe ssful for the detection of S. sonnei amid high levels of competitive microorganisms.

PAGE 145

132 APPENDIX A PREPARATION OF BUFFERS AND SOLUTIONS Binding/Washing (B/W) Buffer (2X) Formula: 10 mM Tris-HCl, pH 7.5, 1.0 mM EDTA, 2.0 M NaCl 1. Calculate amount of stock 1.0 M Tris-HCl, pH 7.5 needed. (1.0 M) (x ml) = (0.01 M) (200 ml) = 2.0 ml of 1.0 M Tris-HCl, pH 7.5 2. Calculate amount of stock 0.25 M EDTA needed. (0.25 M) (x ml) = (0.001 M) (200 ml) = 0.8 ml 0.25 M EDTA 3. Calculate the amount of NaCl needed to make 200 ml. 2.0 mol/L 58.44 g/mol 0.2 L = 23.4 g NaCl 3. Dissolve 23.4 g NaCl in 197.2 ml DI water. 4. Add 2.0 ml of stock 1.0 M Tris-HCl, pH 7.5, and 0.8 ml of stock 0.25 M EDTA. 5. Add 0.2 ml diethyl pyrocarbonate (DEPC) and shake vigorously. 6. Incubate the solution at room temperature for 1 h. 7. Autoclave at 121C for 15 min. Congo Red Solution Formula: 1 mM Congo red 1. Calculate the amount of Congo Red needed. 0.001 mol/L 696.65 g/mol 0.2 L = 0.14 g Congo Red 2. Dissolve 3.4 g Congo Red in 200 ml DI water. 3. Filter sterilize the solution by passage through a 0.2 micron filter. 4. Store at room temperature.

PAGE 146

133 Dynabeads Solution A Formula: 0.1 M NaOH, 0.05 M NaCl 1. Calculate the amount NaOH needed to make 200 ml. 0.1 mol/L 40.00 g/mol 0.2 L = 0.8 g NaOH 2. Calculate the amount of NaCl needed to make 200 ml. 0.05 mol/L 58.44 g/mol 0.2 L = 0.58 g NaCl 3. Dissolve 0.8 g NaOH and 0.58 g Na Cl in 200 ml DI water. 4. Add 0.2 ml DEPC and shake vigorously. 5. Incubate the solution at room temperature for 1 h. 6. Autoclave at 121C for 15 min. Dynabeads Solution B Formula: 0.1 M NaOH 1. Calculate the amount NaOH needed to make 200 ml. 0.1 mol/L 40.00 g/mol 0.2 L = 0.8 g NaOH 2. Dissolve 0.8 g NaOH in 200 ml DI water. 3. Add 0.2 ml DEPC and shake vigorously. 4. Incubate the solution at room temperature for 1 h. 5. Autoclave at 121C for 15 min. Hybridization Buffer 1 Formula: 100 mM Tris-HCl, pH 7.5, 50 mM EDTA, 150 mM NaCl 1. Calculate amount of stock 1.0 M Tris-HCl, pH 7.5 is needed. (1.0 M) (x ml) = (0.1 M) (100 ml) = 10.0 ml of 1.0 M Tris-HCl, pH 7.5 2. Calculate amount of stock 0.25 M EDTA is needed. (0.25 M) (x ml) = (0.05 M) (100 ml) = 20 ml of 0.25 M EDTA

PAGE 147

134 3. Calculate the amount of stock 3.75 M NaCl needed. (3.75 M) (x ml) = (0.15 M) (100 ml) = 4.0 ml of 3.75 M NaCl 4. Combine 10.0 ml of stock 1.0 M Tris-HCl pH 7.5, 20 ml of stock 0.25 M EDTA and 4.0 ml of 3.75 M NaCl with 66.0 ml water.. 6. Add 0.1 ml DEPC and shake vigorously. 7. Incubate the solution at room temperature for 1 h. 8. Autoclave at 121C for 15 min. Hybridization Buffer 2 Formula: 5X Saline-Sodium Citrate (SSC), 0.1% Sarkosyl, 0.02% SDS, 1X Denhardt’s Blocking Reagent 1. Calculate amount of stock 20X SSC needed. (20X) (x ml) = (5X) (100 ml) = 25.0 ml of 20X SSC 2. Calculate amount of stoc k 1.0% Sarkosyl needed. (1.0%) (x ml) = (0.1%) (100 ml) = 10.0 ml of 1.0% Sarkosyl 3. Calculate the amount of stock 1.0% SDS needed. (1.0%) (x ml) = (0.02%) (100 ml) = 2.0 ml of 1.0% SDS 4. Calculate the amount of stock 50X Denhardt’s Blocking Reagent needed. (50X) (x ml) = (1X) (100 ml) = 2.0 ml of 50X Denhardt’s Blocking Reagent 5. Combine 25.0 ml 20X SSC, 10.0 ml 1.0% Sa rkosyl, 2.0 ml 1.0% SDS and 2.0 ml 50X Denhardt’s Bocki ng Reagent with 61.0 ml water. 6. Add 0.1 ml DEPC and shake vigorously. 7. Incubate the solution at room temperature for 1 h. 8. Autoclave at 121C for 15 min.

PAGE 148

135 Low-Salt Wash Buffer Formula: 0.15 M NaCl, 20 mM Tr is-HCl, pH 7.5, 1 mM EDTA 1. Calculate amount of stock 1.0 M Tris-HCl, pH 7.5 is needed. (1.0 M) (x ml) = (0.02 M) (200 ml) = 4.0 ml of 1.0 M Tris-HCl, pH 7.5 2. Calculate amount of stock 0.25 M EDTA is needed. (0.25 M) (x ml) = (0.001 M) (200 ml) = 0.8 ml 0.25 M EDTA 3. Calculate the amount of NaCl needed. 0.15 mol/L 58.44 g/mol 0.2 L = 1.8 g NaCl 4. Dissolve 1.8 g NaCl in 195.2 ml DI water. 5. Add 4.0 ml of stock 1.0 M Tris-HCl, pH 7.5, and 0.8 ml of stock 0.25 M EDTA. 6. Add 0.2 ml DEPC and shake vigorously. 7. Incubate the solution at room temperature for 1 h. 8. Autoclave at 121C for 15 min. Phosphate Buffered Saline, pH 7.4 Formula: 0.058 M Na2HPO4, 0.01 M NaH2PO4, 0.085 M NaCl 1. Calculate the amount of Na2HPO4 needed to make 1 L. 0.058 mol/L 141.08 g/mol = 8.2 g Na2HPO4 2. Calculate the amount of NaH2PO4 needed to make 1 L. 0.01 mol/L 119.96 g/mol = 1.2 g NaH2PO4 3. Calculate the amount of NaCl needed to make 1 L. 0.085 mol/L 58.44 g/mol = 5.0 g NaCl 4. Dissolve 8.2 g Na2HPO4, 1.2 g NaH2PO4, and 5.0 g NaCl in 1 L DI water. 5. Adjust pH if necessary using 1.0 N NaOH or 1.0 N HCl. 6. Autoclave at 121C for 15 min.

PAGE 149

136 Sodium Acetate Buffer, pH 4.0 Formula: 0.1 M NaC2H3O2 glacial acetic acid to pH 4.0 1. Calculate the amount of NaC2H3O2*3H2O needed to make 250 ml. 0.1 mol/L 136.08 g/mol 0.25 L = 3.4 g NaC2H3O2*3H2O 2. Dissolve 3.4 g NaC2H3O2*3H2O in 200 ml DI water. 3. Titrate to pH 4.0 with glacial acetic acid. 4. Adjust final volume to 250 ml using DI water. 5. Filter sterilize the solution by passage through a 0.2 micron filter. Sodium Bicarbonate Buffer, pH 8.6 Formula: 0.1 M Na2CO3, 0.1M NaHCO3 1. Make stock solution of 0.1M Na2CO3 (pH 8.07). 0.1 mol/L 105.99 g/mol = 10.6 g Na2CO3/L 2. Make stock solution of 0.1 M NaHCO3 (pH 11.0). 0.1 mol/L 84.01 g/mol = 8.4 g NaHCO3/L 3. Transfer 100 ml 0.1 M NaHCO3 and titrate to pH 8.6 (about 4.0 ml). 4. Filter sterilize the solution by passage through a 0.2 micron filter. Wash Buffer Formula: 0.5 M NaCl, 20 mM Tris-H Cl, pH 7.5, 1 mM EDTA 1. Calculate amount of stock 1.0 M Tris-HCl, pH 7.5 is needed. (1.0 M) (x ml) = (0.02 M) (200 ml) = 4.0 ml of 1.0 M Tris-HCl, pH 7.5 2. Calculate amount of stock 0.25 M EDTA is needed. (0.25 M) (x ml) = (0.001 M) (200 ml) = 0.8 ml 0.25 M EDTA 3. Calculate the amount of NaCl needed. 0.5 mol/L 58.44 g/mol 0.2 L = 5.9 g NaCl

PAGE 150

137 4. Dissolve 5.9 g NaCl in 195.2 ml DI water. 5. Add 4.0 ml of stock 1.0 M Tris-HCl, pH 7.5, and 0.8 ml of stock 0.25 M EDTA. 6. Add 0.2 ml DEPC and shake vigorously. 7. Incubate the solution at room temperature for 1 h. 8. Autoclave at 121C for 15 min.

PAGE 151

138 APPENDIX B ALIGNMENT OF CHROMOSOMALLY-LOCATED ipaH GENES OF Shigella sonnei ClustalW was used to align the sequen ces of the five chromosomally-located ipaH genes of S. sonnei Ss046. The ClustalW output is give n below with the conserved bases designated with asterisks (*). CLUSTAL W (1.82) multiple sequence alignment ipaH2 ---------ATGCTCCCGACAAATAACAATCACAGATTAATTTCAAATTCGTTCTCCACT 51 ipaH5 ------------ATGCGAGAAATTAATATGCTCAGAAATATTTCATCCTGTTTATTTCCA 48 ipaH3 ATGTCCATCATGTTACCGATAAATAATAA-CTTTTCATTGTCCCAAAAT---TCTTTTTA 56 ipaH4 ---------ATGCTTCCTGTAAATAATCC-CCCCCTATCCACTGGAAACGTCTCTTTTTA 50 ipaH1 ---------ATGAAACCTGCCCACAATCCTTCTTTTTTCCGCTCCTTTTGTGGTTTAGGA 51 ** ipaH2 TATTCAATCGACACTAG---CCGTGCATATGAAA------ATTATCTAACCCATTGGACT 102 ipaH5 CAT---ATCAGCACAAT---TACATCCCCCAACC------ATTATTTGTCCGAATGGGAT 96 ipaH3 TAA--CACTATTTC------CGGTACATATGCTG------ATTACTTTTCAGCATGGGAT 102 ipaH4 CAG--AACTACATCAATCGACAATGTTCACAATA------ATTATCTCTCCGAATGGGTT 102 ipaH1 TGTATATCCCGTTTATCCGTAGAAGAGCAAAATATCACGGATTATCACCGCATCTGGGAT 111 **** *** ipaH2 GAATGGAAAAATAACCGCATACAAGAAGAACAACGAGACATCGCTTTTCAGCGACTAGTA 162 ipaH5 GATTGGGAGAAACAGGGGTTACCGGAAGAACAGCGTACTGAGGCGGTAAGAAGACTTCGT 156 ipaH3 AAATGGGAAAAACAAGCGCTCCCCGGTGAAAATCGGAATGAAGCGGTCTCCCTACTTAAA 162 ipaH4 GAATGGACTAAAAACAGCATTTCCGGAGAAAACAGGGAAACTGCTTTTACCCGGCTCCAA 162 ipaH1 AACTGGGCCAAGGAAGGTGCTGCAACAGAAGACCGAACACAGGCAGTTCGATTACTGAAA 171 *** ** *** * ** ** ipaH2 TCATGTCTACAAAACCAAGAGACG-AACCTGGACTTGTCTGAATTAGGCCTGACAACATT 221 ipaH5 GCATGTCT-TACCTCTAAGGGGCATAAACTGGACCTGCGAGCCTTGGCGCTTTCCTCGTT 215 ipaH3 GAATGTCT-CATCAATCAGTTCAGTGAGCTTCAACTGAATCGTTTAAATCTGTCCTCGCT 221 ipaH4 TTATGTCT-GG------AGAACAGTGA-----AAC---ATCGTTGGA------------194 ipaH1 ATATGTCTGGCTTTTCAAGAGCCA-GCCCTCAATTTAAGTTTACTCAGATTACGCTCTCT 230 ****** ** ipaH2 ACCTGA--AATCCCCCCGGAAATTAAATCAATTAATATAAGTAAAAATAATTTAAGCTTA 279 ipaH5 ACCTGT--ACTCCCTGCTTGCATTAAAAAGCTTGATGTGAGCTGTAATAAATTAACCATC 273 ipaH3 ACCTGACAACTTACCACCTCAAATCACTGTTCTGGAAATTACTCAGAATGCCCTAATATC 281 ipaH4 ---------CTTATC-----------TTGTTTAGG-----TCTCAGAT--CTCTACCACG 227 ipaH1 CCCATAC--CTGCCCCCGCACATACAAGAACTTAACATC-TCTAGCAATGAGCTACGCTC 287 *

PAGE 152

139 ipaH2 ATCTCC-CCATTGCCTGCGTCCCTTACACAGCTTAATGTCAGCTATAACAGACTTATTGA 338 ipaH5 CTTACT-GATTTACCTGAAAATATTAAAGAACTTATTGCAAGAGATAATTTCTTAACACA 332 ipaH3 ATTACCAGAATTGCCAGCATCGCTGGAATACCTTGACGCCTGTGACAATCACCTGTCAAC 341 ipaH4 ATTGCCTGA-----CAATCTTGATGAAAT----TAATGTAAGCAATAACCAACTATCAAT 278 ipaH1 TCTGCCAGAACTCCCTCCGTCCTTAACTGTACTTAAAGCCAGCGATAACAGACTGAGCAG 347 * * * ** ipaH2 ACTGCCTGCTTTGCCTCAAGGACTTAAATTATTGAATGCGTCCCACAATCAA---CTAAT 395 ipaH5 TATATCTGCATTACCACATTATCTAATAACTTTGGATGTGTCCGAAAATCAA---TTAGA 389 ipaH3 ACTTCCTGAATTACCCGCATCTCTGAAACATCTTGATGTAGATAACAACCAA---CTAAC 398 ipaH4 GCTCCCCGAGCTACCAAGGGCATTGAAAGAGCTGAATGCAAGCAGTAATCAA---TTATC 335 ipaH1 GCTCCCGGCTCTTCCGCCTCACCTGGTCGCTCTTGATGTTTCACTTAACAGAGTTTTAAC 407 * * ** * *** ** ** ipaH2 CACACTACCCACACTCCCCATATCTTTGAAGGAGCTTCATGTCTCAAATAATCAATTATG 455 ipaH5 GAATCTGCCGTTATTACCAGACACCATCAAATCACTAAGCGCAGAGTATAATAGGTTATC 449 ipaH3 CATGCTTCCTGAATTGCCTGCATTGCTGGAATATATTAATGCAGATAACAATCAGCTAAC 458 ipaH4 TGCACTTCCTGAATTACCAGTGTCGCTGGAATATATAAATGTGAGTGATAACCATTTGTT 395 ipaH1 ATGTTTGCCTTCTCTTCCATCTTCCTTGCAGTCACTCTCAGCCCTTCTCAATAGCCTGGA 467 ** ** * * ** ipaH2 TTCTCTTCCTGTTTTACCAGAACTACTGGAAACATTAGATGTATCATGTAATGGGCTGGC 515 ipaH5 CACACTGCCTTCATTACCCTTGAATTTAAAAAAACTTGAGGTTAGGAACAACGAACTGCA 509 ipaH3 CATGCTTCCTGAATTACCTACATCGCTGGAAGTGCTCTCAGTAAGAAATAACCAGCTGAC 518 ipaH4 CGCACTTCCTGAATTACCTGCGTCACTAGAATATATTAATGTAAGTGACAATCACCTGTC 455 ipaH1 GACGCTACCTGATCTTCCCCCGGCTCTACAAAAACTTTCTGTTGGCAACAACCAGCTTAC 527 ** *** ** ** ** ** ** ipaH2 AGTTTTACCACCTTTACCATTTTCTTTACAAGAGATTAGCGCAATAGGGAATCTTCTTAG 575 ipaH5 AACTCTTCCATCTCTGCCTTCTAATCTTAAGATACTTAAGGTTGCGCACAACCATCTTAC 569 ipaH3 ATTTCTCCCTGAGTTACCTGAATCACTGGAAGCGCTCGATGTAAGTACTAATCTTCTGGA 578 ipaH4 TGTACTTCCGAGGTTACCAATGTCATTGGAATTACTTGATGCAGCCAGAAATGCTTTGGA 515 ipaH1 TGCCTTACCAGAATTACCATGTGAACTACAGGAACTAAGTGCTTTTGATAACAGATTACA 587 ** ** * * ** ipaH2 TGAACTCCCCCCTCTACCTCACAACATTCACTCCATATGGGCAATCGACAATATGTTAAC 635 ipaH5 TGAACTGCCCCCTTTACCTAGGAGACTGCAACTTCTTTTTGCATATAGCAATAGATTAAG 629 ipaH3 AAGCCTACCAGCCGTACCT----------------------------GTAAGAAATCAT609 ipaH4 AGTAATACCAGATTTTCCA----------------------------GAAAGAGATGAT546 ipaH1 AGAGCTACCGCCCCTTCCTCAAAATCTGAGGCTTTTAAACGTTGGGGAAAACCAACTACA 647 ** ** ** ipaH2 CGATATTCCATACCTGCCGGAAAAT----TTAAGGAACGGTT-----ATTTTGACATAAA 686 ipaH5 CAACTTACCAAACATCCAAGAAAATATTATCATGAGAAGATTTTTTTATTTTGAAAACAA 689 ipaH3 ------------CACTCAGAGGAA-------ACCGAGATATTTTTCCGGTGCCGCGAGAA 650 ipaH4 ------------CAT----------------ATTATAAGAATATTCTGGCTTAATCAGAA 578 ipaH1 CAGACTGCCCGAACTTCCACAACG---------TCTGCAATCACTATATATCCCTAACAA 698 ** ipaH2 TCAGATAAGTCATATCCCGGAAAGCATTCTTAATCTGAGGAATGAATGTTCAATAGATAT 746 ipaH5 CCAAATAACTACAATCCCGACAAATCTTTTTCGTTTAGATCCTCATATAACTATTGAGAT 749 ipaH3 TCGCATCACATACATTCCGGAAAATATACTTAGCCTTGATCCGACCTGCACTATCATCCT 710 ipaH4 CCGGATCACGGCAATTCCGGAAAGCATACTTGGCCTCAGTTCTGATAGCGTTGTCAATCT 638 ipaH1 TCAGCTGAACACATTGCCAGACAGTATCATGAATCTGCACATTTATGCAGATGTTAATAT 758 * * ** * * *

PAGE 153

140 ipaH2 TAGTGATAACCCATTGTCATCCCATGCTCTGCAATCCCTG-------------------C 787 ipaH5 TGCAAATAACCCCTTATCAGATCAAACTCTGCTATTCTTAATACAGCAAACTTCGGTTCC 809 ipaH3 CGAAGACAATCCTCTGTCCTCACGGATCAGGGAGTCTCTG-------------------750 ipaH4 TAGAGAAAATCAACTATCTCCCAGAATAATGCAAACTTTG-------------------678 ipaH1 TTATAACAATCCATTGTCGACTCGCACTCTGCAAGCCCTG-------------------C 799 ** * ** * ipaH2 AAAGATTAAC-----------------CTCTTCGCCGGACTACCACGGCCCGC------823 ipaH5 AAATTTTAACGGGCCTCAGTTTCGTATTTCCCTGTCAGACCAAAACAGACTGTTTTTACG 869 ipaH3 ---TCGCAACAAAC-------------CGCCCAACCGGACTACCACGGCCCAC------787 ipaH4 ---TTACAACAAAC-------------CGCCCAACCGGACTACCACGGCCCAC------715 ipaH1 AAAGATTAAC-----------------CTCTTCGCCGGACTACCACGGCCCAC------835 *** * *** ** * ipaH2 ---------------AGATTTACTTCT--------------CCATGAGTGACGGACAACA 854 ipaH5 CCAGATGTTGCCGCAAAATTTACATTCGCGCCATATCAGAGTCATCACTGAAGGGGGGCA 929 ipaH3 ---------------GGATTTACTTCT--------------CCATGAGTGACGGACAACA 818 ipaH4 ---------------GGATTTACTTCT--------------CCATGAGTGACGGACAACA 746 ipaH1 ---------------GGATTTACTTCT--------------CCATGAGTGACGGACAACA 866 ****** *** *** ** ** ipaH2 GAATACACTCCATCGCCCCCTGGCTGATGCCGTGACAGCATGGTTCCCGGAAAACAAACA 914 ipaH5 GAACTTTCAGATCCCCCCTCTTCCCGAAACTGTGGCAGCCTGGTTTCCTGAAGCAGATCG 989 ipaH3 GAATACACTCCATCGCCCCCTGGCTGATGCCGTGACAGCATGGTTCCCGGAAAACAAACA 878 ipaH4 GAATACACTCCATCGCCCCCTGGCTGATGCCGTGACAGCATGGTTCCCGGAAAACAAACA 806 ipaH1 GAATACACTCCATCGCCCCCTGGCTGATGCCGTGACAGCATGGTTCCCGGAAAACAAACA 926 *** * *** ** ** *** **** ***** ** *** * ipaH2 ATCTGATGTATCACAGATATGGCATGCTTTTGAACATGAAGAGCATGCCAACACCTTTTC 974 ipaH5 TCGGGAGGTTTCTACACAATGGACTTCTTTTTCCACCGAGGAGAATTCCCGGGCATTCTC 1049 ipaH3 ATCTGATGTATCACAGATATGGCATGCTTTTGAACATGAAGAGCACGCCAACACCTTTTC 938 ipaH4 ATCTGATGTATCACAGATATGGCATGCTTTTGAACATGAAGAGCACGCCAACACCTTTTC 866 ipaH1 ATCTGATGTATCACAGATATGGCATGCTTTTGAACATGAAGAGCACGCCAACACCTTTTC 986 ** ** ** **** ***** ** *** ** ** ** ipaH2 CGCGTTCCTTGACCGCCTTTCCGATACCGTCTCTGCACGCAATACCTCCGGATTCCGTGA 1034 ipaH5 CGCGTTCCTTGACCGCCTTTCCGATACCGTCTCTGCACGCAATACCTCCGGATTCCGTGA 1109 ipaH3 CGCGTTCCTTGACCGCCTTTCCGATACCGTCTCTGCACGCAATACCTCCGGATTCCGTGA 998 ipaH4 CGCGTTCCTTGACCGCCTTTCCGATACCGTCTCTGCACGCAATACCTCCGGATTCCGTGA 926 ipaH1 CGCGTTCCTTGACCGCCTTTCCGATACCGTCTCTGCACGCAATACCTCCGGATTCCGTGA 1046 ************************************************************ ipaH2 ACAGGTCGCTGCATGGCTGGAAAAACTCAGTGCCTCTGCGGAGCTTCGACAGCAGTCTTT 1094 ipaH5 ACAGGTCGCTGCATGGCTGGAAAAACTCAGTGCCTCTGCGGAGCTTCGACAGCAGTCTTT 1169 ipaH3 ACAGGTCGCTGCATGGCTGGAAAAACTCAGTGCCTCTGCGGAGCTTCGACAGCAGTCTTT 1058 ipaH4 ACAGGTCGCTGCATGGCTGGAAAAACTCAGTGCCTCTGCGGAGCTTCGACAGCAGTCTTT 986 ipaH1 ACAGGTCGCTGCATGGCTGGAAAAACTCAGTGCCTCTGCGGAGCTTCGACAGCAGTCTTT 1106 ************************************************************ ipaH2 CGCTGTTGCTGCTGATGCCACTGAGAGCTGTGAGGACCGTGTCGCGCTCACATGGAACAA 1154 ipaH5 CACTGTTGCTGCTGATGCCACTGAGAGCTGTGAGGACCGTGTCGCGCTCACATGGAACAA 1229 ipaH3 CGCTGTTGCTGCTGATGCCACTGAGAGCTGTGAGGACCGTGTCGCGCTCACATGGAACAA 1118 ipaH4 CACTGTTGCTGCTGATGCCACTGAGAGCTGTGAGGACCGTGTCGCGCTCACATGGAACAA 1046 ipaH1 CGCTGTTGCTGCTGATGCCACTGAGAGCTGTGAGGACCGTGTCGCGCTCACATGGAACAA 1166 **********************************************************

PAGE 154

141 ipaH2 TCTCCGGAAAACCCTCCTGGTCCATCAGGCATCAGAAGGCCTTTTCGATAATGATACCGG 1214 ipaH5 TCTCCGGAAAACCCTCCTGGTCCATCAGGCATCAGAAGGCCTTTTCGATAATGATACCGG 1289 ipaH3 TCTCCGGAAAACCCTCCTGGTCCATCAGGCATCTGAAGGCCTTTTCGATAATGATACCGG 1178 ipaH4 TCTCCGGAAAACCCTCCTGGTCCATCAGGCATCAGAAGGCCTTTTCGATAATGATACCGG 1106 ipaH1 TCTCCGGAAAACCCTCCTGGTCCATCAGGCATCTGAAGGCCTTTTCGATAATGATACCGG 1226 ********************************* ************************** ipaH2 CGCTCTGCTCTCCCTGGGCAGGGAAATGTTCCGCCTCGAAATTCTGGAGGACATTGCCCG 1274 ipaH5 CGCTCTGCTCTCCCTGGGCAGGGAAATGTTCCGCCTCGAAATTCTGGAGGACATTGCCCG 1349 ipaH3 CGCTCTGCTCTCCCTGGGCAGGGAAATGTTCCGCCTCGAAATTCTGGAGGACATTGCCCG 1238 ipaH4 CGCTCTGCTCTCCCTGGGCAGGGAAATGTTCCGCCTCGAAATTCTGGAGGACATTGCCCG 1166 ipaH1 CGCTCTGCTCTCCCTGGGCAGGGAAATGTTCCGCCTCGAAATTCTGGAGGACATTGCCCG 1286 ************************************************************ ipaH2 GGATAAAGTCAGAACTCTCCATTTTGTGGATGAGATAGAAGTCTACCTGGCCTTCCAGAC 1334 ipaH5 GGATAAAGTCAGAACTCTCCATTTTGTGGATGAGATAGAAGTCTACCTGGCCTTCCAGAC 1409 ipaH3 GGATAAAGTCAGAACTCTCCATTTTGTGGATGAGATAGAAGTCTACCTGGCCTTCCAGAC 1298 ipaH4 GGATAAAGTCAGAACTCTCCATTTTGTGGATGAGATAGAAGTCTACCTGGCCTTCCAGAC 1226 ipaH1 GGATAAAGTCAGAACTCTCCATTTTGTGGATGAGATAGAAGTCTACCTGGCCTTCCAGAC 1346 ************************************************************ ipaH2 CATGCTCGCAGAGAAACTTCAGCTCTCCACTGCCGTGAAGGAAATGCGTTTCTATGGCGT 1394 ipaH5 CATGCTCGCAGAGAAACTTCAGCTCTCCACTGCCGTGAAGGAAATGCGTTTCTATGGCGT 1469 ipaH3 CATGCTCGCAGAGAAACTTCAGCTCTCCACTGCCGTGAAGGAAATGCGTTTCTATGGCGT 1358 ipaH4 CATGCTCGCAGAGAAACTTCAGCTCTCCACTGCCGTGAAGGAAATGCGTTTCTATGGCGT 1286 ipaH1 CATGCTCGCAGAGAAACTTCAGCTCTCCACTGCCGTGAAGGAAATGCGTTTCTATGGCGT 1406 ************************************************************ ipaH2 GTCGGGAGTGACAGCAAATGACCTCCGCACTGCCGAAGCCATGGTCAGAAGCCGTGAAGA 1454 ipaH5 GTCGGGAGTGACAGCAAATGACCTCCGCACTGCCGAAGCCATGGTCAGAAGCCGTGAAGA 1529 ipaH3 GTCGGGAGTGACAGCAAATGACCTCCGCACTGCCGAAGCCATGGTCAGAAGCCGTGAAGA 1418 ipaH4 GTCGGGAGTGACAGCAAATGACCTCCGCACTGCCGAAGCCATGGTCAGAAGCCGTGAAGA 1346 ipaH1 GTCGGGAGTGACAGCAAATGACCTCCGCACTGCCGAAGCCATGGTCAGAAGCCGTGAAGA 1466 ************************************************************ ipaH2 GAATGAATTTACGGACTGGTTCTCCCTCTGGGGACCATGGCATGCTGTACTGAAGCGTAC 1514 ipaH5 GAATGAATTTACGGACTGGTTCTCCCTCTGGGGACCATGGCATGCTGTACTGAAGCGTAC 1589 ipaH3 GAATGAATTTACGGACTGGTTCTCCCTCTGGGGACCATGGCATGCTGTACTGAAGCGTAC 1478 ipaH4 GAATGAATTTACGGACTGGTTCTCCCTCTGGGGACCATGGCATGCTGTACTGAAGCGTAC 1406 ipaH1 GAATGAATTTACGGACTGGTTCTCCCTCTGGGGACCATGGCATGCTGTACTGAAGCGTAC 1526 ************************************************************ ipaH2 GGAAGCTGACCGCTGGGCGCTGGCAGAAGAGCAGAAATATGAGATGCTGGAGAATGAGTA 1574 ipaH5 GGAAGCTGACCGCTGGGCGCTGGCAGAAGAGCAGAAATATGAGATGCTGGAGAATGAGTA 1649 ipaH3 GGAAGCTGACCGCTGGGCGCTGGCAGAAGAGCAGAAATATGAGATGCTGGAGAATGAGTA 1538 ipaH4 GGAAGCTGACCGCTGGGCGCTGGCAGAAGAGCAGAAATATGAGATGCTGGAGAATGAGTA 1466 ipaH1 GGAAGCTGACCGCTGGGCGCTGGCAGAAGAGCAGAAATATGAGATGCTGGAGAATGAGTA 1586 ************************************************************ ipaH2 CCCTCAGAGGGTGGCTGACCGGCTGAAAGCATCAGGTCTGAGCGGTGATGCGGATGCGGA 1634 ipaH5 CCCTCAGAGGGTGGCTGACCGGCTGAAAGCATCAGGTCTGAGCGGTGATGCGGATGCGGA 1709 ipaH3 CCCTCAGAGGGTGGCTGACCGGCTGAAAGCATCAGGTCTGAGCGGTGATGCGGATGCGGA 1598 ipaH4 CCCTCAGAGGGTGGCTGACCGGCTGAAAGCATCAGGTCTGAGCGGTGATGCGGATGCGGA 1526 ipaH1 CCCTCAGAGGGTGGCTGACCGGCTGAAAGCATCAGGTCTGAGCGGTGATGCGGATGCGGA 1646 ************************************************************

PAGE 155

142 ipaH2 GAGGGAAGCCGGTGCACAGGTGATGCGTGAGACTGAACAGCAGATTTACCGTCAGCTGAC 1694 ipaH5 GAGGGAAGCCGGTGCACAGGTGATGCGTGAGACTGAACAGCAGATTTACCGTCAGCTGAC 1769 ipaH3 GAGGGAAGCCGGTGCACAGGTGATGCGTGAGACTGAACAGCAGATTTACCGTCAGCTGAC 1658 ipaH4 GAGGGAAGCCGGTGCACAGGTGATGCGTGAGACTGAACAGCAGATTTACCGTCAGCTGAC 1586 ipaH1 GAGGGAAGCCGGTGCACAGGTGATGCGTGAGACTGAACAGCAGATTTACCGTCAGCTGAC 1706 ************************************************************ ipaH2 TGACGAGGTACTGGCCCTGCGATTGTCTGAAAACGGCTCACAACTGCACCATTCATAA 1752 ipaH5 TGACGAGGTACTGGCCCTGCGATTGTCTGAAAACGGCTCACAACTGCACCATTCATAA 1827 ipaH3 TGACGAGGTACTGGCCCTGCGATTGTCTGAAAACGGCTCACAACTGCACCATTCATAA 1716 ipaH4 TGACGAGGTACTGGCCCTGCGATTGTCTGAAAACGGCTCACAACTGCACCATTCATAA 1644 ipaH1 TGACGAGGTACTGGCCCTGCGATTGCCTGAAAACGGCTCACAACTGCACCATTCATAA 1764 ************************* ********************************

PAGE 156

143 LIST OF REFERENCES Achi, R., L. Mata, X. Silas and A.A. Lindberg. 1996. Immunomagnetic separation and PCR detection show Shigellae to be comm on faecal agents in children from urban marginal communities of Costa Rica. J. Infect. 32: 211-218. Achi-Berglund, R. and A.A. Lindberg. 1996. Rapid and sensitive detection of Shigella sonnei in feces by the use of an O-antigen-specific monoclonal antibody in a combined immunomagnetic separation-pol ymerase chain reaction assay. Clin. Microbiol. Infect. 2: 55-58. al-Lahham A.B., M. Abu-Saud and A.A. Shehabi. 1990. Prevalence of Salmonella Shigella and intestinal parasites in food handler s in Irbid, Jordan. J Diarrhoeal Dis. Res. 8: 160-2. Amagliani, G., E. Omiccioli, A. del Campo, I. J. Bruce, G. Brandi and M. Magnani. 2006. Development of a magnetic capture hybridization-PCR assay for Listeria monocytogenes direct detection in milk sample s. J. Appl. Microbiol. 100: 375-383. [Anonymous]. 2002. AboutShigella Available at: www.about-shigella.com Accessed 05 Nov 2004. Arthur, T.M., J.M. Bosilevac, X. Nou and M. Koohmaraie. 2005. Evaluation of cultureand PCR-based detection methods for Escherichia coli O157:H7 in inoculated ground beef. J. Food Prot. 68: 1566-1574. Bagamboula, C., M. Uyttendaele and J. Debevere. 2002. Acid tolerance of Shigella sonnei and Shigella flexneri J. Appl. Microbiol. 93: 479-486. Bhat, P. and D. Rajan. 1975. Comparative ev aluation of desoxycholate citrate medium and xylose lysine desoxycholate medium in the isolation of Shigellae. Am. J. Clin. Pathol. 64: 399-403. Blostein, J. 1991. Shigellosis from swimming in a park pond in Michigan. Public Health Rep.106: 317-322. Bouvet, P.J. and S. Jeanjean. 1992. Evaluation of two colored latex kits, the Wellcolex Colour Salmonella Test and the Wellcolex Colour Shigella Test, for serological grouping of Salmonella and Shigella species. J. Clin. Microbiol. 30: 2184-2186.

PAGE 157

144 Carlin, N.I.A. and A.A. Lindberg. 1 986. Monoclonal antibodies specific for Shigella flexneri lipopolysaccharides: clones binding to type I and type II: 6,7,8 antigens, group 6 antigen, and a core epitope. Infect. Immun. 53: 103-109. Castora, F.J. and J.J. Greene 1998. Hybridization probes. In: Recombinant DNA principles and methodologies pp.335-351. Greene, J.J. and Rao, V.B., Eds., Marcel Dekker, Inc., New York, NY. [CDC] Centers for Disease Control and Prevention. 1999. Outbreaks of Shigella sonnei infection associated with eating fresh pa rsley--United States and Canada, July – August 1998. Morb. Mortal. Wkly. Rep. 48: 285-289. [CDC] Centers for Disease Control and Pr evention. 2000. Public health dispatch: outbreak of Shigella sonnei infections associated with eating a nationally distributed dip -California, Ore gon, and Washington, January 2000. Morb. Mortal. Wkly. Rep. 49: 60-61. [CDC] Centers for Disease Control and Prev ention. 2001. Shigellosis outbreak associated with an unchlorinated fill-and-drai n wading pool--Iowa, 2001. Morb. Mortal. Wkly. Rep. 50: 797-800. [CDC] Centers for Disease Control and Prevention. 2002. PHLIS surveillance data. Shigella : annual summary, 2002. Available at: http://www.cdc.gov/ncidod/dbmd/phlisdata/shigella.htm Accessed 04 Jun 2006. [CDC] Centers for Disease Control and Prevention. 2003. PHLIS surveillance data. Shigella : annual summary, 2003. Available at: http://www.cdc.gov/ncidod/dbmd/phlisdata/shigella.htm Accessed 04 Jun 2006. [CDC] Centers for Disease Control and Prevention. 2004. PHLIS surveillance data. Shigella : annual summary, 2004. Available at: http://www.cdc.gov/ncidod/dbmd/phlisdata/shigella.htm Accessed 04 Jun 2006. [CDC] Centers for Disease Control and Prevention. 2006a. Disease information: shigellosis. Available at: http://www.cdc.gov/ncidod/dbmd/di seaseinfo/shigellosis_t.htm Accessed 05 Nov 2004. [CDC] Centers for Disease Control and Pr evention. 2006b. Preliminary FoodNet data on the incidence of infection with path ogens transmitted commonly through food --10 States, United States, 2005. Morb Mortal. Wkly. Rep. 55: 392-395. Chen, J. and M.W Griffiths. 2001. Detection of Salmonella and simultaneous detection of Salmonella and Shiga-like toxin-producing Escherichia coli using magnetic capture hybridization polymerase chain reacti on. Lett. Appl. Microbiol. 32: 7-11. Chen, J., R. Johnson, and M.W. Griff iths. 1998. Detection of verotoxigenic Escherichia coli by magnetic capture-hybridization. Appl Environ. Microbiol. 64: 147-152.

PAGE 158

145 [CDPH] Chicago Department of Public Health. 1999. Small cl uster of a rare Shigella serotype. Available at: http://egov.cityofchicago .org/webportal/COCWebPortal/COC_EDITORIAL/shig_a rbo_may_99.pdf Accessed 05 Nov 2004. Cook, K.A., T. Boyce, C. Langkop, K. Kuo, M. Swartz, D. Ewert, E. Sowers, J. Wells, and R. Tauxe. 1995. Scallions and shigello sis: a multistate outbreak traced to imported green onions. Abstract. Proc. 44t h Ann. Conf. Epid. Intel. Serv. p. 36. March 27-31; Atlanta, GA. Cudjoe, K.S. and R. Krona. 1997. Detection of Salmonella from raw food samples using Dynabeads antiSalmonella and a conventional reference method. Int. J. Food Microbiol. 37: 55-62. Davis, H.J., P. Taylor, J.N. Perdue, G.N. Stelma, J.M. Humphreys, R. Rowntree, and K.D. Greene. 1988. A shigellosis outbreak traced to commercially distributed lettuce. Am. J. Epidemiol. 128: 1312-1321. Del Gallo, M.L., F. Lucarelli, E. Cosulich, E. Pistarino, B. Santamaria, G. Marrazza, and M. Mascini. 2005. Steric factors cont rolling the surface hybridization of PCR amplified sequences. Anal. Chem. 77: 6324-6330. Dmitriev, B.A., Y.A. Knirel, O.K. Sheremet, A.A. Shashkov, N.K. Kochetkov, and I.L. Hofman. 1979. Somatic antigens of Shigella The structure of the specific polysaccharide of Shigella Newcastle ( Sh. flexneri type 6) lipopolysaccharide. Eur. J. Biochem. 98: 309-316. Drysdale, M., M. MacRae, N.J.C. Strachan, T.M.S. Reid, and I.D. Ogden. 2004. The detection of non-O157 E. coli in food by immunomagnetic separation. J. Appl. Microbiol. 97: 220-224. Dunn, R.A., W.N. Hall, J.V. Altamirano, S.E. Dietrich, B. Robinson-Dunn, and D.R. Johnson. 1995. Outbreak of Shigella flexneri linked to salad prepared at a central commissary in Michigan. Public Health Reports. 110: 580-586. Echeverria, P., O. Sethabutr, and C. Pita rangsi. 1991. Microbiology and diagnosis of infections with Shigella and enteroinvasive Escherichia coli Rev. Infect. Dis. 13(Suppl 4): S220-225. Escartin, E.F., A.C. Ayala, and J.S. Lozano. 1989. Survival and growth of Salmonella and Shigella on sliced fresh fruit. J. Food Prot. 52: 471-472, 483. Evans, D.A., D.J. Hankinson, and W. Litsky. 197 0. Heat resistance of certain pathogenic bacteria in milk using a commercial plat e heat exchanger. J. Dairy Sci. 53: 16591665.

PAGE 159

146 Fernandez-Prada, C.M., D.L. Hoover, B.D. Ta ll, A.B. Hartman, J. Kopelowitz, and M.M. Venkatesan. 2000. Shigella flexneri IpaH7.8 facilitates escape of virulent bacteria from the endocytic vacuoles of mouse and human macrophages. Infect. Immun. 68: 3608-3619. Fleming, C.A., D. Caron, J.E. Gunn, M.S. Ho rine, B.T. Matyas, and M.A. Barry. 2000. An outbreak of Shigella sonnei associated with a recr eational spray fountain. Am. J. Public Health. 90: 1641-1642. Flessa, S., R.F. Vogel, and L.J. Harris. 2003. Survival of Shigella sonnei during dessication on surfaces is dependant upon density of inoculum and inoculum carrier. Poster presented at International Associati on for Food Protection Annual Meeting. August 10-13; New Orleans, LA. [FDA] Food and Drug Administration. 1998. Bacteriological analytical manual Shigella 8th Edition. AOAC International, Arlington, VA. [FDA] Food and Drug Administration. 2001a. F DA survey of imported fresh produce, FY 1999 field assignment. Available at: http://www.cfsan.fda.gov/~dms/prodsur6.html Accessed 05 Nov 2004. [FDA] Food and Drug Administration. 2001b. F DA survey of imported fresh produce: imported produce assignment FY 2001. Available at: http://www.cfsan.fda.gov/~dms/prodsur7.html Accessed 05 Nov 2004. [FDA] Food and Drug Administration. 2003. Surv ey of domestic fresh produce, FY 2000/2001 field assignment. Available at: http://www.cfsan.fda.gov/~dms/prodsu10.html Accessed 05 Nov 2004. Fredlund, H., E. Back, L. Sjoberg, and E. Tornquist. 1987. Water-melon as a vehicle of transmission of shigellosis. Scand. H. Infect. Dis. 19: 219-219. Frost, J.A., B. McEvoy, C.A. Bentley, Y. A ndersson, and A. Rowe. 1995. An outbreak of Shigella sonnei infection associated with consum ption of iceberg lettuce. Emerg. Infect. Dis. 1: 26-29. Fukiya, S., H. Mizoguchi, T. Tobe, and H. Mori. 2004. Extensive genomic diversity in pathogenic Escherichia coli and Shigella strains revealed by comparative genomic hybridization microarray. J. Bacteriol. 186: 3911-3921. Galanakis, E., M. Tzoufi, M. Charisi, S. Levidiotou, and Z. Papadopoulou. 2002. Rate of seizures in children with shigello sis. Acta Paedia trica. 91: 1001-2. Gamian, A. and E. Romanowska. 1982. The core structure of Shigella sonnei lipopolysaccharide and the linkage between O-specific polysaccharide and the core region. Eur. J. Biochem. 129: 105-109.

PAGE 160

147 Garca-Fulgueiras, A., S. Snchez, J.J. Guilln, B. Marsilla, A. Aladuena, and C. Navarro. 2001. A large outbreak of Shigella sonnei gastroenteritis associated with consumption of fresh pasteurised milk cheese. Eur. J. Epidemiol. 17: 533-538. Hale, T. 1991. Genetic basis of virulence in Shigella species. Microbiol. Rev. 55: 206224. Haraga, A. and S.I. Miller. 2003. A Salmonella enterica serovar typhimurium translocated leucine-rich repeat effector protein inhibits NF-kappa B-dependant gene expression. Infect. Immun. 71: 4052-4058. Health Canada. 2004. Compendium of analy tical methods. MFLP-25: Isolation and identification of Shigella spp. from foods. Available at: http://www.hcsc.gc.ca/food-aliment/mh-dm/mhe-dme/ compendium/volume_3/e_mflp-2501.html Accessed 05 Nov 2004. Hsih, H.Y. and H.Y. Tsen. 2001. Combina tion of immunomagnetic separation and polymerase chain reaction for the simultaneous detection of Listeria monocytogenes and Salmonella spp. in food samples. J. Food Prot. 64: 1744-1750. Ingersoll, M., E.A. Groisman, and A. Zychlinsky. 2002. Pathogenicity islands of Shigella Curr. Topics Microbiol. Immunol. 264: 49-65. [ICMSF] International Commission on Microb iological Specificat ions for Foods. 1996. Shigella In: Microorganisms in Foods, 5 pp. 280-298. Roberts, T.A., BairdParker, A.C., and Tompkins, R.B., Eds ., Blackie Academic & Professional. Chapman & Hall, London, UK. Invitrogen Corporation. 2006. Auto mated IMS from Dynal Biotech. http://www.invitrogen.com/content/sfs/brochures/Beadretriever.pdf Accessed 15 Jun 2006. Islam, D. and A.A. Li ndberg. 1992. Detection of Shigella dysenteriae type 1 and Shigella flexneri in feces by immunomagnetic isolati on and polymerase chain reaction. J. Clin. Microbiol. 30: 2801-2806. Islam, D., S. Tzipori, M. Islam, a nd A.A. Lindberg. 1993. Rapid detection of Shigella dysenteriae and Shigella flexneri in faeces by an immuno magnetic assay with monoclonal antibodies. Eur. J. Cli n. Mibrobiol. Infect. Dis. 12: 25-32. Islam, M.S., M.A. Hossain, S.I. Khan, M.N.H. Khan, R.B. Sack, M.J. Albert, A. Huq, and R.R. Colwell. 2001. Survival of Shigella dysenteriae type 1 on fomites. J. Health Popul. Nutr. 19: 177-182. Islam, M.S., M.K. Hasan, and S.I. Khan. 1993. Growth and survival of Shigella flexneri in common Bangladeshi foods under various conditions of time and temperature. Appl. Environ. Microbiol. 59: 652-654.

PAGE 161

148 Jacobson, A.P., M.L. Johnson, T.S. Hammac k, and W.H. Andrews. 2002. Evaluation of the Bacteriological Analytical Manual (BAM) culture method for the detection of Shigella sonnei in selected types of produce. Po ster presented at FDA Science Forum. February 20-21; Washington, D.C. Jansson, P.-E., A.A. Lindberg, B. Lindberg, and R. Wollin. 1981. Structural studies on the hexose region of the core in lipopolysaccharides from Enterobacteriaceae Eur. J. Biochem. 115: 571-577. Ji, N.N., M.M. Liu, X.R. Huang., J. Zhen, S. S. Li, S. Jiang, H.H. Yu, S.Y. Wang, and X.X. Peng. 2006. Immunocapture UPPCR combined with DGGE for rapid detection of Shigella species. J. Appl. Microbiol. 10:795-799. Jin, Q., Z. Yuan, J. Xu, Y. Wang, Y. Shen, W. Lu, J. Wang, H. Liu, J. Yang, F. Yang, X. Zhang, J. Zhang, G. Yang, H. Wu, D. Qu, J. Dong, L. Sun, Y. Xue, A. Zhao, Y. Gao, J. Zhu, B. Kan, K. Ding, S. Chen, H. Cheng, Z. Yao, H. Bingkun, R. Chen, D. Ma, B. Qiang, Y. Wen, Y. Hou, and J. Yu. 2002. Genome sequence of Shigella flexneri 2a: insights into pathogenicity through comparison with genomes of Escherichia coli K12 and O157. Nucleic Acids Res. 30: 4432-4441. June, G.A., P.S. Sherrod, R.M. Amaguana W.A. Andrews, and T.S. Hammack. 1993. Evaluation of the Bacteriological Analytical Manual culture method for the recovery of Shigella sonnei from selected foods. J. AOAC Int. 76: 1240-1248. Kakinuma, K., M. Fukushima, and R. Kawa guchi. 2003. Detection and identification of Escherichia coli Shigella and Salmonella by microarrays using the gyrB gene. Biotech. Bioeng. 83: 721-728. Kapperud, G., L.M. Rorvik, V. Hasseltvedt, E. A. Hoiby, B.G. Iverson, K. Staveland, G. Johnson, J. Leitao, H. Herikstad, Y. Ande rsson, Y. Langeland, B. Gondrosen, and J. Lassen. 1995. Outbreak of Shigella sonnei infection traced to imported iceberg lettuce. J. Clin. Microbiol. 33: 609-614. Kenne, L., B. Lindberg, K. Petersson, and E. Romanowska. 1977. Basic structure of the oligosaccharide rep eating subunit of the Shigella flexneri O-antigens. Carbohydr. Res. 56: 363-370. Kimura, A.C., K. Johnson, M.S. Palumbo, J. Hopkins, J.C. Boase, R. Reporter, M. Goldoft, K.R. Stefonek, J.A. Farrar, T.J. Van Gilder, and D.J. Vugia. 2004. Multistate shigellosis outbreak and comm ercially prepared food, United States. Emerg. Infect. Dis. 10: 1147-1149. Kocka, F.E., F.D. Dorigan, Q.T. Abbasi, A.L. Swiatlo, and M. Hubbard-Shepard. 1992. Clinical evaluations of the Wellcolex Colour Shigella and Salmonella tests. Diagn. Microbiol. Infect. Dis. 15: 1-4. Kothary, M.H. and U.S. Babu. 2001. Infec tive dose of foodborne pathogens in volunteers: a review. J. Food Safety. 21: 49-73.

PAGE 162

149 Kramer, M.F. and D.M. Coen. 1994. The polymerase chain reaction. In: Current protocols in molecular biology pp15.11-15-19. Ausbel, F.M., Brent, R. and Kingston, R.E., Eds., Wiley-Interscience. New York, NY. Lampel, K.A. 2001. Shigella In: Compendium of methods for the microbiological examination of foods, 4th ed. pp. 381-385. Downes, F.P. and Ito, K., Eds., American Public Health Association. Washington, D.C. Lampel, K.A. and A.T. Maurelli. 2001. Shigella species. In: Food microbiology: fundamentals and frontiers 2nd ed pp. 216-227. Doyle, M.P., Beuchat, L.R., and Montville T.J., Eds., American Society for Microbiology Press, Washington, D.C. Lampel, K.A. and P.A. Orlandi. 2002. Polymera se chain reaction detection of invasive Shigella and Salmonella enterica in food. Meth. Mol. Biol. 179: 235-244. Lampel, K.A., J.A. Jagow, M. Trucksess, a nd W.E. Hill. 1990. Polymerase chain reaction for detection of invasive Shigella flexneri in food. Appl. Environ. Microbiol. 56: 1536-1540. Lampel, K.A., P.A. Orlandi, and L. Korneg ay. 2000. Improved template preparation for PCR-based assays for the detection of food-borne bacterial pathogens. Appl. Environ. Microbiol. 66: 4539-4542. Lee, L.A., S.M. Ostroff, H.B. McGee, D. R. Johnson, F.P. Downes, D.N. Cameron, N.H. Bean, and P.M. Griffin. 1991. An outbreak of shigellosis at an outdoor music festival. Am. J. Epidemiol. 133: 608-615. Lefebvre, J., F. Gosselin, J. Ismal, M. Lorange, H. Lior, and D. Woodward. 1995. Evaluation of commercial antisera for Shigella serogrouping. J. Clin. Microbiol. 33: 1997-2001. Lew, J.F., D.L. Swerdlow, M.E. Dance, P.M. Griffin, C.A. Bopp, M.J. Gillenwater, T. Mercantante, and R.I. Glass. 1991. An outbreak of shigellosis aboard a cruise ship caused by a multiple-antibiotic-resistant strain of Shigella flexneri American J. Epidemiol. 134: 413-420. Lindqvist, R. 1999. Detection of Shigella spp. in food with a nested PCR method – sensitivity and performance compared with a conventional culture method. J. Appl. Microbiol. 86: 971-978. Lising, M., K. Reese, D. Walton, A. Shields, T. Bittner, R. Schrock, and M. Healy. 2004. Identification of Gram-negative bacteria using the DiversiLab system. [poster]. 104th General Meeting, Am. Soc. Microbiol. March 23-27; New Orleans, LA. Available at: http://www.bacbarcodes.com/posters/ASM_2004_Enteric_Poster.pdf Accessed 05 Nov 2004.

PAGE 163

150 Long, S.M., G.K. Adak, S.J. O’Brien, and I.A. Gillespie. 2002. General outbreaks of infectious intestinal disease linked with salad vegetables and fruit, England and Wales, 1992-2000. Common Dis. Pu blic Health. 5: 101-105. Lynch, M.J.B., C.G. Leon-Velarde, S. Mc Ewen, and J.A. Odumeru. 2004. Evaluation of an automated immunomagnetic separati on method for the rapid detection of Salmonella species in poultry environmenta l samples. J. Microbiol. Meth. 58: 285288. Maher, N., H.K. Dillon, S.H. Vermund, a nd T.R. Unnasch. 2001. Magnetic beads capture eliminates PCR inhibitors in samples collected from the airborne environment, permitting detection of Pneumocystis carinii DNA. Appl. Environ. Microbiol. 67: 449-452. Mangiapan, G., M. Vokurka, L. Schouls, J. Cadranel, D. Lecossier, J. van Embden and A.J. Hance. 1996. Sequence-capture PCR im proves detection of Mycobacterial DNA in clinical samples. J. Clin. Microbiol. 34: 1209-1215. Martin, D.L., T.L. Gustafson, K.W. Pelo si, L. Suarez, and G.V. Pierce. 1986. Contaminated produce -a common source for two outbreaks of Shigella gastroenteritis. Am. J. Epidemiol. 124: 299-305. Mehlman, I.J., A. Romero, and B.A. Wentz. 1985. Improved enrichment for recovery of Shigella sonnei from foods. J. AOAC. 68: 552-555. Melito, P.L., D.L. Woodward, J. Munro, J. Wa lsh, R. Foster, P. Tilley, A. Paccagnella, J. Issac-Renton, J. Ismail, and L.K. Ng. 2005. A novel Shigella dysenteriae serovar isolated in Canada. J. Clin. Microbiol. 43: 740-744. Mensah, P., D. Yeboah-Manu, K. Owusu-Darko, and A. Ablordey. 2002. Street foods in Accra, Ghana: how safe are they? B. World Health Organ. 80: 546-554. Monack, D.M. and J.A. Theriot. 2001. Actin-based motility is sufficient for bacterial membrane protrusion formation and host cell uptake. Cell. Microbiol. 3: 633-647. Nakamura, M. 1962. The survival of Shigella sonnei on cotton, glass, wool, paper and metal at various temperatures. J. Hyg. 60: 35-39. Neidhardt, F.C. 2004. Bact erial structures. In: Sherris Medical Microbiology, 4th ed pp. 11-25. Ryan, K.J., and Ray C.G., Eds. McGraw-Hill, New York. Ogawa, M. and C. Sasakawa. 2006. Intracellular survival of Shigella Cell Microbiol. 8: 177-184. Ogawa, M., T. Suzuki, I. Tatsuno, H. Abe, and C. Sasakawa. 2003. IcsB, secreted via the type III secretion system, is chaperoned by IpgA and required at the post-invasion stage of Shigella pathogenicity. Mol Mi crobiol. 48: 913-931.

PAGE 164

151 Ogawa, M., T. Yoshimori, T. Suzuki, H. Sagara, N. Mizushima, and C. Sasakawa. 2005. Escape of intracellular Shigella from autophagy. Science. 307: 727-731. Olsvik, ., T. Popovic, E. Skjerve, K.S. C udjoe, E. Hornes, J. Ugelstad, and M. Uhln. 1994. Magnetic separation techniques in di agnostic microbiology. Clin. Microbiol. Rev. 7: 43-54. Orlandi, P.A. and K.A. Lampel. 2000. Extracti on-free, filter based template preparation for rapid and sensitive PCR detection of pathogenic parasitic protozoa. J. Clin. Microbiol. 38: 2271-2277. Parsot, C. and P.J. Sansonetti. 1996. Invasion and the pathogenesis of Shigella infections. Curr Topics Microbiol. Immunol. 209: 25-42. Peng, X., W. Luo, J. Zhang, S. Wang, and S. Lin. 2002. Rapid detection of Shigella species in environmental sewage by an immunocapture PCR with universal primers. Appl. Environ. Microbiol. 68: 2580-2583. [PHLS] Public Health Laboratory Servi ce. 1997. Outbreaks of foodborne illness in humans, England and Wales: Quar terly Report. CDR Wkly. 7: 207-216. Pupo, G.M., R. Lan, and P.R. Reeves. 2000. Multiple independent origins of Shigella clones of Escherichia coli and convergent evolu tion of many of their characteristics. Proc. Natl. Acad. Sci. USA. 97: 10567-10572. Rafii, F. and P. Lunsford. 1997. Survival and detection of Shigella flexneri in vegetables and commercially prepared sa lads. J. AOAC Int. 80: 1191-1197. Rahman, S.R. and W.H. Stimson. 2001. Charact erization of monocl onal antibodies with specificity for the core oligosaccharide of Shigella lipopolysaccharide. Hybrodoma. 20: 85-90. Raza, S., J. Huong, T. Bittner, and M. Healy. 2003. Identification and strain differentiation of food and waterborne bact erial threat agents using automated repPCR. Poster presented at BioDefense Research Meeting, American Society of Microbiology. March 9-12; Ba ltimore, MD. Available at: http://www.bacbarcodes.com/posters /BBCI_Poster_2003_Biodefence_Poster_03_1 1_03_SR_FINAL.pdf Accessed 05 Nov 2004. Reeve, G., D.L. Martin, J. Pappas, R.E. Thompson, and K.D. Greene. 1989. An outbreak of shigellosis associated with the cons umption of raw oysters. N. Engl. J. Med. 321: 224-227. Rolland, K., N. Lambert-Zechovsky, B. Picard, and E. Denamur. 1998. Shigella and enteroinvasive Escherichia coli strains are derived from dist inct ancestral strains of E. coli. Microbiol. 144: 2667-2672.

PAGE 165

152 Sansonetti, P.J., D.J. Ko pecko, and S.B. Formal. 1981. Shigella sonnei plasmids: Evidence that a large plasmid is necessary for virulence. Infect. Immun. 34: 75-83. Sapsford, K.E., A. Rasooly, C.R. Tait t, and F.S. Ligler. 2004. Detection of Campylobacter and Shigella species in food samples us ing an array biosensor. Anal. Chem. 76: 433-440. Satchell, F.B., P. Stephenson, W.H. Andrews, L. Estela, and G. Allen. 1990. The survival of Shigella sonnei in shredded cabbage. J. Food Prot. 53: 558-562, 624. Sethabutr, O., M. Venkatesan, G.S. Mur phy, B. Eampokalap, C.W. Hoge, and P. Echeverria. 1993. Detection of Sh igellae and enteroinvasive Escherichia coli by amplification of the invasion plasmid antigen H DNA sequence in patients with dysentery. J. Infect. Dis. 167: 458-461. Sethabutr, O., M. Venkatesan, S. Yam, L.W. Pang, B.L. Smoak, W.K. Sang, P. Echeverria, D.N. Taylor, and D.W. Isenba rger. 2000. Detection of PCR products of the ipaH gene from Shigella and enteroinvasive Escherichia coli by enzyme linked immunosorbent assay. Diagn. Micr obiol. Infect. Dis. 37: 11-16. Shchepinov, M.S., S.C. Case-Green, and E.M. Southern. 1997. Steric factors influencing hybridization of nucleic acids to oligonucleotide arrays Nucleic Acids Res. 25: 1155-1161. Shiferaw, B., S. Shallow, R. Marcus, S. Se gler, D. Soderlund, F.P. Hardnett, and T.V. Gilder, for the Emerging Infections Program FoodNet Working Group. 2004. Trends in population based active survei llance for shigellosis and demographic variability in FoodNet s ites, 1996-1999. Clin. Infect. Dis. 38(Suppl 3): S175-S180. Sigma-Genosys website. 2006. DNA calculator. Available at: http://www.sigmagenosys.com/calc/DNACalc.asp Accessed 20 Jun 2006. Smith, J.L. 1987. Shigella as a foodborne pathogen. J. Food Prot. 50: 788-801. Song, T., C. Toma, N. Nakasone, and M. Iw anaga. 2005. Sensitive and rapid detection of Shigella and enteroinvasive Escherichia coli by a loop-mediated isothermal amplification method. FEMS Mi crobiol. Lett. 243: 259-263. Sonjai, K., R. Soisangwan, Y. Sakolva ree, H. Kurazono, M. Chongsa-nguan, P. Tapchaisri, Y. Mahakunkijcharoen, G.B. Na ir, H. Hayashi, and W. Chaicumpa. 2001. Validation of salmonellosis and shigellosis diagnostic te st kits at a provincial hospital in Thailand. Asian Pac. J. Allergy Immunol. 19: 115-127. Spicer, C.C. 1959. The survival of Shigella sonnei on cotton threads. J. Hyg. 57: 210-215. Stevens, K.A. and L.A. Jaykus. 2004. Bact erial separation and concentration from complex sample matrices: a review Crit. Rev. Microbiol. 30: 7-24.

PAGE 166

153 Suzuki, T. and C. Sasakawa. 2001. Molecular basis of the intrace llular spreading of Shigella Infect. Immun. 69: 5959-5966. Taylor, W.I. and D. Schelhart. 1969. Isolati on of Shigellae VII. Comparison of gramnegative broth with Rappaport’s enrich ment broth. Appl. Microbiol. 18: 393-395. Terajima, J., K. Tamura, K. Hirose, H. Iz umiya, M. Miyahara, H. Konuma, and H. Watanabe. 2004. A multi-prefectural outbreak of Shigella sonnei infections associated with eating oysters in Japan. Microbiol. Immunol. 48: 49-52. Theron, J., D. Morar, M. Du Preez, V.S. Brzel, and S.N. Venter. 2001. A sensitive seminested PCR method for the detection of Shigella in spiked environmental water samples. Wat. Res. 35: 869-874. Tollison, S.B. and M.G. Johnson. 1985. Sensitivity to bile salts of Shigella flexneri sublethally heat stressed in buffer or broth. Appl. Envir on. Microbiol. 50: 337-341. [TPH] Toronto Public Health. 2002. Health advisory: shigellosis food poisoning investigation. Available at: http://www.city.toronto.on.ca/he alth/advisory_shigella.htm Accessed 05 Nov 2004. Toyotome, T., T. Suzuki, A. Kuwae, T. Nonaka, H. Fukuda, S. Imajoh-Ohmi, T. Toyofuku, M. Hori, and C. Sasakawa. 2001. Shigella protein IpaH(9.8) is secreted from bacteria within mammalian cells a nd transported to the nucleus. J. Biol. Chem. 276: 32071-32079. Trevejo, R.T., S.L. Abbott, M.I. Wolfe, J. Meshulam, D. Yong, a nd G.R. Flores. 1999. An untypeable Shigella flexneri strain associated with an outbreak in California. J. Clin. Microbiol. 37: 2352-2353. Tsai, Y.-L., J.Y. Le and B.H. Olson. 2003. Magnetic bead hybridization to detect enterotoxigenic Escherichia coli strains associated with cattle in environmental water sources. Can. J. Microbiol. 49: 391-398. Uyttendaele, M., C.F. Bagamboula, E. De Sm et, S. Van Wilder, and J. Debevere. 2001. Evaluation of culture media for enrichment and isolation of Shigella sonnei and S. flexneri Int. J. Food Mi crobiol. 70: 255-265. Vantarakis, A., G. Komninou, D. Venier i, and M. Papapetropoulou. 2000. Development of a multiplex PCR detection of Salmonella spp. and Shigella spp. in mussels. Lett. Appl. Microbiol. 31: 105-109. Villalobo, E. and A. Torres. 1998. PCR for the detection of Shigella spp. in mayonnaise. Appl. Environ. Microbiol. 64: 1242-1245.

PAGE 167

154 Viret, J.-F., U. Bruderer, and A. B. Lang. 1992. Characterization of the Shigella serotype D ( S. sonnei ) O polysaccharide and the enterobact erial R1 lipopolysaccharide core by use of mouse monoclonal antibod ies. Infect. Immun. 60: 2741-2747. Warren, B.R. 2003. Comparison of conventiona l culture methods and polymerase chain reaction for the detection of Shigella spp. on tomato surfaces. Masters thesis. University of Florida, Gainesville, Fl. Warren, B.R., M.E. Parish, and K.R. Schne ider. 2005a. Comparison of chromogenic Shigella spp. plating medium with sta ndard media for the recovery of Shigella boydii and Shigella sonnei from tomato surfaces. J. Food Prot. 68: 621-624. Warren, B.R., M.E. Parish, and K.R. Sc hneider. 2005b. Comparison of conventional culture methods and FTA filtration-nested PCR for the detection of Shigella spp. on tomato surfaces. J. Food Prot. 68: 1606-1612. Warren, B.R., M.E. Parish, and K.R. Schneider. 2006. Shigella as a foodborne pathogen and current methods for detection. Crit. Rev. Food. Sci. Nutr. Accepted and in press. Wessel, D. and U.I. Flgge. 1984. A method for the quantitative recovery of protein in dilute solution in the presence of deterg ents and lipids. Anal Biochem. 138: 131133. Whatman website. BioScience FAQs. Available at: http://www.whatman.com/support/?pageID=3.6.124.112#fa112 Accessed 12 Nov 2004. Wood, L.V., L.E. Ferguson, P. Hogan, D. Thurman, D.R. Morgan, H.L. DuPont, and C.D. Ericsson. 1983. Incidence of bacterial enteropathogens in foods from Mexico. Appl. Environ. Microbiol. 46: 328-332. Wu, F.M., M.P. Doyle, L.R. Beuchat, J.G. Wells, E.D. Mintz, and B. Swaminathan. 2000. Fate of Shigella sonnei on parsley and methods of disinfection. J. Food Prot. 63: 568-572. Yagupsky, P., M. Loeffelholz, K. Bell, and M.A. Menegus. 1991. Use of multiple markers for investigation of an epidemic of Shigella sonnei infections in Monroe County, New York. J. Clin. Microbiol. 29: 2850-2855. Yang, F., J. Yang, X. Zhang, L. Chen, Y. Ji ang, Y. Yan, X. Tang, J. Wang, Z. Xiong, J. Dong, Y. Xue, Y. Zhu, X. Xu, L. Sun, S. Chen, H. Nie, J. Peng, J. Xu, Y. Wang, Z. Yuan, Y. Wen, Z. Yao, Y. Shen, B. Qiang, Y. Hou, J. Yu, and Q. Jin. 2005. Genome dynamics and diversity of Shigella species, the etiological agents of bacillary dysentery. Nucleic Acids Res. 33: 6445-6458.

PAGE 168

155 Yuk, H.G., B.R. Warren, and K.R. Schneider 2006. Preliminary evaluation of flowthrough immunocapture real-time PCR for the detection of Salmonella on tomato surfaces within eight hours. J. Food Prot. Accepted and in press. Zaika, L. 2001. The effect of temp erature and low pH on survival of Shigella flexneri in broth. J. Food Prot. 64: 1162-1165. Zaika, L. 2002a. The effect of NaCl on survival of Shigella flexneri in broth as affected by temperature and pH. J. Food Prot. 65: 774-779. Zaika, L. 2002b. Effect of organic ac ids and temperature on survival of Shigella flexneri in broth at pH 4. J. Food Prot. 65: 1417-1421. Zammatteo, N., I. Alexandre, I. Ernest, L. Le, F. Brancart, and J. Remacle. 1997. Comparison between microwell and bead supports for the detection of human cytomegalovirus amplicons by sandwich hybridization. Anal. Biochem. 253: 180189. Zhuang, R.-H., L.R. Beuchat, and F.J. Angulo. 1995. Fate of Salmonella Montevideo on and in raw tomatoes as affected by temper ature and treatment w ith chlorine. Appl. Environ. Microbiol. 61:2127-31.

PAGE 169

156 BIOGRAPHICAL SKETCH Benjamin Ray Warren was born in Brandon, FL, on June 29, 1975. In 1998, he received his Bachelor of Science from the Un iversity of Florida in food science. After graduation he took the position of Food Safety and Product Development Manager with Blood’s Hammock Groves, Inc., a grower/shipp er/processor of fresh Florida citrus. In 2003, he received his Master of Science from the University of Fl orida in food science where his thesis won the IFAS Award of Ex cellence for Graduate Research. The author worked part-time at Deibel Laboratories of Gainesville, Inc. during his master’s and doctoral research. On May 23, 2004, the author married Nicole Leigh Sanson at the Baughman Center on the University of Florida campus. On Ma y 2, 2006, the author and his wife celebrated the birth of their first son, Zachary Ray Warren. Upon completion of his doctoral degree, the author plans to pursue a career in research and academics in food microbiology.


Permanent Link: http://ufdc.ufl.edu/UFE0015665/00001

Material Information

Title: Improved Sample Preparation for the Molecular Detection of Shigella sonnei in Foods
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015665:00001

Permanent Link: http://ufdc.ufl.edu/UFE0015665/00001

Material Information

Title: Improved Sample Preparation for the Molecular Detection of Shigella sonnei in Foods
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015665:00001


This item has the following downloads:


Full Text












IMPROVED SAMPLE PREPARATION FOR THE MOLECULAR DETECTION OF
,nge/ll/ sonnei IN FOODS
















By

BENJAMIN RAY WARREN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Benjamin Ray Warren


























To Nikki, for your undying love and support; to Zachary, for bringing so much love into
our lives; to my parents, for never losing faith in me; and to all my friends along the way,
for without all of you this would not have been possible.















ACKNOWLEDGMENTS

First and foremost, I would like to thank my committee chair, Dr. Keith Schneider,

for all of his support and guidance over my graduate studies. Additionally, I would like to

thank Dr. Mickey Parish for seeing my potential and inspiring me to return to the

University of Florida to pursue my Ph.D., and Dr. Schneider for always having an open

door and encouraging my questions and ideas. I would further like to thank the remaining

members of my graduate committee, Dr. Douglas Archer, Dr. Eric Triplett and Dr. Keith

Lampel, for all of their assistance with this project; each of them brought a unique

perspective to this project and provided excellent support.

Furthermore, I would like to thank my wife, Nicole, whose hard work and sacrifice

during my graduate studies has made all of this possible; she will always have my

admiration and love. I would also like to thank my parents, Dennis and Linda Warren, for

their never-ending love and support not only in recent years but throughout my life; I

thank them for setting such a good examples.

Statistical assistance was provided by University of Florida, IFAS Statistics, with

special thanks to Meghan Brennan. This project was funded in part by the USDA-

CSREES IFAFS Grant number 00-52102-9637.
















TABLE OF CONTENTS


A C K N O W L E D G M E N T S ............................................. .............................................. iv

LIST OF TABLES .............. .......... .. ....... ........... ....... ix

LIST OF FIGURES ......... ........................................... ............ xi

ABSTRACT .............. ..................... .......... .............. xii

CHAPTER

1 INTRODUCTION ............... ................. ........... ................. ... ..... 1

2 LITER A TU RE REV IEW .................................................. ............................... 6

,/lge/llt as a Foodborne Pathogen ..................................................................... .... 6
E pidem iology of / ge// l ................... ... ..... .....................................................7
Demographic Variability of Infections with .l/nge/lt spp...............................8
Recently Identified Serotype of S. dysenteriae .............................................9
Foodborne Outbreaks Involving Shigella................. ........................... 9
Prevalence of.\//lge/l t Food and Food Handlers ...............................................11
Survival of ./ng / .............. .... .. ............. ................................ .......... .... 12
Environmental Factors on Survival of Shigella............................................12
Survival of.\h/ge//At on Fom ites ......................................................... ....... 14
Survival of.\,/ge//At in Food and W ater ............. .............................................. 15
Current Understanding of.\h/ge//At Pathogenesis ..................................................17
\/nlge/ll Invasion of Epithelial Cells .......................................................18
Potential Roles of the IpaH Effector Proteins ..............................................18
IpaH7.8 facilitates escape from endocytic vacuoles ...................................18
Subversion of host cell signaling by IpaH9.8 ............................................19
Blockage of Autophagy by IcsB ..................... ............... ............... ... 20
Genetic Relationship Between .\l/gell,// and Escherichia coli............................... 21
Detection Methods for .l/ge/ in Foods ..................................... ..................23
Conventional Culture Methods for ./nge//ll ............... .. ..................23
Traditional microbiological media for enrichment and isolation of
lgl. l,. ........................ ................. ................ ................ 23
The FDA Bacteriological Analytical Manual culture method for
detection of.\h/ge//At in foods ............................................................ 27
Other culture methods for the detection of \s/ge//At in foods ...................28









Immunological M ethods for .\/nge//l t Detection .............. ..............................29
Differences in S. sonnei form I and form II lipopolysaccharide ..................30
Immunological Detection Methods for Bacteria............. ...... ............... 31
Latex agglutination methods for .l/tigel/ .......................................... 33
Enzyme immunoassay methods for lge/ ............................... .......... 35
Immunomagnetic separation methods for \/nge//lt detection ....................36
Immunomagnetic separation using the Pathatrix .............................. 37
Additional technologies for immunological detection of Shigella ............38
Molecular Microbiological Methods for .\llgel/t Detection in Foods ..............39
Polymerase chain reaction detection of .\/hlge//t in foods..........................39
Improved PCR detection of.\/lge/lt by FTA filtration..............................41
Additional molecular microbiological techniques for detection of
S ...................... ..................................................... ............ 4 3
Detection of Pathogens by Sequence Capture...........................................................46

3 M ATERIALS AND M ETHODS ........................................ .......................... 50

Preliminary Studies......................... ....................................... ... 50
Preparation of Microbiological Media ................................ ................51
Acquisition and Maintenance of \/nge//lt sonnei Cultures.......................... 50
Adaptation of Cultures to Rifampicin ...................................... ............... 50
Acquisition/Preparation of Food M atrices ........................................................51
Acquisition and Maintenance of Anti-.\/lgel//t Antibodies..............................52
Binding of Antibodies to Paramagnetic Beads..................................................53
Evaluation and Optimization of Immunocapture Using Anti-.\/Nlge//A, Beads ....54
Crude DNA Extraction from Bacteria by Boiling.........................................55
DNA Extraction from Anti-.\/lgel//l Beads using the DNeasy Kit...................56
Preparation of HeLa Cell Extracts .............................................. ...............56
RN A Extraction U sing the RN easy K it ...............................................................57
DNase Treatment of RNA Extracts Prior to RT-PCR ........................................58
Induction and Expression of ipaH RNA in S. sonnei .........................................59
Identification of\/nlge//l -Specific Genetic Loci ............................................59
Development of Primers/Probes for the Detection of/lg//........................60
Evaluation of Primer/Probe Specificity................... ............... ..................60
Binding of Biotinylated Capture Probes to Streptavidin-Coated Paramagnetic
B e a d s .......................................................................................................... 6 4
Inoculum Preparation ..................... .................... ............... ... 65
Calculation of Generation Time of S. sonnei in ,'1/rlge//t Broth..........................65
Preliminary Experiments with Anti-.\/egel//t Beads .........................................66
Separation of S. sonnei from Food Matrices Using Low-Speed Centrifugation.66
Survival Studies ....................... .... ... .. .................. ....... .......... 67
Sample Inoculation and Subsequent Recovery ................................................67
Three-Tube Most Probable Number Estimation of Survivors ............................68
Evaluation of Detection Methods ...................... ........................................ 68
Inoculation of Samples and Subsequent Recovery ..........................................70
Modified BAM Culture Method for S. sonnei ................. ............................70
Flow-Through Immunocapture (FTI) Using the Pathatrix..............................71









Sequence Capture of ./lge// l DN A ...................................................................72
Real-Time PCR and Reverse Transcriptase (RT) PCR ..................................... 74
Recording of Data and Statistical Analysis ......... .... ............... .................. 75

4 R E S U L T S .............................................................................7 7

P relim in ary Stu dies................................... ................... .. .. ............ .... ......... 7 7
Calculation of Generation Time of S. sonnei in ,\l///al t Broth..........................77
Growth curve of S. sonnei ATCC 9290 ....................................... .......... 78
Growth curve of S. sonnei ATCC 29031 ......................................... 78
Growth curve of S. sonnei ATCC 29030 ............................................. 78
Growth curve of S. sonnei ATCC 25931 ......................................... 80
Growth curve of S. sonnei ATCC 29930 ..................................................80
Evaluation Anti-.\/lgel//l Antibodies for Use with Flow-Through
Im m unocapture ........................ .................................. ......... .. ...... .. 82
Preliminary Experiments with Anti-.\/lgel// t Beads ...........................................84
Optimization of Anti-.\/lgel//l Bead Concentration for Flow-Through
Im m unocapture of S. sonnei ............................................................ .... 85
Identification of Potentially .\/lge/l, t-Specific Genetic Loci ..............................87
Specificity of Primers Developed for Potentially \,/lge/l, t-Specific Genetic
L o c i ..................... ........ ........ .... ... .. ......................................8 7
Separation of S. sonnei from Food Matrices by Low-Speed Centrifugation ......90
Development of DNA Sequence Capture (DSC) for the Detection of S.
son n ei ............................... ... .. .... . ......... .. ........................ ....... 9 1
Expression of ipaH RNA in Log and Stationary Phase S. sonnei....................95
Survival Studies ..................................... ..... .... ........................96
Survival of S. sonnei on Smooth Tomato Surfaces..........................................96
Survival of S. sonnei in Potato Salad........................................ ............... 97
Survival of S. sonnei in Ground Beef....................................... ............... 99
Evaluation of D election M methods ............................................ ........................ .101
Detection of S. sonnei in Selected Foods by a Modified FDA Bacteriological
AnalyticalManual (BAM) .\/n/ge// Culture Method ...................................101
Detection of S. sonnei in Selected Foods by Flow-Through Immunocapture
(F T I) .......................................................................... ............... 1 0 2
Detection of S. sonnei in Selected Foods by DNA Sequence Capture (DSC) ..104

5 DISCUSSION AND CONCLUSIONS .............. .............................................106

P relim inary Studies................... .. .............. .................. ...... ............ .. .. ... .......106
Growth Characteristics of S. sonnei in .\lt/ge/ll Broth (SB) ...........................107
Expression and Induction of the ipaH Gene of S. sonnei...............................108
Identification of Potentially \/nlge/, t-Specific Genetic Loci...........................109
Survival Studies ......................... ...................... ...... .......... ..........110
Rapid S. sonnei Inactivation on Tomato Surfaces............... ............111
S. sonnei Survives in Potato Salad and Ground Beef ............. .... ............ ..112
Evaluation of Detection M ethods ................................ ............ ............... 112
Recovery of S. sonnei by the BAM .l/n/ge,// Culture Method ..........................112









Analysis of lowest detection levels of the BAM ,/ngel//At culture method. 113
Evaluation of Flow-Through Immunocapture for the Detection of S. sonnei in
F ood ............................................. .... ...... ..... .... .. .. ............... 115
Operational issues with anti-./hige//l beads in flow-through
im m unocapture .................................................... .... .......................115
Non-specific immunocapture of Enterobacter cloacae and Escherichia
c o li ..................................... .. .... .................. ....................... ....... 1 1 7
Analysis of lowest detection levels of the FTI-MAC and FTI-PCR
m ethods ................ .......... ........ .......................... ... ....... 119
Future research involving flow-through immunocapture for the detection
of S. sonnei ....................... ........................................................ .... 122
Evaluation of DNA Sequence Capture (DSC) for the Detection of S. sonnei in
F o o d ......................................................................... .. 1 2 3
Evaluation of hybridization buffers for DSC ............................................123
Non-specific adsorption of DNA to CP-.\/ige//At beads............................. 124
Analysis of lowest detection levels of the DSC method............................126
Future research to improve DSC for the detection of S. sonnei in food ....128
Sources of Variation Among Inoculated Studies .......... .............. .............129
Practical Applications of Flow-Through Immunocapture and DNA Sequence
C capture ............................................................................................. ........ 130
C o n c lu sio n s......................................................................................................... 1 3 1

APPENDIX

A PREPARATION OF BUFFERS AND SOLUTIONS ..........................................132

Binding/W ashing (B/W ) Buffer (2X).................................................................... 132
C ongo R ed Solution ........... .............................................................. ........ .. ....... .. 132
D ynabeads Solution A ............. ................................................. ........ 133
D ynabeads Solution B .......................................... ........................ 133
H ybridization B uffer 1..................... ........................ .. .. ...... ...............133
H ybridization B uffer 2 ......... ................. ................. .................... ............... 134
L ow -Salt W ash B uffer........... ........................................................ ............... 135
Phosphate Buffered Saline, pH 7.4............................................... ........ ....... 135
Sodium A cetate Buffer, pH 4.0 ........................................ .......................... 136
Sodium Bicarbonate Buffer, pH 8.6 ........................................ ...... ............... 136
W a sh B u ffe r ...................................... .............................................. 13 6

B ALIGNMENT OF CHROMOSOMALLY-LOCATED ipaH GENES OF ,/ngel/ht
son n ei ............................................................. .... ........... ..... 13 8

LIST OF REFEREN CES ........................................... ........................ ............... 143

BIOGRAPHICAL SKETCH ............... ................. ............... 156
















LIST OF TABLES


Table p

2-1. Percentage of ,/Nige//t isolates in the United States reported by PHLIS in recent
years .......... .................................. ..................

2-2. Selected foodborne outbreaks involving .\/ngel// .......... ....................................10

3-1. Antibodies investigated for immunocapture of S. sonnei .......................................54

3-2. Primers designed for the detection of. /ge//t ..................................................61

3-3. \/nge//At and non-./,'Nge/l a strains tested for specificity................. ...............62

4-1. Evaluation of anti-.\/Nge//A, antibodies for flow-through immunocapture (FTI) of
S. sonn ei. ............................................................................83

4-2. Optimization of anti-.\/lgel//l bead concentration for flow-through
im munocapture (FTI) of S. sonnei.. ................................................................... 86

4-3. Genetic targets identified with potential specificity for ./nge//lt spp. or for S.
sonnei alone ................................................................. ..... .........88

4-4. Evaluation of primer specificity among stock \,/ilge/ll cultures and by
comparative analysis against previously sequenced \/ gel//t genomes..................89

4-5. Effects of low-speed centrifugation on S. sonnei populations in sample
su p e rn a ta n t ...............................................................................................................9 1

4-6. Evaluation of hybridization buffers for DSC for the detection of S. sonnei ..............92

4-7. Comparison of paramagnetic beads for use with CP-.\/igel//t beads.......................94

4-8. Specific capture of ipaH DNA in the presence of non-target DNA by CP-.\/'igel'//
b ead s ............................. .............................. ...... .......................... 9 5

4-9. Sensitivity of DNA sequence capture method................................. ............... 96

4-10. Transcriptional induction of the ipaH gene using HeLa cell extracts and the dye
Congo red ............... ......... ..................... ............. ............ 97

4-11. Number of samples positive for S. sonnei by various detection methods..............102









5-1. Lowest detection levels of the BAM ,Nl//nge//l, culture method............................. 114

5-2. Lowest detection levels of the FTI-MAC and FTI-PCR methods............................120

5-3. Lowest detection levels of the D SC method ............................... ........1.........127
















LIST OF FIGURES


Figure page

2-1. Invasion of epithelial cells by \ltgell'//t spp.. ................................... ............... 19

2-2. Structural differences between hexose regions of ./ilge//At sonnei and \//ge,//t
flexneri lipopolysaccharides......................................................... ............. 30

2-3. O-antigen repeating subunits of S. sonnei and S. flexneri lipopolysaccharide...........31

2-4. Lipopolysaccharides of S. sonnei ............................ ..... ................................. 32

2-5. The Pathatrix system for flow-through immunomagnetic separation ......................38

3-1. Design of the capture probe and CP-.\/lge//At beads............................................65

3-2. Flow diagram of the experiments involving inoculated food samples.....................69

3-3. Flow diagram of the DNA sequence capture method.........................................73

4-1. Growth curve: S. sonnei ATCC 9290 in '1/lgel//A broth ..........................................79

4-2. Growth curve: S. sonnei ATCC 29031 in '1/lgel//A broth ........................................79

4-3. Growth curve: S. sonnei ATCC 29030 in .\/lrge/t broth ...........................................80

4-4. Growth curve: S. sonnei ATCC 25931 in .\/ige/l t broth ........................................ 81

4-5. Growth curve: S. sonnei ATCC 29930 in '1/lgel//A broth .......................................81

4-6. Survival of a five-strain S. sonnei cocktail on the smooth surfaces of tomatoes .......98

4-7. Survival of five-strain S. sonnei cocktail in potato salad ..................................99

4-8. Survival of five-strain S. sonnei cocktail in ground beef .............. ............... 100

5-1. Location of anti-.\/lgel//l beads in the capture phase of the Pathatrix during FTI...118















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 Philosophy

IMPROVED SAMPLE PREPARATION FOR THE MOLECULAR DETECTION OF
.1/ge/lt sonnei IN FOODS

By

Benjamin Ray Warren

August 2006

Chair: Keith R. Schneider
Major Department: Food Science and Human Nutrition

,1/Ngel/i, the causative agent of bacillary dysentery, was the third most reported

foodborne bacterial pathogen in the U.S. for 2005, and most of the isolates were

identified as S. sonnei. Methods for the detection of /ge//At in food, however, remain

problematic. In preliminary studies, a chromosomally-located genetic target for specific

detection of.h/ilge/ht RNA was investigated by comparing the available genomes of

\/, ge//At and E. coli. All DNA sequences identified with potential specificity for all

,/ngel/t spp. or with potential specificity for only S. sonnei using database searches

tested positive for E. coli strains that had been isolated from ground beef.

Additionally, the survival of a five-strain S. sonnei cocktail on tomato surfaces, in

potato salad and in ground beef was investigated using a most probable number (MPN)

method. Inoculated tomatoes were stored at 130C at 85% relative humidity, while potato

salad and ground beef samples were stored at 2.50C and 80C. On tomato surfaces, S.

sonnei populations declined to undetectable levels by day 3. In potato salad and ground









beef samples, S. sonnei populations detected at the end of the product shelf-life (28 days

and 11 days, respectively) were not significantly different (P > 0.05) than initial

populations. These studies suggest that S. sonnei survives well in foods when not

desiccated.

Flow-through immunocapture (FTI) followed by analysis of recovered anti-

/nlge//t beads by spread-plate using MAC (FTI-MAC), FTI followed by analysis of

recovered anti-.\/lgl, i/ beads by real-time PCR (FTI-PCR) and DNA sequence capture

(DSC) were compared to the .\lnge//t culture method of the FDA Bacteriological

AnalyticalManual (BAM) for the detection of S. sonnei on tomato surfaces, in potato

salad and in ground beef. FTI-MAC was significantly better (P > 0.05) than the BAM

.\l/ge/ll culture method for the analysis of tomatoes, but not potato salad or ground beef.

FTI-PCR and DSC were significantly better (P > 0.05) than the BAM .\/ge//At culture

method for the analysis of tomatoes and potato salad, but not ground beef.














CHAPTER 1
INTRODUCTION

The ability to analyze food products for the presence of pathogenic bacteria is

essential for verifying the safety of foods, identifying agents of foodborne illness and

determining sources of foodborne outbreaks. Conventionally, the microbiological

analysis of food involves culture enrichment followed by isolation on selective media.

Confirmation of presumptive isolates is generally through biochemical characterization

and/or serology. Conventional culture methods, however, are often problematic, in that

many are time-consuming and require several days to complete, appropriate selective

media are not currently available for all bacterial foodborne pathogens, some bacterial

pathogens require specific atmospheric or other growth conditions which may be difficult

to simulate in the laboratory and some bacterial pathogens may not be culturable by

currently available methods. In addition, most culture enrichment procedures used for the

detection of bacterial foodborne pathogens detect only the presence or absence of the

target pathogen. For enumeration, a most probable number (MPN) method must be

employed; however media required for MPN analysis for some bacterial foodborne

pathogens is not currently available.

For bacterial foodborne pathogens whose conventional culture methods are

problematic, alternative sample preparation methods may be used to improve sensitivity.

Alternative sample preparation methods provide a means of separating and concentrating

bacterial pathogens or components (proteins, nucleic acids, etc.) of bacterial pathogens

from food matrices. If molecular-based detection is to follow, the alternative sample









preparation method must also contend with any potential inhibitors which may be present

in the food. Unfortunately, there is no alternative sampling method suitable for the

separation and concentration of all types bacterial pathogens from all forms of food;

therefore each combination of sampling method, food matrix and bacterial pathogen must

be investigated independently.

Once bacterial pathogens are separated and concentrated from food, there are many

options for detection by rapid methods, such as immunoassays (enzyme linked

immunosorbent assay (ELISA) and lateral flow devices), DNA hybridizations or the

polymerase chain reaction (PCR). Most rapid methods require the presence of>103

colony forming units (CFU)/ml of the target bacterial pathogen for consistent and

dependable detection (Stevens and Jaykus, 2004); therefore efficient separation and

concentration of bacterial pathogens from food are critical. Because bacterial pathogens

may be present in food at low, yet potentially infectious populations, selective or non-

selective enrichment is often performed prior to sample analysis to increase the bacterial

population. For some sample preparation methods, very short enrichment times (4-5 hr)

may be sufficient to increase the bacterial populations to detectable levels, allowing the

assay to be performed in a same-day format (Yuk et al., 2006; Schneider and Warren,

unpublished data).

The detection of.\/i/ge/lt spp. in food is one example of where the use of

alternative sample preparation methods may be used to improve analysis over

conventional culture methods. The most commonly used .\lige//ll culture method in the

U.S. is found in the U.S. Food and Drug Administration's Bacteriological Analytical

Manual (BAM). The enrichment media recommended in the BAM is \l/gel//t broth (SB),









a low-carbohydrate medium used to limit the decrease in pH associated with acid

production from the microbiological metabolism of sugars. SB, however, does not

provide adequate specificity for .\/igel/h spp. and other members of the family

Enterobacteriaceae have been reported to out-compete and/or overgrow .\/ngel// spp.

during enrichment (Uyttendaele et al., 2001; Warren, 2003; Warren et al., 2005a). Other

enrichment media, such as Enterobacteriaceae Enrichment (EE) broth (Uyttendaele et

al., 2000) and Gram-negative (GN) broth (CMMEF), have been suggested for the

enrichment of .\igell//A spp.; however EE broth has been reported to be inhibitory to S.

boydii serotype 18 (Warren, 2003; Warren et al., 2005b) and GN broth contains bile salts

and sodium desoxycholate, which have been shown to inhibit stressed shigellae (Tollison

and Johnson, 1985; Uyttendaele et al., 2001).

Isolation media are also problematic for .\nge//At spp. The BAM recommends

MacConkey agar (MAC), which is selective for Gram-negative bacteria and differential

based on the utilization of lactose. Typically, .\ngel//t spp. are lactose negative; however

some serotypes of S. boydii have been reported as lactose positive. During the Gulf War

in the early 1990's, lactose positive S. sonnei were isolated from U.S. military personnel

(Dr. D.J. Kopecko, FDA, personal communication). Upon further investigation, it was

discovered that typical lactose negative S. sonnei mutate at high frequency to lactose

positive phenotypes during stationary phase on lactose-containing microbiological media

(Dr. D.J. Kopecko, FDA, personal communication). It has yet to be determined whether

lactose positive S. sonnei mutants can form in contaminated food products and whether

this mutation would allow S. sonnei a means to evade detection by conventional culture

methods, such as the BAM.









Previously, a filtration-based sample preparation method using FTA filters

(Whatman, Clifton, NJ) in combination with nested PCR was investigated for the

detection of S. sonnei and S. boydii on tomato surfaces (Warren, 2003; Warren et al.,

2005b). When a sample is applied to FTA filters, the moisture from the sample activates

chemical denaturants, chelating agent buffers and free radical traps embedded in the filter

which lyse cells, denature enzymes, inactivate pathogens and immobilize genomic DNA

(Whatman, 2006). Using a tandem filter funnel system in which the first filter funnel was

for size exclusion of sample material and the second filter funnel contained an FTA filter,

100 ml PBS rinses of tomatoes were analyzed and the captured \liigell/ DNA was

amplified by nested PCR. This FTA filtration-nested PCR assay was able to detect S.

sonnei and S. boydii on tomato surfaces at inoculation levels as low as 6.2-7.4

CFU/tomato. Unfortunately, the FTA filtration system was not as successful when

applied to other types of produce, such as strawberries, cantaloupes or retail Valencia

oranges. The analysis of these types of produce in the FTA filtration system resulted in

clogged filters and poor detection limits as compared to those performed on tomato

rinses.

Recently, a novel device for flow-through immunomagnetic separation (IMS), the

Pathatrix, has been developed (Matrix MicroScience, Inc., Golden, CO). IMS methods

involve the coupling of specific antibodies to paramagnetic beads, which exhibit

magnetic qualities only when placed in a magnetic field. In IMS methods, the antibodies

are used to capture target bacteria and then the bead-bacteria complexes are separated and

concentrated from the food matrix using a magnet. Traditional IMS methods analyze

small sample volumes (1.0 ml), whereas the Pathatrix allows the analysis of a much









larger sample volume (250 ml) by using a tubing system in which IMS beads are

immobilized in a capture phase and the sample is continuously pumped through the

capture phase using a peristaltic pump. Once recovered, the IMS beads can be analyzed

by spread-plate for the isolation of viable colonies or applied directly to DNA extraction

methods for subsequent analysis by PCR.

Finally, sample preparation methods based on the specific isolation of bacterial or

viral DNA/RNA from various types of matrices using oligonucleotide probes

immobilized on paramagnetic beads have been reported. Most commonly, biotinylated

probes are attached to paramagnetic beads pre-coated with streptavidin. The bead-probes

are then used to sample mixtures containing sample DNA/RNA in hybridization buffer.

Following hybridization, the captured DNA/RNA can be removed from the bead-probes

using heat and used as template in PCR, RT-PCR or other molecular-based detection

assay.

The hypothesis of this study was that specific antibodies and/or specific

oligonucleotide probes may be attached to paramagnetic beads and used for the detection

of S. sonnei with increased sensitivity over the current conventional culture methods. The

specific objectives of this study were as follows:

1. To investigate the survival of S. sonnei on the tomato surfaces, in potato salad and
in ground beef when held under standard refrigerated conditions.

2. To investigate chromosomally-located genetic targets for the specific detection of
all .\/nge/lt spp. or only S. sonnei.

3. To develop and compare flow-through immunocapture, using the Pathatrix, for the
detection of S. sonnei on the tomato surfaces, in potato salad and in ground beef
with the BAM .\l/igel// culture method.

4. To develop a nucleotide sequence capture method for the rapid detection of S.
sonnei on the tomato surfaces, in potato salad and in ground beef.














CHAPTER 2
LITERATURE REVIEW

.\hige/at as a Foodborne Pathogen

Background Information

.h/gell// spp. are the causative agents of shigellosis, or "bacillary dysentery," first

discovered over 100 years ago by Kiyoshi Shiga, a Japanese scientist (Anonymous,

2002). Shigellae are members of the bacterial family Enterobacteriaceae and are nearly

identical genetically to Escherichia coli and are also closely related to Salmonella and

Citrobacter spp. (Lampel, 2001). Shigellae are characterized as Gram-negative,

facultatively anaerobic, non-sporeforming, non-motile rods that typically do not ferment

lactose. In addition, they are lysine-decarboxylase, acetate, and mucate negative and do

not produce gas from glucose, although some S. flexneri six serotypes have been reported

to produce gas (Echeverria et al., 1991; International Commission on Microbiological

Specifications for Foods (ICMSF), 1996). There are four serogroups of.\//gel/h S.

dysenteriae (serogroup A; 15 serotypes), S. flexneri (serogroup B; eight serotypes divided

into 11 subserotypes), S. boydii (serogroup C; 20 serotypes), and S. sonnei (serogroup D;

one serotype) (Centers for Disease Control and Prevention (CDC), 2004).

.hige/ll has been classically characterized as a waterborne pathogen (Smith, 1987)

and outbreaks have been reported from contaminated community water sources that were

un- or under-chlorinated (Blostein, 1991; Fleming et al., 2000; CDC, 2001). Foodborne

outbreaks of.\/ige//At are also common, especially with foods that are subjected to

processing or preparation by hand, are exposed to a limited heat treatment, or are









served/delivered raw to the consumer (Wu et al., 2000). Some examples of food products

from which ./ngel//t spp. have been isolated are potato salad, ground beef, bean dip, raw

oysters, fish, and raw vegetables.

The infective dose for ./lge//At is very low: 10 cells of S. dysenteriae to 500 cells of

S. sonnei (Kothary and Babu, 2001). At-risk populations, such as the very young, very

old, or persons with decreased immune function, may be more susceptible to infection.

Due to the low infective dose of \/hige'/hi, person-to-person transmission is common,

especially in day-care settings (S. sonnei) where toddlers commonly practice poor

personal hygiene. Typical symptoms of infection include bloody diarrhea, abdominal

pain, fever, and malaise. Although the mechanism is unknown, seizures have been

reported in 5.4% of shigellosis cases involving children (Galanakis et al., 2002). Chronic

sequelae from S. dysenteriae serotype 1 infections can include hemolytic uremic

syndrome (HUS), while S. flexneri infections are associated with later development of

reactive arthritis, especially in persons with the genetic marker HLA-B27 (CDC, 2006a).

Reactive arthritis is characterized by joint pain, eye irritation, and painful urination

(CDC, 2006a).

Epidemiology of .\tige/all

The four serogroups of .\/igell// differ in epidemiology (Ingersoll et al., 2002). S.

dysenteriae is primarily associated with epidemics (Ingersoll et al., 2002) with serotype 1

associated with the highest fatality rate (5-15%) (CDC, 2003). S. flexneri predominates in

areas of endemic infection, while S. sonnei has been implicated in source outbreaks in

developed countries (Hale, 1991). S. boydii has been associated with source outbreaks in

Central and South America but is most commonly restricted to the Indian subcontinent. S.









boydii is rarely isolated in North America; however slight increases in the numbers of

isolates have been observed in both 2003 and 2004 (CDC, 2004).

According to the CDC Emerging Infections Program, Foodborne Diseases Active

Surveillance Network (FoodNet), \/nge//At was the third most reported foodborne

bacterial pathogen in 2005 (CDC, 2006b). Of 16,614 laboratory-diagnosed cases,

/nlge/ll accounted for 2,078 cases (12.5% of total cases) behind only Salmonella (6,471

cases) and Campylobacter (5,655 cases) (CDC, 2006b). From 1996-1998 to 2005 the

estimated annual incidence of ,/Nge//At spp. in the U.S. decreased by 43% (CDC, 2006b).

While S. sonnei continued to be the most isolated serogroup in the U.S. in 2004, the rate

of isolation has declined while the rate of isolation of S. flexneri and S. boydii have

increased slightly (Table 2-1).

Demographic Variability of Infections with .lnge,//At spp.

FoodNet data on shigellosis in the U.S. collected from 1996 to 1999 were recently

analyzed for trends in demographic variability (Shiferaw et al., 2004). The overall

incidence of shigellosis was highest among the following groups: children aged 1-4

years, male patients, blacks, Hispanics and Native Americans (Shiferaw et al., 2004).

There were also marked demographic differences between infection with S. sonnei and S.

flexneri with respect to age, sex and race. While the incidence of both S. sonnei and S.

flexneri were higher among those aged 1-4 years, there was a second peak of S. flexneri

infection among those aged 30-39 years (Shiferaw et al., 2004). The incidence of S.

sonnei among men and women were similar; however the incidence of S. flexneri among

men was almost twice that of women (Shiferaw et al., 2004). In addition, the incidence of

S. sonnei among blacks and whites was higher than that of S. flexneri, while the incidence

of S. flexneri was higher among Native Americans (Shiferaw et al., 2004).










Table 2-1. Percentage of .\/gellt isolates in the United States reported by PHLIS in
recent years.
Serogroup 2002 2003 2004
S. sonnei 83.5% 80.2% 68.9%
S. flexneri 12.2% 14.4% 17.2%
S. boydii 0.8% 1.1% 1.8%
S. dysenteriae 0.3% 0.4% 0.4%
Ungrouped 3.2% 3.9% 11.7%

Total isolates 12,992 11,552 9,343
(CDC, 2002; CDC, 2003; CDC, 2004)

Recently Identified Serotype of S. dvsenteriae

During 2001 to 2003, six biochemically, serologically and genetically identical

.\/ige/ll strains were isolated in geographically distant locations in Canada. When

analyzed biochemically, the suspect strains displayed reactions consistent with that of

.\/ige/ll spp. (Melito et al., 2005). When analyzed serologically, the suspect strains

produced weak reactions with S. dysenteriae serovars 4 and 16 and E. coli 0159 and

0173 antisera, however antisera prepared from one of the suspect isolates was

completely absorbed by antigens from S. dysenteriae serotype 4 and E. coli 0159 (Melito

et al., 2005). In addition, all six strains tested PCR positive for the ipaH gene and the

invasion associated locus. Molecular typing by PCR-RFLP of the rfb gene produced a S.

dysenteriae serovar 2 and E. coli 0112ac pattern (Melito et al., 2005). Based on these

analyses, the authors proposed the six suspect isolates represented a novel serovar of S.

dysenteriae.

Foodborne Outbreaks Involving l/tige//at

A summary of selected recent foodborne shigellosis outbreaks is given in Table 2-

2. Characteristic of foodborne shigellosis, several recent outbreaks have been associated

with foods consumed raw (Martin et al., 1986; Fredlund et al., 1987; Davis et al., 1988;









Table 2-2. Selected foodborne outbreaks involving .\/nlge/l
Year Serogroup Food Product(s) Implicated Reference(s)
1983 S. sonnei Tossed salad Martin et al., 1986
1986 S. sonnei Shredded lettuce Davis et al., 1988
1986 S. sonnei Raw oysters Reeve et al., 1989
1987 S. sonnei Watermelon Fredlund et al., 1987
1988 S. sonnei Uncooked tofu salad Yagupsky et al., 1991;
Lee et al., 1991
1989 S. flexneri 4a German potato salad Lew et al., 1991
1992 S. flexneri 2 Tossed salad Dunn et al., 1995
1994 S. sonnei Iceberg lettuce Long et al. 2002;
Kapperud et al., 1995;
Frost etal., 1995
1995-6 S. sonnei Fresh pasteurized milk cheese Garcia-Fulgueiras et
al., 2001
1996 S. flexneri Salad vegetables PHLS, 1997
1998 S. sonnei Uncooked, chopped curly CDC, 1999
parsley
1998 S. flexneri Restaurant-associated, source Trevejo et al., 1999
unknown
1999 S. boydii 18 Bean salad (parsley or cilantro) CDPH, 1999
2000 S. sonnei Five layer bean dip CDC, 2000;
Kimura et al., 2004
2001 S. sonnei Raw oysters Terajima et al., 2004
2002 .N/igel/t spp. Greek-style pasta salad TPH, 2002

Cook etal., 1995; Dunn etal., 1995; Frost etal., 1995; Kapperud etal., 1995; Public

Health Laboratory Service (PHLS), 1997; CDC, 1999) and processed or prepared by

hand (Martin etal., 1986; Lee etal., 1991; Lew etal., 1991; Yagupsky etal., 1991; Dunn

et al., 1995; Chicago Department of Public Health (CDPH), 1999; Trevejo et al., 1999;

Toronto Public Health (TPH), 2002). In 2000, a multi-state outbreak of shigellosis was

traced to a commercially prepared five-layer bean dip (Kimura et al., 2004). This

outbreak involved 406 persons (14 hospitalizations, 0 deaths) across 10 states. After

extensive epidemiological investigation, numerous problems were identified in the

manufacturing process and investigators determined that the source of the outbreak was

most likely an infected food-handler (Kimura et al., 2004). This outbreak demonstrates









the vulnerability of our food supply to point-source contamination with /ngel// followed

by wide distribution and subsequent infection of many consumers (Kimura et al., 2004).

Prevalence of ,\/ige/,t Food and Food Handlers

Despite the high incidence of shigellosis, there is limited data on the prevalence of

,\/ige/ll among food handlers or on food products. A study investigating the presence of

enteropathogens among food handlers in Irbid, Jordan, isolated ./nge//At from the stools

of four out of 283 examined food handlers (al-Lahham et al., 1990). Mensah et al. (2002)

evaluated 511 food items from the streets of Accra, Ghana, from which S. sonnei was

isolated from one sample of macaroni. It was noted that the macaroni was served using

bare hands instead of clean utensils, which may have led to the S. sonnei contamination.

Wood et al. (1983) examined foods from Mexican homes, commercial sources in

Guadalajara, Mexico, and from restaurants in Houston, TX, for contamination with

bacterial enteropathogens. While no ,h/ilge/l was isolated from foods sampled from 12

Houston restaurants or from food commercially prepared in Guadalajara, Mexico, four

isolates were obtained from meals prepared in Mexican homes. These studies

demonstrate the importance of proper food handling and the role food handlers in the

transmission of \//lgel//i

In response to President Clinton's National Food Safety Initiative (January 1997)

and Produce & Imported Foods Safety Initiative (October 1997), the U.S. Food and Drug

Administration (FDA) has investigated the presence of foodborne pathogens, including

.\/lg/ll, on imported and domestic produce (FDA, 2001a; FDA, 2001b; FDA, 2003). An

FDA survey of imported broccoli, cantaloupe, celery, parsley, scallions, loose-leaf

lettuce, and tomatoes found .\/ige/ll contamination in nine of 671 total samples: three of

151 cantaloupe samples, two of 84 celery samples, one of 116 lettuce samples, one of 84









parsley samples, and two of 180 scallion samples (FDA, 200 la). Another FDA survey of

domestically grown fresh cantaloupe, celery, scallions, parsley and tomatoes found

,/Nige/lt contamination in five of 665 total samples: one of 164 cantaloupe samples, three

of 93 scallion samples, and one of 90 parsley samples (FDA, 2003). An additional survey

of imported produce was conducted; however at the time of this publication results were

not publicly available (FDA, 2001b).

Survival of \l.,li.ge/l

Environmental Factors on Survival of \/. ,,e/ht

\/nge//At spp. are heat sensitive, acid resistant, salt tolerant bacteria that can

withstand low levels of organic acids (Zaika, 2001; Zaika, 2002a; Zaika 2002b). Zaika

(2001) studied the survival of S. flexneri strain 5348 in brain heart infusion (BHI) broth

as a function of pH (2 to 5) and temperature (4 to 370C). When inoculated into BHI broth

adjusted to pH 5, S. flexneri demonstrated growth when held at 19, 28, and 37C, while

counts declined over time at temperature of 120C or lower. When inoculated into BHI

broth adjusted to pH 2, 3, or 4, inoculated S. flexneri counts declined over time at all

temperatures tested. S. flexneri in BHI broth adjusted to pH 2 reached undetectable levels

in 1 to 3 days when held at temperature of 19C or lower. In general, S. flexneri survival

was greater in BHI broth incubated at lower temperatures and adjusted to higher pH. This

study suggests that S. flexneri is acid resistant and that acidic foods may support the

survival of.\/i/gel/t over a long period of time (Zaika, 2001).

Zaika (2002a) studied the survival of S. flexneri strain 5348 in BHI broth (pH 4 to

6) containing 0.5 to 8% NaC1. In BHI adjusted to pH 6, S. flexneri grew in the presence

of <6% NaCl when held at 19 and 370C, and in the presence of <7% NaCl when held at

280C. Growth of S. flexneri was also observed in BHI broth adjusted to pH 5 containing









<2, <4, <4, <0.5% NaCl when held at 37, 28, 19, and 120C, respectively (Zaika, 2002a).

S. flexneri populations gradually declined in BHI adjusted to pH 4 at all incubation

temperatures and all levels of NaCl tested (Zaika, 2002a). Results from this study suggest

that S. flexneri is salt tolerant and may survive in salty foods such as pickled vegetables,

caviar, pickled herring, dry cured ham, and certain cheeses for extended periods of time

(Zaika, 2002a).

S. flexneri survival was also studied in BHI broth supplemented with organic acids

commonly found in fruits and vegetables (citric, malic, and tartaric acid) or fermentation

acids commonly used as preservatives (acetic and lactic acid) at 0.04M and adjusted to

pH 4 with HC1 or NaOH (Zaika, 2002b). Fermentation acids (acetic and lactic acid) had a

greater effect on survival than citric, malic, and tartaric acids (Zaika, 2002b). When

incubated at 370C, S. flexneri survived for 1 to 2 days in the presence of each organic

acid tested. At 40C, S. flexneri survived in the presence of all the organic acids tested for

longer than 55 days (Zaika, 2002b). This study suggests that organic acids may aid in the

inactivation of .\/lge//, however foods with low levels of acids stored at low

temperatures may support the survival of the bacterium for extended periods of time

(Zaika, 2002b).

Temperature is an important factor in survival of shigellae. Freezing (-200C) and

refrigeration (40C) temperatures support survival, but not growth, of.\//lge/la

(International Commission on Microbiological Specifications for Foods (ICMSF), 1996).

When studied in nutrient broth, the observed temperature ranges which permitted growth

for S. sonnei and S. flexneri were 6.1 to 47.10C and 7.9 to 45.20C, respectively (cited in

ICMSF, 1996); however .\/nlge//t can survive for extended periods of time when stored at









-200C or at 40C. Elevated temperatures are less permissive for .\/nge//A survival, and

traditional pasteurization and cooking temperatures are sufficient for inactivation. Evans

et al. (1970) calculated the decimal reduction time (D-value) of S. dysenteriae in

pasteurized whole milk to be 0.0008 sec at 82.20C. When studied in nutrient broth, most

strains of S. sonnei and S. flexneri were inactivated within 5 min at 630C (cited in

ICMSF, 1996). Sublethal heat exposure can sensitize \/ngel//t to selective components of

microbiological media. Tollison and Johnson (1985) demonstrated that S. flexneri

sublethally heat-stressed by exposure to 50.00C for 30 minutes in phosphate buffer

became sensitive to 0.85% bile salts and 0.50% sodium desoxycholate. Since these

compounds are ingredients in several enrichment and isolation media used for detection

of .\/ige//i, the thermal history of the food sample to be analyzed should be known.

Survival of \/,ge//,t on Fomites

Inanimate objects, or fomites, can serve as vectors for transmission of .\/ige//At and

there have been several reports on the survival of.\/ilgel//t on various surfaces (Spicer,

1959; Nakamura, 1962; Islam et al., 2001). Spicer (1959) studied the survival of S.

sonnei dried on cotton threads at room temperature and under refrigeration at various

levels of relative humidity. In general, survival was better at refrigerated temperatures

and at high (84%) and low (0%) relative humidity (RH). S. sonnei remained detectable on

the cotton threads after 12 days at 5-100C (84% RH) (Spicer, 1959). Similar results were

observed using 10 different strains of S. sonnei inoculated on cotton, glass, wood, paper,

and metal at various temperatures (-200C to 450C) (Nakamura, 1962). When held at -

200C, most of the strains survived for more than 14 days on each surface, however

surfaces held at 450C did not support survival of most strains (Nakamura, 1962).

Investigations on the survival of S. dysenteriae serotype 1 on cloth, wood, plastic,









aluminum, and glass objects suggest that 1.5 to 4 hours post-inoculation, S. dysenteriae

serotype 1 enters a viable but non-culturable (VBNC) state (Islam et al., 2001). While no

S. dysenteriae serotype 1 could be recovered after 5 days by conventional culture

methods, viable cells could be observed using fluorescent antibody techniques (Islam et

al., 2001). Whether or not .\lNge/tl is able to achieve a true VBNC state or if these results

demonstrate the inadequacy of plating media in recovering viable ./nge//At has not been

fully investigated. Nevertheless, these studies demonstrate that fomites can sustain viable

\/ngel//l for an extended time and serve as vehicles in transmission of the pathogen.

Survival of \l/ige/hl, in Food and Water

.\/lgellt can survive in water with little decline in population. Rafii and Lunsford

(1997) inoculated distilled water with S. flexneri at 2.8 x 108 colony forming units

(CFU)/ml and held the samples at 40C. After 26 days, 9.2 x 107 CFU/ml of S. flexneri

survived. This high survival rate of S. flexneri in water supports the historical association

of shigellosis outbreaks and water sources.

Fruits and vegetables can support the growth or survival of.\/ige//t Escartin et al.

(1989) artificially contaminated fresh cut papaya, jicama, and watermelon with S. sonnei,

S. flexneri, or S. dysenteriae and within 6 hours at room temperature, growth was

observed. Rafii and Lunsford (1997) inoculated raw cabbage, onion, and green pepper

with S. flexneri and although counts decreased slightly at 40C, survival was observed

after 12 days on onion and green pepper, at which time sampling was terminated due to

spoilage (Rafii and Lunsford, 1997). S. flexneri survival was observed on the cabbage for

26 days. Wu et al. (2000) studied survival of S. sonnei on whole and chopped parsley

leaves. At 210C, growth on chopped parsley was observed at a similar rate to that in









nutrient broth (Wu et al., 2000). At 40C, populations declined on both chopped and

whole parsley throughout the 14-day storage period, however S. sonnei survived

regardless of initial population (Wu et al., 2000). These studies demonstrate .\/nge//t

survival on refrigerated produce for periods of time that exceed expected shelf life.

Low pH foods can support survival of.\/Nige//t when held at refrigerated

temperatures. Bagamboula et al. (2002) demonstrated S. sonnei and S. flexneri survival in

apple juice (pH 3.3 to 3.4) and tomato juice (pH 3.9 to 4.1) held at 7C for 14 days. No

reduction was observed in the tomato juice, while a 1.2 to 3.1 logo CFU reduction was

observed in apple juice over the 14 day study. Rafii and Lunsford (1997) observed S.

flexneri survival in carrot salad (pH 2.7 to 2.9), potato salad (pH 3.3 to 4.4), coleslaw (pH

4.1 to 4.2), and crab salad (pH 4.4 to 4.5) held at 40C. Sampling was terminated at day 11

for the carrot and the potato salads, at which time S. flexneri counts decreased from 4.3 x

106 to 4.2 x 102 CFU/g and from 1.32 x 106 to 8.5 x 102 CFU/g, respectively. Sampling of

the coleslaw and the crab salads ceased due to product spoilage on days 13 and 20,

respectively, however S. flexneri survived at levels of 2.16 x 104 and 2.4 x 105 CFU/g,

respectively. These studies indicate that ./nge//lt survived at refrigerated temperatures

despite the presence of background microflora and low pH.

Prepared foods can also support the survival of ./ngel//l Islam et al. (1993b)

investigated the growth and survival of S. flexneri in boiled rice, lentil soup, milk, cooked

beef, cooked fish, mashed potato, mashed brinjal, and raw cucumber. All food samples,

except raw cucumber, were autoclaved prior to inoculation. Ten gram or 10 ml samples

of each food were inoculated with 105 cells of S. flexneri, incubated at 5, 25, or 37C and

sampled over 72 hr. All of the foods tested supported growth up to 108 tolO10 cells per g









or ml within 6 to 18 hr after inoculation at 250C and 370C (Islam et al., 1993b). Initial

inoculum levels were maintained throughout the 72 hr holding period for all foods,

except rice and milk. S. flexneri counts in rice decreased by approximately 1 log after 72

hr at 50C, whereas counts in milk dropped after 48 hr but then returned to the initial

inoculum level by 72 hr. These results demonstrate the ability of.\/ige//At to grow and

survive in a variety of prepared foods that may be contaminated by an infected food

handler.

\/ngel//l is able to survive on produce packaged under vacuum or modified

atmosphere. Satchell et al. (1990) investigated the survival of S. sonnei in shredded

cabbage packaged under vacuum or in a modified atmosphere of nitrogen and carbon

dioxide when stored at room temperature or under refrigeration. When test samples were

stored at room temperature, counts of inoculated S. sonnei remained at level for up to

three days, at which time populations began to drop. When test samples were stored

under refrigeration, S. sonnei counts did not drop after seven days. In the latter study,

refrigerated samples maintained a constant pH throughout the study, while samples stored

at room temperature had a drop in pH that may have contributed to the decline in S.

sonnei cell numbers.

Current Understanding of \/,ige/ht Pathogenesis

This discussion will be limited to an overview of epithelial cell invasion by

\/lNgel//, potential roles of the IpaH effector proteins and blockage of autophagy by IcsB.

The reader is directed to the following reference for a more complete review of.\/ige//At

pathogenesis and the toxins produced by .\/nge//t spp.: Warren et al., 2006.









.\/nlgel// Invasion of Epithelial Cells

The invasion of the local epithelium of the colon (large intestine) is presented in

Figure 2-1. Once ingested, shigellae move through the gastrointestinal tract to the colon,

where they translocate the epithelial barrier via M cells that overlay the solitary lymphoid

nodules (Suzuki and Sasakawa, 2001). Upon reaching the under side of the M cells,

\/nge//lt infect macrophages and induce cell apoptosis (Suzuki and Sasakawa, 2001).

Once released from the macrophage, .l/ngel//t enters neighboring epithelial cells. .\l/gell//

first forms a membrane bound protrusion into the adjacent cell. This protrusion must

distend two membranes: one from the donor cell, and another from the recipient cell

(Parsot and Sansonetti, 1996). As the protrusion pushes further into the recipient cell, it is

taken up by the recipient cell resulting in the bacteria enclosed in a double-membrane

vacuole (Monack and Theriot, 2001). Intercellular spread is completed when .\/nlg//t

escape from the double-membrane vacuole, releasing it into the cytosol of the recipient

cell. In response to invasion, epithelial cells produce pro-inflammatory cytokines that

contribute to inflammation of the colon (Suzuki and Sasakawa, 2001).

Potential Roles of the IpaH Effector Proteins

IpaH7.8 facilitates escape from endocytic vacuoles

Fernandez-Prada et al. (2000) reported that IpaH7.8 of S. flexneri was required for

efficient escape from endocytic vacuoles. Human monocyte-derived macrophages

(HMDM) and the J744 mouse macrophage cell line were infected with S. flexneri 2457T

and pWR700, an ipaH7.8 deletion mutant of S. flexneri 2457T. After the infected

HMDM and J744 cells were incubated in the presence of gentamicin and chloroquine,

results showed that more pWR700 than 2457T was present within endocytic vacuoles,

suggesting that IpaH7.8 is required for escape from the vacuole. The contrast between











I
Meters through M cells that overlay the solitary lymphicelld nodules, infect the


| WT _



- - -- -



Macrophage 0 '


Figure 2-1. Invasion of epithelial cells by ,Nlell, spp. In the large intestine, ,liell,
enters through M cells that overlay the solitary lymphoid nodules, infect the
resident macrophage and induce cell apoptosis. Once released from the
macrophage, .\/nge//At enters the neighboring enterocytes and escape from the
double membrane vacuole that encompasses them. .\/ilge//l multiply in the
cytoplasm of the host cell and polymerize actin for motility. IcsB is required
to evade autophagic recognition; therefore icsB mutants are degraded once
they escape from the vacuole. Figure reproduced from Ogawa and Sasakawa,
2006.

pWR700 and 2457T localization within endocytic vacuoles was more pronounced in the

J744 cell line. One explanation for this was that the HMDM in tissue culture represented

a heterogenous population of cells, at various stages of differentiation. The authors

further suggested that the ipaH genes may play a bigger role in monocytes than

macrophages (Fernandez-Prada et al., 2000). It is noteworthy to mention that ipaH4.5

and ipaH9.8 mutants behaved like the wild-type 2457T in both HMDM and J744 cells,

suggesting their role in virulence differs from that of ipaH7.8.

Subversion of host cell signaling by IpaH9.8

Toyotome et al. (2002) investigated the secretion of IpaH proteins from S. flexneri

in broth cultures and determined that IpaH proteins are exported by type III secretion









after entry into the host cell. Further investigation showed that once secreted, IpaH9.8

accumulates in the nucleus, while small amounts are present in the cytoplasm. IpaH9.8

has similar structure to the Salmonella Typhimurium protein SspH1, which belongs to the

bacterial LPX repeat protein family. Upon infection of the local epithelium, intracellular

pathogens, such as .\,/ge//l and Salmonella, elicit the secretion of proinflammatory

cytokines, such as interleukin 8 (IL-8) (Haraga and Miller, 2003). Production of IL-8 and

other cytokines in response to bacterial invasion are dependant, in part, on activation of

transcription factor NF-kappa B. Haraga and Miller (2003) demonstrated that SspH1 and

IpaH9.8 both localize to the mammalian nucleus and inhibit nuclear factor kappa B (NF-

kappa B)-dependent gene expression (Haraga and Miller, 2003). In this way, IpaH9.8

serves to subvert host cell signaling events involved in the immune response to epithelial

invasion.

Blockage of Autophagy by IcsB

Autophagy is a critical process in eukaryotic cells in which undesirable cellular

components or organelles, including invading microbes, are degraded. Recently, Ogawa

et al. (2005) identified IcsB as critical in the camouflage against autophagic recognition.

IcsB is an effector protein exported by type III secretion and is located on the cell

surface. IcsB mutants are fully invasive and capable of escaping from the vacuole, but

defective in its ability to multiply within the host cell (Ogawa et al., 2003). In the absence

of IcsB, the autophagy protein Atg5 recognizes and binds to IcsA (VirG), thus initiating

autophagosome formation. Ogawa et al. (2005) demonstrated the IcsA (VirG) binding

region for both Atg5 and IcsB is the same, and Atg5 binding to IcsA (VirG) is inhibited

by IcsB in a dose-dependent manner. By blocking the binding if IcsA (VirG) by Atg5,









IcsB inhibits the autophagic recognition of ,\/nge//t within the host cell cytoplasm, thus

contributing to intracellular survival.

Genetic Relationship Between .\lhigel// and Escherichia coli

While it has been generally accepted that .\ngel//t are within the species E. coli,

recent studies have indicated that ,/nge,//i, like the other forms of pathogenic E. coli,

derived from different evolutionary origins, suggesting convergent evolution of the

.\liige'//A phenotype (Pupo et al., 2000). Rolland et al. (1998) used restriction fragment

length polymorphism of rDNA (ribotyping) to group 75 strains of ./,ge//l 13 strains of

enteroinvasive E. coli (EIEC) and 72 E. coli strains of the E. coli Reference (ECOR)

Collection, which have been classified into four phylogenic groups (A, B B2 and D).

The S. sonnei, S. flexneri and most S. dysenteriae ribotypes were closely related to

phylogenic group D, while S. dysenteriae serotype 1 strains were closely related to

phylogenic group B and S. boydii strains were spread between phylogenic group D and

Bl (Rolland et al., 1998). In contrast, the ribotypes of EIEC strains were widely

distributed among phylogenic groups A, B and B2. This evidence suggests that ,\/1ge,//

and EIEC derived from different origins.

Pupo et al. (2000) sequenced eight housekeeping genes from four regions of the

chromosome for 46 strains of.\hi/ge//t representing all four serotypes. Three distinct

clusters of,/hige//At were identified and although S. sonnei and S. dysenteriae serotype 1,

8 and 10 did not group in the main three clusters, they fell well within the species E. coli

(Pupo et al, 2000). As with the study by Rolland et al. (1998), S. boydii serotype 13 was

distantly related to the other .nge//l t strains. Cluster 1 contained most of the S. boydii

and S. dysenteriae strains along with S. flexneri serotypes 6 and 6A. Cluster 2 contained

seven S. boydii strains and S. dysenteriae serotype 2. Cluster 3 contained S. flexneri









serotypes 1-5 and S. boydii serotype 12. Unlike the results from ribotyping, the use of

multiple genes for phylogenic analysis revealed greater genetic diversity among the

strains of.\/hge//,i, further suggesting that .ligel//A, derived from different evolutionary

origins.

Fukiya et al. (2004) used comparative genomic hybridization microarray analysis

to compare the gene content of E. coli K-12 with that of 22 pathogenic E. coli and

.\iigell.A strains. When compared to the E. coli K-12 genome, the genomes of S. sonnei,

S. boydii and S. flexneri 2a were missing only 613, 533 and 716 open reading frames

(ORFs). The genomes of the other pathogenic E. coli strains were missing similar

numbers of ORFs. Subsequent phylogenic analysis revealed a close relationship between

three of four EIEC strains and the three strains of.\ligeill, which suggests EIEC and

.h/gell// form a single E. coli pathovar (Fukiya et al. 2004; Yang et al. 2005).

Providing further to the body of evidence that ,./ige/ll and EIEC derived from

different origins, the complete genomes of S. boydii serotype 4 (strain 227), S.

dysenteriae serotype 1 (strain 197) and S. sonnei (strain 046) have recently been

sequenced (Yang et al., 2005). Comparative genomics among the newly sequenced

.liige/ll genomes and the previously sequenced genomes of S. flexneri 2a (strain 301)

and E. coli K-12 (strain MG1655) supported previous work by Fukiya et al. (2004).

While the genomes of .\hlge//At share most of their genes with that of E. coli K-12, the

.h/gell// phenotype is a result of the gain and loss of functions through bacteriophage-

mediated gene acquisition, insertion sequence (IS)-mediated DNA rearrangements and

formation of pseudogenes (Yang et al., 2005). For example, the chromosome and

virulence plasmid of S. sonnei strain 046 contained 327 and 28 intact IS elements and 67









and 68 partial IS elements, respectively. In contrast, the E. coli K-12 genome contained

only 37 intact IS elements and seven partial IS elements. These studies taken together,

demonstrate that the i.\ige/ll have evolved from distinct E. coli ancestors through

convergent evolution.

Detection Methods for ',\/nge/Al in Foods

Conventional Culture Methods for ./n.,e/l t

Traditional microbiological media for enrichment and isolation of.\/ng'e/,t

Traditional microbiological techniques make use of selective and differential media

for the enrichment and isolation of.\//nge//t Many variants of enrichment and plating

media have been investigated for optimal recovery, often with conflicting results.

Although ./nge//At is readily isolated from clinical samples, food samples are more

problematic. Isolation of .\1/ge//t from food samples can be inhibited by indigenous

microflora, especially the coliform bacteria and Proteus spp. (ICMSF, 1996). The

addition of the antibiotic novobiocin to liquid and solid media has been shown to improve

the isolation of S. flexneri and S. sonnei from investigated foods (cited in ICMSF, 1996).

Contamination of food products with ,.l/ge/ll results primarily through a food handler

with poor personal hygiene; therefore the concentration of /ge//at may be very low

compared to that of the indigenous microflora (Lampel and Maurelli, 2001). Currently,

selective media are not available that adequately suppress the growth of background

microflora, therefore .\h/gel// is often overgrown by competing microorganisms (Lampel

and Maurelli, 2001). More research is needed to determine more appropriate selective

media and enrichment conditions for the isolation of.\,/ge//at from food samples.

Two enrichment broths initially used for the isolation of/Nige//at were Selenite-F

(SF) and Tetrathionate (TT) broth. These broths were originally designed for the isolation









of Salmonellae, but due to the lack of specific enrichment media for shigellae they were

used as all-purpose enteric enrichment broths (Taylor and Schelhart, 1969). Sodium

selenite, although selective for salmonellae, is toxic to ,.hige/la (and most enterics);

therefore it is no longer used in enrichment procedures for .\Shge/l/ TT is a peptone-

based broth with bile salts and sodium thiosulfate that inhibits growth of most Gram-

positives and Enterobacteriaceae. While TT is routinely used for the enrichment of

Salmonella, it is rarely used for .hiigll./ Gram-negative (GN) broth is a peptone-based

broth with glucose and mannitol. The concentration of mannitol in GN broth is higher

than glucose to promote mannitol fermentors, like .lilgel/i Both TT and GN broths

contain bile salts, which can be inhibitory to stressed cultures (Tollison and Johnson,

1985). Furthermore, GN broth contains sodium desoxycholate, which has been shown to

inhibit heat-stressed shigellae (Uyttendaele et al., 2001). Regardless, GN broth is listed as

an alternate enrichment medium for the detection of ./hige//t from food by some standard

methods (Lampel, 2001).

Current enrichment procedures (FDA, 1998; Lampel, 2001, Health Canada, 2004)

use a low carbohydrate medium, .h/ge/lt broth (SB) with addition of novobiocin, for the

detection/isolation of.\h/ige//t One study reported that acids produced by other

Enterobacteriaceae during the fermentation of carbohydrates were toxic to shigellae

(Mehlman et al., 1985); however, other studies have shown that .iige/llt can grow at pH

4.5 to 4.75 (Bagamboula et al., 2002) and survive at pH 4.0 (Zaika, 2002). Nevertheless,

the use of SB limits the production of acids, and thereby limits the introduction of low

pH, during enrichment. SB is also less stringent than TT broth and GN broth for the

enrichment of ,hgel//At since it contains neither bile salts nor sodium desoxycholate. A









recent study investigated SB, GN broth, tryptic soy broth, and Enterobacteriaceae

Enrichment (EE) broth with the addition of novobiocin for enrichment/detection of

shigellae (Uyttendaele et al., 2001). When incubated in GN broth, S. sonnei was unable

to grow to comparable levels as observed in SB and EE broths. EE broth, however, has

been reported inhibitory to S. boydii (Warren, 2003; Warren et al., 2005b).

Multiple plating media with differing selectivity can be used to increase the

chances of.\/i'ge//t isolation. The most common low selectivity medium used for plating

.\lge/ll is MacConkey Agar (MAC). Eosin methylene blue (EMB) or Tergitol-7 (T7)

agars can also be used. Since differentiation on MAC is solely based on lactose

fermentation, .\nge//at colonies look similar to those of many lactose negative

competitors (Uyttendaele et al., 2001). On MAC, ./hige/hi colonies are translucent or

slightly pink, with or without rough edges. .\/nge//t produce colorless colonies on EMB

and bluish colonies on the yellowish-green T7 agar (Lampel, 2001).

Intermediate selectivity media useful in isolating .\nge//At are desoxycholate citrate

agar (DCA) and xylose lysine desoxycholate agar (XLD). .\/ige/ll spp. produce colorless

colonies on both DCA and XLD. Bhat and Rajan (1975) reported XLD superior to DCA

for the isolation of.\//ge//at since DCA required a 48 hr incubation to show clear colony

morphology as opposed to overnight incubation for XLD. Unfortunately, D-xylose,

which serves as a differentiating agent on XLD agar, is fermented by some strains of S.

boydii while most .nge//lt cannot ferment xylose (Bhat and Rajan, 1975). Thus some

strains of.\/hge//At will be missed if XLD is used as the sole plating medium.

Highly selective media for .\/nge//t spp. include Salmonella-.ilgell/t agar (SSA)

and Hektoen Enteric agar (HEA). A problem associated with SSA and HEA is that they









may be too stringent for some strains of.\/ngell, especially if the culture is stressed

(Lampel, 2001; Uyttendaele et al., 2001). .h/ge//At spp. produce colorless, translucent

colonies on SSA and green colonies on HEA.

A newly developed plating medium, Chromogenic .\hge//At spp. Plating Medium

(CSPM; R&F Laboratories, West Chicago, IL), offers medium selectivity (bile salts,

antibiotic supplementation) with an alternative to differentiation based on lactose

fermentation. Instead, differentiation on CSPM is based on proprietary components

consisting of select carbohydrates, pH indicators, and chromogens (Dr. Larry Restaino,

personal communication). .\hnge/ll spp. produce white to clear colonies on CSPM while

competitors produce colored colonies. CSPM has been compared to MAC and SSA for

the recovery of S. boydii and S. sonnei from tomato surfaces with no significant

differences in recovery observed (Warren, 2003; Warren et al., 2005a). Further

evaluation of CSPM against other strains of S. boydii and S. sonnei as well as the other

serogroups of.\ /ge//At is needed.

Recent studies at the FDA, Laboratory for Enteric and Sexually Transmitted

Diseases have demonstrated stable lactose-positive mutations in stationary phase S.

sonnei (Dr. D.J. Kopecko, FDA, personal communication). DNA sequencing experiments

have revealed slip-strand mutations within the lac repressor (lacl) that are responsible for

the lactose-positive phenotype. These mutations are significant for detection of S. sonnei

in food since typical colonies on plating media (MAC) are differentiated based on the

utilization of lactose, the typical S. sonnei phenotype being lactose-negative.









The FDA BacteriologicalAnalytical Manual culture method for detection of .\/ngel/a in
foods

The FDA Bacteriological Analytical Manual (BAM) culture method for the

isolation and detection of .\/nge//t spp. from food utilizes a combination of low

carbohydrate enrichment, anaerobic conditions, and elevated temperature (FDA, 1998).

Briefly, a 25 g sample of the food product is transferred to 225 ml of.\/nge//t broth (SB)

to which novobiocin (0.5 tg/ml for S. sonnei; 3.0 tg/ml for other .\nge//lt spp.) has been

added. Samples are held at room temperature for 10 min and periodically shaken. Sample

supernatants are transferred to an Erlenmeyer flask and the pH adjusted to 7.0 + 0.2 with

sterile 1 N NaOH or 1 N HC1. Flasks are incubated anaerobically for 20 hr (44C for S.

sonnei; 42C for all other .\lnge//t spp.) and the enrichments are streaked on MAC.

Confirmation of suspicious colonies involves tests for motility, H2S, gas formation, lysine

decarboxylase, and fermentation of sucrose or lactose. Isolates negative for all

confirmatory tests are further tested using biochemical reactions including adonitol,

inositol, lactose, potassium cyanide, malonate, citrate, salicin, and methyl red. Shigellae

are negative for all except methyl red. Antisera agglutination is then used to identify any

culture displaying typical .\/nlge/t characteristics.

June et al. (1993) evaluated the effectiveness of the BAM culture method for

\/nlge/ll Two strains of S. sonnei, strains 9290 and 25931, were inoculated on potato

salad, chicken salad, cooked shrimp salad, lettuce, raw ground beef, and raw oysters.

Using either unstressed or chilled stressed cells, acceptable recovery was achieved for

both strains from the potato salad, chicken salad, cooked shrimp salad and lettuce

samples, but not from the ground beef and raw oyster samples. An approximate 4-log unit

difference in recovery from ground beef samples was observed between the two strains,









suggesting high strain variability. Given the low infective dose of .///lel/At (as low as 10

cells), the BAM was considered ineffective for the evaluation of raw ground beef and raw

oysters.

In 2002, the BAM culture method for ./nge//At was evaluated using two strains of

unstressed, chill-stressed, or freeze-stressed S. sonnei (strains 357 and 20143) on selected

types of produce (Jacobson et al., 2002). Acceptable recovery of unstressed cells (<1.0 x

101 CFU/25g) and chill-stressed or freeze-stressed cells (<5.2 x 101 CFU/25g) were

observed for all produce types tested (Jacobson et al., 2002). More recently, similar tests

of a modified BAM protocol showed unacceptable recovery of unstressed S. sonnei

(patient isolate from an outbreak involving an unknown source) and S. boydii (patient

isolate from an outbreak involving bean salad) on tomatoes at 1.9 x 102 CFU/tomato and

>5.3 x 105 CFU/tomato, respectively (Warren, 2003). These results support the

observation that significant variation exists among strains of S. sonnei and demonstrate

the importance of including the other serogroups of.\/ige//At when evaluating culture

methods for detection in food.

Other culture methods for the detection of \/,le/ht in foods

Alternate culture methods for the detection of .\/ige//t in foods can be found in

Health Canada's Compendium of Analytical Methods (Health Canada, 2004) and the

American Public Health Association's Compendium of Methods for the Microbiological

Examination of Foods (CMMEF) (Lampel, 2001). The Health Canada method is based

on the BAM culture method with a few modifications; most notably enrichments for all

\/nge//At spp. (including S. sonnei) are supplemented with 0.5 tg/ml novobiocin and

incubated anaerobically at 420C. The CMMEF culture method is also similar to the BAM

culture method, however the level of novobiocin in S. sonnei enrichment media is lower









(0.3 pg/ml) and enrichment conditions are aerobic at 370C. The CMMEF further suggests

that two to three plates of various selective media be used to streak the enriched cultures:

MAC for low selectivity, XLD for intermediate selectivity, and HEA for high selectivity.

Confirmation of suspicious colonies is similar to methods described above for the BAM

culture method.

Recently, the CMMEF culture method and an enrichment procedure involving EE

broth have been compared to the BAM culture method for the detection of S. boydii and

S. sonnei on tomato surfaces (Warren, 2003). Natural tomato microflora was found to

have a great impact on recovery of S. sonnei and completely inhibited recovery of S.

boydii in all three culture methods. No significant differences (P > 0.05) were observed

among the culture methods for detection of S. sonnei, or between the BAM and CMMEF

culture methods for the detection of S. boydii. EE broth was found to be inhibitory to S.

boydii. These results suggest the need for more selective enrichment protocols for the

evaluation of.\/lgel//t spp. in food.

Immunological Methods for .\li.ge// Detection

Lipopolysaccharide (LPS) of \/nhl./l/t

Gram-negative bacterial LPS consists of three distinct regions: lipid A, core

oligosaccharide, and a serotype-specific O-polysaccharide chain (O-antigen) (Neidhardt,

2004). The lipid A portion anchors the LPS molecule to the outer membrane. The core

oligosaccharide is composed of two regions: the inner core and the outer core. The inner

core is common to many enterobacterial species and is composed of heptose and 2 keto-3

deoxyoctulosonate (KDO), while the outer core is rich in hexose and tends to be more

species-specific (Tsang et al., 1987). Studies of the core structure of S. flexneri have

indicated that serotypes 1 to 5 and variants X and Y share the E. coli R3 core (Carlin and









D-Galp(a -2)D-Galp(a 1-2)D-Glcp(a 1-3)D-G Icp(( 1 -
3

A Pl
o-Glep



D-G lp(u I -2)D-G [cp(a -2)D-Galp(a 1-3)D-G Ip(u 1 -
3
B I
al
D-GlcpNAc

Figure 2-2. Structural differences between hexose regions of.\/l/g/e//t sonnei and \/nlg///t
flexneri lipopolysaccharides. A) S. sonnei and S. flexneri serotype 6 share the
hexose region of the E. coli R1 core. B) S. flexneri serotypes 1-5 and X and Y
variants have a hexose region identical to the E. coli R3 core. (Structures
compiled from Jansson et al., 1981). Abbr.: Gal = galactose, Glc = glucose,
GlcNAc = 2-acetamido-N-deoxyglucose.

Lindberg, 1986), while S. flexneri serotype 6 and S. sonnei share the E. coli R1 core

(Gamian and Romanowska, 1982; Carlin and Lindberg, 1986; Viret et al., 1992) (Figure

2-2). Antibodies have been developed with specificity to the inner core of.\/i/ge//t spp.

(Rahman and Stimson, 2001). The O-antigen is a polymer of repeating saccharides that is

highly variable among species. Strains sharing identical O-antigen repeating units are of

the same serotype. The O-antigens of all S. flexneri serotypes are a repeating

tetrasaccharide (Carlin and Lindberg, 1986), while the O-antigen of S. sonnei is a

repeating disaccharide (Kenne et al., 1977) (Figure 2-3).

Differences in S. sonnei form I and form II lipopolysaccharide

Virulent S. sonnei produce smooth colonies, termed form I, which result from

expression of the O-antigen. Unlike other ./iige//t species, the LPS genes (rfc and rfb) of

S. sonnei are located on a large virulence plasmid, which can be spontaneously lost at









A L-AItNAcAp(aI -3)4-N-DFucNAcp

B -4)D-GalA(f 1-3)D-GalNAc(fl -2)L-Ac3Rha(a -2)L-Rha(a l-

C -2)L-Rhap(a 1-2)L-Rhap(a -3)L-Rhap(al 1-3)D-GIcpNAc(fll-


Figure 2-3. O-antigen repeating subunits of S. sonnei and S. flexneri lipopolysaccharide.
(A) The disaccharide repeating subunit of the S. sonnei O-antigen, (B) the
tetrasaccharide repeating subunit of S. flexneri serotype 6, and (C) the
tetrasaccharide repeating subunit of S. flexneri serotypes 1-5 and X and Y
variants. (Structures compiled from Carlin and Lindberg, 1986; Dmitriev et
al., 1979; and Gamian and Romanowska, 1982). Abbr: LAltNAcA = N-acetyl-
L-altrosaminuronic acid, 4-N-DFucNac = 2-acetamido-4-amino-2,4,6-trideoxy-
D-galactose, GalA = galacturonic acid, GalNAc = N-acetylgalactosamine,
Ac3Rha = 3-O-acetylrhamnose, Rha = rhamnose, GlcNAc = 2-acetamido-N-
deoxyglucose.

high frequency. Sansonetti et al. (1981) investigated the stability of form I plasmids and

observed 1 to 45% plasmid loss from re-streaking form I colonies onto MAC and

incubating 24 hr at 370C. When the large virulence plasmid is lost, rough (avirulent)

colonies, termed form II, are produced that express the Enterobacteriaceae R1

lipopolysaccharide core (Sansonetti et al., 1981; Gamian and Romanowska 1982). A

defective mutant of form II S. sonnei LPS, termed R-form, is characterized by an

incomplete core region (Gamian and Romanowska, 1982) (Figure 2-4). As antibodies

specific for S. sonnei O-antigen will bind form I, but not form II or R-form LPS,

immunological detection methods with specificity for the O-antigen of.\l/ge//t are

compromised when the virulence plasmid is lost.

Immunological Detection Methods for Bacteria

Immunological methods, such as latex agglutination (LA), enzyme immunoassay

(EIA), or immunomagnetic separation (IMS), have been utilized for the detection of










A D-Galp(a 1-2)D-Galp(a 1 -2)u-Glcp(a 1-3)D-Glcp(a -3)LaD-Hepp( 1-3)LaD-Hepp( 1-5)dOc LA
3 7 4
I I
fll 1 P-PEtN
D-GIcp LaD-Hepp
3 7

fll al
[L-AItNAcAp(a 1 -3)4-N-DFucNAcp]o= 0-4 D-GIcNp



B P-PEtN
I
4
D-Galp(a I-2)D-Galp(a 1-2)D-Glcp(a 1-3)D-Glcp(a 1-3)LaD-Hepp(1-3)Lau-Hepp(1-5)dOclA
3 7 4

361 1 P-PEtN
D-Glcp Lao-Hepp




C P-PEtN
I
4
D-G Icpa(a -3)D-GIcp(a 1-3)LaD- Hepp( 1-3)LaD-Hepp( 1-5)dOclA
7 4
I I
1 P-PEtN
Lao-Hepp
7
I
al
D-GIcNpa

Figure 2-4. Lipopolysaccharides of S. sonnei. A) Form I, B) form II, and C) R-form.
(Structures compiled from Gamien and Romanowska, 1982).

many foodborne bacterial pathogens. LA assays involve latex particles coated with

antibodies specific for target bacteria. Binding causes a visible clumping of the latex

particle-bacteria complexes that can be seen with the naked eye. EIA is a term used to

describe many assay formats in which enzyme-labeled antibodies are used to bind

antigens with detection via a colorimetric reaction using the enzyme label. Microplate









readers are generally used for detection of the colorimetric signal and these assays can be

performed in high-throughput format. IMS involves small paramagnetic beads coated

with antibodies against the surface antigens of bacterial cells. Paramagnetic beads exhibit

magnetic properties only when placed within a magnetic field and show no residual

magnetism when removed from this field. After target bacteria are bound by the

antibody-coated beads, a magnetic field is used to separate the bead-bacteria complexes.

Once the magnetic field is removed, the bead-bacteria complexes return to suspension

(Olsvik et al., 1994). Detection of bead-bacteria complexes can be performed via direct

plating on microbiological media, direct microscopy, or nucleic acid amplification. LA,

EIA and IMS methods require the use of monoclonal or polyclonal antibodies to bind

specific antigens present on the surface of bacterial cells. For this reason, development of

antibodies with sufficient specificity is critical for the performance of immunological

methods. A summary of these and other rapid methods for .\/igel/h detection is presented

in Table 2-3.

Latex agglutination methods for .,/ltigel/

LA requires the prior isolation of a suspect colony on solid media. For this reason

LA methods cannot be used for detection of.\/ige//t directly from a food sample, but

rather can confirm or aid in characterization of suspected ./nge//lt colonies. Several latex

agglutination serotyping kits for .\higeli// are commercially available, however only one

representative kit will be discussed in this review.

The Wellcolex Colour .\/ligel/ Test (WCT-.\l/get/A/, Remel Inc., Lenexa, KS)

allows the identification of isolates to the species level using only two reagents, each

consisting of a mixture of red and blue latex particles coated with antibodies specific for











Table 2-3. Selected rapid methods for detection of.\/nlgel//


Target Matrix Sample Preparation Detection Detection Limit Reference
ipaH Tomatoes FTA filtration Nested PCR < 10' CFU/tomato Warren, 2003


Serogroup
S. sonnei
S. boydii
S. flexneri
S. dysenteriae 1

S. dysenteriae

S. flexneri
S. flexneri
S. sonnei form I

S. dysenteriae
S. flexneri
S. boydii
S. sonnei
S. dysenteriae
S. dysenteriae 1
S. flexneri
S. sonnei
S. dysenteriae 1
S. flexneri
S. dysenteriae
S. flexneri
S. boydii
S. sonnei
S. dysenteriae 1
S. flexneri


Water
Mayonnaise

Mussels

Lettuce
Various foods
Feces

Rectal swabs



Various foods
Feces


Feces

Sewage



Feces


SE, boiled extract
CEP

PE, CEP

Alkaline denaturation
SE, BDC
IMS

PE, DPSM



None
IMS


IMS

Immunocapture



IMS


Semi-nested PCR
PCR

PCR

PCR
Nested PCR
PCR

EIA



Biosensor
PCR


PCR

UPPCR



EIA


> 1.1 x 10 CFU/ml
102 -103 CFU/ml

101 -102 CFU/ml

1.0 x 104 CFU/g
1.0 x 101 CFU/g
1.0-1.5 x 101 CFU/ml

ND



> 4.9 x 104 CFU/ml
ND


1.0 x 101 CFU/g

5.0 x 10' CFU/ml



1.0 x 103 CFU


Theron et al., 2001
Villalobo and Torres,
1998
Vantarakis et al.,
2000
Lampel et al., 1990
Lindqvist, 1999
Achi-Berglund and
Lindberg, 1996
Sonjai et al., 2001



Sapsford et al., 2004
Achi et al., 1996


Islam and Lindberg,
1992
Peng et al., 2002



Islam et al., 1993a


Abbreviations: SE selective enrichment, PE pre-enrichment, ND not determined, LPS lipopolysaccharide, EIA enzyme
immunoassay, BDC buoyant density centrifugation, UPPCR universal primer PCR, DPSM direct plate on selective media, CEP
- chemical extraction/ ethanol precipitation.


ipaH
virA

virA

plasmid DNA
ial
LPS, ial

LPS



LPS
LPS, ial


LPS, ial

LPS, 16S rRNA



LPS









one of the four different .\/nge//A serogroups. Each reagent is added to one of two

identical sample spots of the suspect isolate. The presence of homologous antigen results

in the agglutination of one color coupled with a change in background color. Color

change combinations due to any of the four ./nge//At serogroups are easily distinguishable

as is the negative reaction in which the particles remain in smooth purple suspension.

Non-specific agglutination results in a purple agglutination with a clear background.

Bouvet and Jeanjean (1992) tested the WCT-,/nlgel//At against 100 .,\/ige//t isolates

from human stools and observed specificity and sensitivity at 100% and 98%,

respectively. The two ,.\/ge/ht strains that did not give visible aggregates were S.

dysenteriae type 2 and S. flexneri type 4. These results were supported by Kocka et al.

(1992). Using the WCT-,/ lgel///, 42 of 42 clinical .\ ilgell/ isolates and seven of eight

stock \/hge//At cultures were correctly grouped (Kocka et al., 1992). The stock ./nge//At

culture that was not correctly grouped had been repeatedly passed in culture that may

have resulted in the loss of some antigenicity. Lefebvre et al. (1995) evaluated WCT-

,\/ige/ll and six commercial slide agglutination ,.\/lge// serogrouping kits for accuracy.

The WCT-,s/ngel//A, was easily performed and interpretation of results was less subjective

than the other tests. WCT-\,/tge,//At met a performance standard of 90% accuracy in these

evaluations.

Enzyme immunoassay methods for \ltige//At

Although EIA methods for the detection of foodborne pathogens are common, the

only commercially available kits for .\l/ge//At spp are the Shigel-Dot A (for S.

dysenteriae), B (for S. flexneri), C (for S. boydii), and D (for S. sonnei) test kits (Science

Development and Management. Ltd., Bangkok, Thailand). The Shigel-Dot kits are

membrane dot-blot enzyme-linked immunosorbent assays (ELISA). These test kits have









been validated using 500 rectal swabs and have been compared to conventional culture

isolation and Western blot analysis. The diagnostic accuracy of the Shigel-Dot A, B, C,

and D was 99.2%, 95.0%, 94.0%, and 96.4%, respectively, when compared to

conventional culture isolation, and all were 100% when compared to the conventional

culture isolation and Western blot combined (Sonjai et al., 2001). Monoclonal antibodies

included in the Shigel-Dot D kit are reported to detect both S. sonnei form I and form II

LPS.

Immunomagnetic separation methods for .\/lge//At detection

IMS has been investigated for the concentration and purification of bacterial

pathogens from food samples (Cudjoe and Krona, 1997; Hsih and Tsen, 2001; Drysdale

et al., 2004; Lynch, et al., 2004) and are reportedly more sensitive than comparable

conventional culture methods (Cudjoe and Krona, 1997; Hsih and Tsen, 2001). Anti-

Salmonella, anti-E. coli 0157:H7, anti-Campylobacter, and anti-Listeria IMS beads are

commercially available (Matrix MicroScience, Cambridgeshire, UK; Dynal Biotech,

Oslo, Norway) however, anti-.\/ngel// IMS beads are not yet commercially available.

IMS has been used as a technique to concentrate .\/nge//t from clinical samples for

downstream detection processes such as EIA (Islam et al., 1993a) or PCR (Islam and

Lindberg, 1992; Achi etal., 1996; Achi-Berglund and Lindberg, 1996). By using IMS,

PCR inhibitors inherent to fecal samples were successfully eliminated. Briefly, IMS was

used to concentrate S. sonnei (Achi et al., 1996; Achi-Berglund and Lindberg, 1996), S.

flexneri (Islam and Lindberg, 1992; Achi et al., 1996), and S. dysenteriae serotype 1

(Islam and Lindberg, 1992; Achi et al., 1996) from feces prior to PCR. The detection

limit for the IMS-EIA and IMS-PCR methods were 1.0 x 103 CFU/ml (Islam et al.,

1993a) and 1.0-1.5 x 101 CFU/ml (Islam and Lindberg, 1992; Achi-Berglund and









Lindberg, 1996), respectively. In addition, the IMS-PCR method was more than two

times as effective then the conventional culture method for the diagnosis of shigellosis in

children with severe diarrhea (Achi et al., 1996). These studies have significance with

respect to ./ngel//t detection in foods, however further investigations are needed to

evaluate the suitability of the IMS techniques for use with food samples.

Immunomagnetic separation using the Pathatrix

Recently, a novel device allowing flow-through IMS, the Pathatrix, has been

developed (Figure 2-5). The Pathatrix allows a 250 ml sample to be continuously pumped

through a tubing system in which IMS beads are immobilized in a capture phase, thereby

allowing the analysis of a complete 25 g sample + 225 ml buffer/enrichment broth

homogenate with or without prior incubation. Each Pathatrix unit can analyze up to five

samples at one time. The incubation pots in which the sample stomacherr bag) is placed

are temperature controlled from room temperature up to 450C. After analysis, the tubing

system is disconnected from the sample and connected to a vessel containing wash buffer

which is then pumped over the capture phase to wash away any food particles or unbound

microorganisms. Once washing is complete the beads are recovered in buffer and the

captured microorganisms may be detected using selective plating, colorimetric assays or

molecular assays.

Arthur et al. (2005) compared the Pathatrix method in combination with selective

plating and PCR using the Lightcycler (Roche Applied Science, Indianapolis, IN) with

the BAX system (DuPont Qualicon, Wilmington, DE) and the Assurance GDS

(BioControl Systems, Inc., Seattle, WA) for detection ofE. coli 0157:H7 in ground beef.

When samples were inoculated at 17 CFU/65 g ground beef, all of the investigated

methods detected E. coli 0157:H7 in 57 of 57 samples. However, when samples were

















0




Figure 2-5. The Pathatrix system for flow-through immunomagnetic separation (IMS).
The Pathatrix base unit (A) can hold up to five samples at one time. The
capture phase (B) utilizes a magnet to immobilize the IMS beads within the
tubing system to allow binding to specific antigen as they pass through.

inoculated at 1.7 CFU/65 g ground beef, the Pathatrix in combination with selective

plating or PCR (4 hr pre-enrichment) detected E. coli 0157:H7 in 98% of samples

whereas the BAX system (8 hr pre-enrichment) and the Assurance GDS system (8 hr pre-

enrichment) detected E. coli 0157:H7 in 66% and 73% of samples, respectively (Arthur

et al., 2005).

Additional technologies for immunological detection of \/n/,,e/t

In a recent report, Sapsford et al. (2004) describe detection of S. dysenteriae in

buffer and on chicken carcasses using an array biosensor developed at the Naval

Research Laboratory. The array biosensor measures total internal reflection fluorescence

using a 25-min sandwich immunoassay for antigen detection. An advantage of the array

biosensor is that little or no sample preparation is required prior to analysis. The detection

limit of the array biosensor for S. dysenteriae in ground turkey, chicken carcass wash,

buffered milk, and a lettuce leaf wash was observed at 7.8 x 105 CFU/g, 4.9 x 104

CFU/ml, 7.8 x 105 CFU/ml, and 2.0 x 105 CFU/ml, respectively. When tested in buffer,

the array biosensor did not respond as efficiently to the other serogroups of.\,/ge//At and









cross-reacted with E. coli, suggesting the specificity of the polyclonal antibodies was a

limiting factor. The use of a monoclonal antibody with higher specificity may improve

the diagnostic ability of this testing format.

Molecular Microbiological Methods for ./lge,//At Detection in Foods

Polymerase chain reaction detection of .l/,,tie/la in foods

Polymerase chain reaction (PCR) methods for detection of .\l/ge//t in food have

previously demonstrated higher sensitivity than comparable culture methods (Warren,

2003). PCR assays for .\/igel/h spp. have targeted the invasion associated locus (ial)

(Islam and Lindberg, 1992; Lindqvist, 1999), the virA gene (Villalobo and Torres, 1998;

Vantarakis et al., 2000), or the ipaH gene (Sethabutr et al., 1993; Sethabutr et al., 2000;

Theron et al., 2001; Lampel and Orlandi, 2002; Warren, 2003) for amplification. The

same PCR primers are used to detect each of the serogroups of.\/ige//t and

enteroinvasive E. coli. The ial and virA genes are located on the large virulence plasmid

(sometimes referred to as the invasion plasmid), however sequencing of the S. flexneri

genome (Jin et al., 2002) revealed the ipaH gene to be encoded multiple times on both

the chromosome and the large virulence plasmid. Thus, detection of the ipaH gene is

possible in the event the large virulence plasmid has been lost, which has been shown to

occur when food samples are stored for a prolonged periods of time prior to analysis

(Lampel and Orlandi, 2002). The discussion below will be limited to PCR detection of

,\/ige/ll in food samples, however several reports of detection in fecal samples appear in

the literature. These methods are summarized along with other rapid methods in Table 2-

3.

Vantarakis et al. (2000) developed a multiplex PCR method to detect both ./nge//At

spp. and Salmonella spp. in mussels. Multiplex PCRs involve the use of two or more sets









of primers in the same PCR such that multiple targets can be amplified in the same

reaction. Artificially inoculated S. dysenteriae and S. Typhimurium were recovered by

homogenizing mussel meat with peptone water. DNA from an aliquot of the homogenate

was purified using a guanidine isothionate method and concentrated via ethanol

precipitation. The virA gene of ,/Nige//At spp. and the invA genes of Salmonella spp. were

amplified. Sample homogenates were inoculated at various concentrations to establish the

lowest detection limit for the method. When samples were not pre-enriched prior to

analysis, the multiplex PCR method was able to detect S. dysenteriae at 1.0 x 103 CFU/ml

in the homogenate. Following 22 hour incubation in buffered peptone water the multiplex

PCR methods was able to S. dysenteriae detect levels as low as 1.0 x 101 to 1.0 x 102

CFU/ml in the homogenate. Similar results were observed for S. Typhimurium.

Villalobo and Torres (1998) investigated PCR for the detection of S. dysenteriae

serotype 1 in mayonnaise. Samples were homogenized in buffered peptone water and

artificially contaminated with various concentrations of S. dysenteriae serotype 1.

Bacterial cells were lysed with detergent, the DNA extracted with phenol-chloroform and

precipitated with ethanol. A multiplex PCR was then used to amplify regions from the

virA gene and the 16S rRNA gene. The lowest level of inoculation detected by this

method was 1.0 x 102 to 1.0 x 103 CFU/ml in the homogenate.

In a study by Lindqvist (1999), nested PCR was investigated for the detection of

the ial locus of.\,//ge//At spp. from spiked lettuce, shrimp, milk, and blue cheese samples.

Nested PCR involves the use of two sequential PCRs in which the target sequence in the

second PCR lies within the amplified sequence in the first PCR. Nested PCRs can be

used to achieve extreme low sensitivities. Food samples inoculated were homogenized in









physiological saline and bacteria were isolated by buoyant density centrifugation

(separation of components based on density). Single PCR, using only the external primer

sets, was able to detect S. flexneri in aqueous solution at 0.5-1.0 x 105 CFU/ml. Nested

PCR was more sensitive and was able to detect 1.0 x 103 CFU/ml. The nested PCR assay

in combination with buoyant density centrifugation was able to detect S. flexneri

inoculated onto all four foods at 1.0 x 101 CFU/g (Lindqvist, 1999).

Theron et al. (2001) investigated a semi-nested PCR for the detection of the ipaH

gene of S. flexneri in spiked environmental water samples. The detection limits in the

various environmental water samples were 2.0 x 103 CFU/ml for well water, 1.4 x 101

CFU/ml for lake water, 5.8 x 102 CFU/ml for river water, 6.1 x 102 CFU/ml for treated

sewage water, and 1.1 x 101 CFU/ml for tap water. Variability in results among the water

samples was attributed to the presence of humic substances that inhibited PCR. Humic

substances are transformed products from soil organic matter that do not belong to the

known classes of biochemistry. These include humic acids, fulvic acids, and humins. Pre-

enrichment in GN broth served to dilute humic substances while allowing the S. flexneri

to multiply, thereby increasing the concentration of target DNA.

Improved PCR detection of .\/ngll, by FTA filtration

FTA filters (Whatman, Clifton, NJ) have been used in template preparation for

PCR detection of pathogenic microorganisms (Lampel et al., 2000; Orlandi and Lampel,

2000; Lampel and Orlandi, 2002; Warren, 2003; Warren et al., 2005b). Moisture in the

sample activates chemicals in the FTA filter that lyse cells, denature enzymes, inactivate

pathogens, and immobilize genomic DNA (Whatman website, 2004). Chemicals in the

FTA filters also protect DNA and RNA from light, free radicals, enzymes, and pathogens

during dry, room-temperature storage. FTA filtration as sample preparation for PCR has









been developed for detection of Cyclospora and Cryptosporidium in water (FDA, 1998).

Recently, this sample preparation technique has been investigated for detection of S.

boydii and S. sonnei from tomato rinses by nested PCR (Warren, 2003; Warren et al.,

2005b). Briefly, tomato rinses were passed through a two-stage filter where the first stage

contained filter paper for size exclusion and the second stage contained FTA filter paper.

After purification and washing, 6-mm punches are taken from the FTA filters and used

directly as template in the first step of the nested PCR. The FTA filtration/nested PCR

(FTA-PCR) assay detected S. boydii and S. sonnei at 6.2 x 100 CFU/tomato and 7.4 x 100

CFU/tomato, respectively (Warren, 2003; Warren et al., 2005b).

Although FTA-PCR had excellent sensitivity when used to analyze tomato rinses,

which are relatively clean, similar results were not observed when recovery of inoculated

S. boydii and S. sonnei was attempted from cantaloupes, strawberries, or retail Valencia

oranges (Schneider and Warren, unpublished data). FTA-PCR was unable to detect

inoculated ./nlge/lt from strawberries, suggesting that the purification and washing

protocol used on the FTA filters may not be adequate to remove PCR inhibitors (Warren,

2003). Despite several size exclusion filtering techniques applied prior to FTA filtration,

the FTA filters were routinely clogged by cantaloupe fibers or the wax coating from retail

oranges, which entered the buffer rinse during recovery procedures (Warren, 2003). The

advantage of the FTA-PCR technique is that no enrichment was necessary to obtain low

detection limits due to the large volume of sample analyzed by the filters. The technique

warrants further investigation for the analysis of foods for .\/nge/l/ contamination as

improvements in pre-filtration techniques could improve the sensitivity.









Additional molecular microbiological techniques for detection of.\/ige//t

An interesting new method, immunocapture universal primer PCR (iUPPCR) has

been recently reported (Peng et al., 2002). Universal primer PCR employs primers

designed against highly conserved regions, such as 16S rRNA genes, thus the resulting

PCR can be used for amplification of almost any bacteria. Typically, the sequence of the

resulting amplicon would be further analyzed for identification of the bacteria. In

iUPPCR however, immunocapture is employed for specificity and universal primer PCR

is used for detection of captured bacteria. Monoclonal antibodies specific for individual

serotypes of S. dysenteriae, S. flexneri, S. boydii, or S. sonnei were coated into the wells

of a 96-well polystyrene plate. Cross-reactivity tests demonstrated the specificity of the

monoclonal antibodies to the strain level. Wells were challenged with bacteria and

captured bacteria were detected using universal primer PCR, i.e. PCR with primers

specific for 16S rRNA sequences conserved among closely related bacteria. The

detection limit for shigellae in broth was 5.0 x 102 CFU/ml (Peng et al., 2002).

Presumably, detection of up to 96 pathogens sharing the conserved 16S rRNA sequence

could be accomplished in the same plate using specific monoclonal antibodies for

bacterial capture (differentiation of cell types) and the universal primer PCR for

detection. Further investigation of this method with respect to food samples as opposed to

bacteria in broth is required.

More recently, Ji et al. (2006) reported iUPPCR in combination with denaturing

gradient gel electrophoresis (DGGE) for the detection and identification of.\/ige//t spp.

In contrast to the iUPPCR method described above, genus-specific polyvalent

monoclonal antibodies were used to coat the wells of a 96-well plate, such that all

\/nge//lt spp. may be captured in each well. Following iUPPCR to amplify 16S rRNA









gene fragments, DGGE was used to identify the specific serotype. Unfortunately, the

iUPPCR-DGGE method was only tested against pure cultures in laboratory conditions.

Although the authors speculate that it would be useful for the detection and identification

of \/nge//At in food and environmental samples, further testing in the presence of

food/environmental matrices and indigenous microflora are required.

Advances in oligonucleotide microarrays have further enabled the simultaneous

detection and identification of bacterial foodborne pathogens. Oligonucleotide

microarrays involve the ordered immobilization of specific probes to a solid surface

followed by hybridization with labeled sample DNA. After hybridization, development of

the label can allow identification/enumeration of complementary sample DNA sequences.

In a recent report, Kakinuma et al. (2003) describes an oligonucleotide microarray using

probes specific for the gyrB gene to detect and differentiate E. coli, .\/n1,//i, and

Salmonella. PCR amplified regions of the gyrB gene were fluorescently labeled and

hybridized to detection probes immobilized on glass slides. Based on reaction patterns,

three species of .\/ige/lu, seven serovars of Salmonella, and one strain of E. coli were

correctly identified at the species level. Identification at the subspecies level of

Salmonella was problematic when multiple serovars were present in the same sample due

to the overlap of microarray patterns. Assay formats such as this could be expanded to

potentially detect and identify a wide range of enteric pathogens.

Song et al. (2005) reported an alternative DNA amplification method, loop-

mediated isothermal amplification (LAMP), for the detection of.\l/i/,ge/ spp. in clinical

samples. In the LAMP assay, four specialized primers designed to specifically recognize

six distinct regions on the target gene are used in combination with a DNA polymerase









with strand displacement activity. The amplification reaction occurs at 650C, therefore no

thermocycler is necessary as with PCR. For a more complete description of the LAMP

assay, the reader is directed to the Eiken Chemical Co., Ltd. (Tokyo, Japan) website:

http://loopamp.eiken.co.jp/e/lamp/index.html (Accessed 07 Jun 2006). Stool samples

homogenized in sterile water were inoculated with various titers of S. flexneri YSH6000

and the DNA was extracted using a boiling method. The sensitivity of the LAMP assay

for S. flexneri in the inoculated stool samples was eight CFU per reaction whereas the

sensitivity of a PCR reaction was determined to be 800 CFU per reaction. A distinct

advantage of the LAMP assay is that amplified DNA may be visualized by the naked eye

as turbidity in the sample, eliminating the need for post-amplification processes, such as

gel electrophoresis.

Finally, repetitive sequence-based PCR (rep-PCR) has been demonstrated for the

identification and molecular typing of members of the family Enterobacteriaceae,

including .\/nlg//I (Raza et al., 2003; Lising et al., 2004). In rep-PCR, primers are

designed complementary to bacterial interspersed repetitive sequences and the regions

lying outward of the primers are amplified resulting in DNA fragments of varying

lengths. These fragments can then be separated by electrophoresis to form a bacterial

fingerprint unique to individual bacterial strains. The DiversiLab System (Spectral

Genomics, Inc., Houston, TX) generates such fragments by rep-PCR that are then

analyzed using microfuidics lab-on-a-chip and the Agilent 2100 Bioanalyzer (Agilent

Technologies, Inc. Palo Alto, CA). Results can be visualized in several formats including

dendograms, electropherograms, gel-like image, or scatter plots (Lising et al., 2004).

Rep-PCR technology has been shown to be reproducible and can allow the differentiation









of species, subspecies and strains in as little as 4 hours. Some .\l/ngel// serogroups show

>95% similarity with the DiversiLab Enteric Kit Beta version, however the DiversiLab

.\/ige/ll kit offers greater differentiation among serogroups and strains (Lising et al.,

2004). Strains of E. coli (non-pathogenic, enterohemorragic, and enterotoxigenic) were

effectively differentiated from .\illgel/, spp. using rep-PCR (Raza et al., 2003; Lising et

al., 2004), however no enteroinvasive E. coli strains were included in the studies.

Detection of Pathogens by Sequence Capture

Sequence capture methods have been investigated for the detection of a wide

variety of bacterial, viral, and fungal pathogens. Typically, a specific oligonucleotide

probe is immobilized on a solid surface and DNA/RNA prepared from the target

microorganism is hybridized to the probe. Zammatteo et al. (1997) compared the

sequence capture of human cytomegalovirus using DNA probes immobilized on 96-well

microtiter plates and paramagnetic beads and found that faster hybridization kinetics

were obtained with the use of beads as the solid support. Although Zammatteo et al.

(1997) used DNA probes covalently bound to the solid supports, it is more common to

use oligonucleotide probes labeled with biotin on their 5' end with paramagnetic beads or

microtiter plates with wells pre-coated with streptavidin. As with more traditional nucleic

acid hybridization, variables such as salt concentration, temperature, contact time and GC

content of the probe sequence all have influence on hybridization efficiency. Del Gallo et

al. (2005) investigated steric factors affecting the hybridization of PCR amplified

sequences to DNA probes immobilized on the screen-printed gold surface of disposable

electrodes. The amount of probe coverage on the surface as well as the relative position

of the probe on the target sequence was found to partially control hybridization

efficiency. When the probe coverage was approx. 2.9 x 1012 molecules/cm2 and when the









probe sequence was located at one of the termini of the target sequence the efficiency of

hybridization was increased (Del Gallo et al., 2005). In addition, the use of spacer

molecules between the solid support and the probe sequence has been shown to

drastically improve hybridization efficiency (Shchepinov et al., 1997; Amagliani et al.,

2006). The method used to purify the probe during synthesis may also affect

hybridization efficiency, with HPLC purification preferred over desalting (Amagliani et

al., 2006). Finally, sequence capture methods may be used to prepare DNA/RNA from

samples that contain PCR inhibitors, since these inhibitors are removed during post-

hybridization washing (Maher et al., 2001; Tsai et al., 2003).

While there are numerous reports of sequence capture for the analysis of clinical

samples for bacterial and viral pathogens, this discussion is limited to the few reports of

sequence capture used for the analysis of foods or environmental samples. In a study by

Chen et al. (1998), sequence capture in combination with PCR was investigated for the

detection of verotoxigenic E. coli (VTEC) in brain heart infusion (BHI) broth cultures

and in artificially contaminated ground beef. Biotin-labeled probes were used to form

hybrids with specific DNA segments and then the hybrids were bound using streptavidin-

coated paramagnetic beads. In BHI broth, detection of initial VTEC concentrations of

103, 102 and 100 CFU/ml was achieved after 5, 7 and 10 hr enrichment at 370C,

respectively. In ground beef samples, the sequence capture method was able to detect

VTEC at levels of 100 CFU/g after 15 hr incubation in BHI broth. For both BHI broth and

ground beef samples, DNA was extracted from one ml aliquots of the enrichments by the

boiling method and the prepared DNA was used as template in the sequence capture

method. In a subsequent study, Chen and Griffiths (2001) modified this sequence capture









method for the simultaneous detection of Salmonella and Shiga-like toxin-producing E.

coli. Again using cultures in BHI broth, the detection limit was determined to be 100

CFU/ml after 10 hr incubation at 370C (Chen and Griffiths, 2001). This study

demonstrates that multiple probes, each specific for a different pathogen, may be used in

combination on solid supports in order to test for different pathogens at the same time.

Tsai et al. (2003) investigated sequence capture in combination with PCR for the

detection of enterotoxigenic E. coli (ETEC) associated with cattle in environmental water

samples. Biotin-labeled probes with specificity for the enterotoxin gene LTIIa were

attached to streptavidin-coated paramagnetic beads and used to form hybrids with

prepared DNA samples. The detection limit of the assay was determined to be 2.5

attogram/tl DNA. In addition, some of the extracted DNA samples were spiked with

humic acids, known inhibitors of PCR, to determine if the sequence capture method was

effective at removing these inhibitors prior to PCR. In the presence of humic acids, the

sequence capture method increased sensitivity 10,000-fold over conventional PCR (Tsai

et al., 2003).

Amagliani et al. (2006) developed sequence capture in combination with PCR for

the detection ofListeria monocytogenes in milk. Using NH2-labeled DNA probes specific

for the hlyA gene coupled to paramagnetic beads, the detection limit of the assay was 101

CFU/ml. The sensitivity was achieved using (CH2)12 spacers between the NH2-label and

the DNA probe sequence. In contrast to many previously reported sequence capture

methods, the method of Amagliani et al. (2006) required no DNA extraction/purification

as capture was performed directly in the milk sample. For this reason a two-step sequence

capture in which biotin-labeled probes formed hybrids with target sequences followed by






49


isolation of the hybrids using streptavidin-coated paramagnetic beads was not successful.

It was hypothesized that indigenous biotin in the milk samples interfered with the biotin-

labeled hybrids binding to the streptavidin-coated beads (Amagliani et al., 2006).














CHAPTER 3
MATERIALS AND METHODS

Preliminary Studies

Acquisition and Maintenance of./,\//,./ sonnei Cultures

The following cultures were purchased from the American Type Culture

Collection (ATCC; Manassas, VA): S. sonnei ATCC 9290, S. sonnei ATCC 29031, S.

sonnei ATCC 29030, S. sonnei ATCC 25931 and S. sonnei ATCC 29930. Each strain

was resuscitated as instructed in the package inserts and the resulting growth was

streaked for isolation on MAC plates. One typical S. sonnei colony from each plate was

transferred to a TSA slant and stored at 40C. A second typical S. sonnei colony was

transferred per product instructions to Protect TM Bacterial Preservers (Scientific Device

Laboratory, Inc., Des Plaines, IL) and stored at -700C.

Adaptation of Cultures to Rifampicin

Subcultures of each of the five S. sonnei strains were adapted to the bactericidal

agent rifampicin by spontaneous mutation. A 10,000 tg/ml (1%) stock solution of

rifampicin was prepared by dissolving 2.0 g rifampicin (Fisherbrand, Fisher Scientific,

Pittsburg, PA) in 200 ml methanol. The stock solution was then filter sterilized and stored

in the dark at 40C.

Stock cultures were grown overnight in 10 ml TSB (370C, 30 rpm). Overnight

cultures were transferred to 10 ml TSB supplemented with 2.5 tg/ml rifampicin and

grown overnight (370C, 30 rpm). With each transfer, the concentration of rifampicin

increased until the final concentration was 200 [tg/ml rifampicin. Once the cultures were









adapted to 200 pg/ml rifampicin, cultures were grown overnight (37C, 30 rpm) three

consecutive times in TSB supplemented with 200 pg/ml rifampicin to ensure well

adapted populations.

Once adaptation was complete, the final overnight culture was streaked for

isolation on MAC and incubated overnight at 370C. One typical colony from the

overnight MAC plate was transferred to a TSA slant supplemented with 200 pg/ml

rifampicin and stored at 40C. Another typical colony was transferred per manufacturer's

instructions to a ProtectTM Bacterial Preserver and stored at -760C.

Preparation of Microbiological Media

Wild-type cultures were grown in Tryptic Soy Broth (TSB; BD Diagnostics,

Franklin Lakes, NJ) containing 10 [M Congo red (TSCR) and maintained on Tryptic Soy

Agar (TSA; BD Diagnostics) slants containing 10 [M Congo red. For survival studies

only, bacterial strains were adapted to 200 [g/ml of the bactericidal agent rifampicin and

experiments were conducted in TSB supplemented with 100 [g/ml rifampicin (TSB rif+).

All dilutions, unless otherwise specified, were performed using 0.1% Peptone (BD

Diagnostics) water. Phosphate Buffered Saline (PBS; pH 7.4) was prepared using the

following formula per liter: 1.2 g NaHPO4, 8.2 g Na2PO4 and 5.0 g NaC1.

\l/ngel// Broth (SB) was prepared according to the U.S. Food and Drug

Administration's (1998) Bacteriological AnalyticalManual (BAM) and supplemented

with novobiocin (ICN Biomedicals Inc., Aurora, OH) at 3.0 ig/ml. MacConkey Agar

(MAC; BD Diagnostics), Triple Sugar Iron Agar (TSI; BD Diagnostics), Lysine Iron

Agar (LIA; BD Diagnostics) and Motility Medium (MM; BD Diagnostics) were all

prepared according to manufacturers instructions. TSI and LIA were prepared as slants in









screw cap tubes. MM was supplemented with 0.005% triphenyltetrazolium chloride

(TTC; BD Diagnostics). When necessary, media pH was adjusted using filter sterilized

IN NaOH.

Acquisition/Preparation of Food Matrices

Mature green, unwashed, unwaxed tomatoes (Florida 47 cultivar) were obtained

from a nearby packinghouse. Tomatoes were held at 40C prior to use. 'Picnic Potato and

Egg Salad' (hereafter referred to as potato salad) was prepared the night before each

experiment using a publicly available recipe (http://southernfood.about.com/od/

potatosalads/r/b100624c.htm?terms=picnic+potato+and+egg+salad ). Briefly, 6.0 lbs

potatoes were skinned, cubed, boiled for 15 min, and then cooled. Eight large eggs were

hard boiled for 15 min, cooled, and chopped. After straining the water from the cooked

potatoes, the chopped eggs, 12 cup chopped red onion, 1 cup chopped fresh celery, 1 and

12 cup mayonnaise (Hellman's Lite, Unilever, Englewood Cliffs, NJ), 3 tbs yellow

mustard (Publix Supermarket brand, Lakeland, FL), 1 tsp salt, and 1 tsp black pepper

were added and mixed well with a large serving spoon. The potato salad was stored at

4C until use. Ground sirloin (90% lean/10% fat; hereafter referred to as ground beef)

was purchased at a local grocer the morning of each experiment.

Acquisition and Maintenance of Anti-.,\/ntie/t Antibodies

All polyclonal and monoclonal antibodies used in the following experiments are

commercially available and are listed in Table 3-1. Anti-.\/i/gel// polyclonal antibodies

AB01 and AB04 were purchased from Virostat, Inc. (catalog number 0901; Portland,

ME) and AbCam, Inc. (catalog number Ab19988; Cambridge, MA), respectively. Goat

anti-rabbit IgG (H+L chain specific) antibodies (AB03) were purchased from Southern

Biotech Associates, Inc. (catalog number 4050-01; Birmingham, AL). Anti-S. sonnei









monoclonal antibodies (AB02) were purchased from Novus Biologicals (catalog number

BM1316; Littleton, CO). Goat anti-mouse IgG (H+L chain specific) antibodies (AB05)

were purchased from Southern Biotech Associates, Inc. (catalog number 1031-01). Upon

receipt, all antibodies were stored at 4.00C if they were to be used within two months

otherwise aliquots of 100 [l were stored frozen at -200C.

Binding of Antibodies to Paramagnetic Beads

The binding of antibodies to MagaCell beads was performed according to Cortex

Biochem, Protocol 503 (available at http://www.cortex-biochem.com/commerce/ccc 1010

-protocols.htm) except that all volumes were reduced proportionately for a starting

volume of 500 [l beads. The MagaCell beads were mixed by inverting the bottle by hand

until no beads were visible as a pellet on the bottom of the bottle. A 500-[l aliquot of the

beads were transferred to a clean, sterile 1.5 ml microcentrifuge tube. All washing steps

were performed using a magnetic particle concentrator (MPC-S; Dynal Biotech) or a

magnetic separator (CD3002; Cortex BioChem) to draw the beads to the side of the tube

allowing the supernatant to be removed using a pipette. The beads were washed two

times in 1000 [l de-ionized water, followed by four times in 1000 [l acetone. The beads

were then resuspended in 250 [l acetone containing 0.12 g 1,1 carbonyldiimidazole per

10 ml. The microcentrifuge tubes were placed on a rugged rotator and mixed by gentle

end-over-end rotation for 1 hr. The beads were then washed four times in 1000 [l

acetone, four times in 1000 [l de-ionized water and four times in 1000 [l 0.1M sodium

bicarbonate buffer, pH 8.6. The beads were then resuspended in 400 pl 0.1M sodium

bicarbonate buffer, pH 8.6, an appropriate amount of antibody solution (Table 3-1) was

added and the volume was adjusted to 500 pl using 0.1M sodium bicarbonate buffer, pH









Table 3-1. Antibodies investigated for immunocapture of S. sonnei.
Volume added
per 500 pl
Designation Clonality Immunogen MagaCell beads
AB01 Polyclonal Mixture of S. boydii, S. flexneri, S. 50
dysenteriae
AB02 Monoclonal S. sonnei, NCTC 9774 100
(Wheeler phase I)
AB03 Polyclonal Pooled antisera from goats 100
hyperimmunized with normal rabbit
IgG
AB04 Polyclonal Membrane extract of S. sonnei and 50
S. flexneri
AB05 Polyclonal Pooled antisera from goats 100
hyperimmunized with mouse IgG
paraproteins

8.6, when necessary. The beads were again placed on a rugged rotator and mixed with

gentle end-over-end rotation for 18-24 hr at room temperature. The next day, the beads

were washed two times in 500 pl 0.1M sodium bicarbonate buffer, pH 8.6. The beads

were then resuspended in 500 pl 0.1M sodium bicarbonate buffer, pH 8.6, containing

ethanolamine (3 ml/liter) and mixed by gentle end-over-end rotation for 1 hr. The beads

were then washed once in 500 pl 0.1M sodium acetate buffer, pH 4.0, resuspended in 500

pl 0.1M sodium acetate buffer, pH 4.0, and mixed by gentle end-over-end rotation for 1

hr. Finally the beads were washed three times in 500 pl PBS containing 0.1% sodium

azide and stored at 40C until used.

Evaluation and Optimization of Immunocapture Using Anti-./gel/h, i Beads

Preliminary challenges of anti-.\l/ge//At beads for immunocapture of S. sonnei

strains were performed in PBS. A 100-tl aliquot of an overnight S. sonnei culture in

TSCR was used to inoculate sterile stomacher bags containing 250 ml PBS such that the

final cell titer was approximately 2.0 x 105 CFU/ml. The stomacher bags were then

placed into Pathatrix incubation pots and the samples were circulated for 30 min at 370C









using 25 .il aliquots of anti-.\/ige//l, beads for immunocapture. Anti-.s/lgel//A beads

prepared using various anti-.\/igell//A antibodies and various concentrations of anti-

,/Ngel//t beads were investigated for FTI. After circulation, each sample was washed in

PBS and the beads were recovered in approximately 250 tl PBS. A 100-rl aliquot of the

recovered beads was analyzed by spread plate using MAC. MAC plates were incubated at

370C for 24 hr and the resulting colonies were counted. Antibodies AB01, AB02 and

AB04 were bound directly to MagaCell beads and tested for primary capture in FTI. In

addition, AB01 and AB02 were tested for immunocapture of S. sonnei in broth culture

followed by secondary capture using MagaCell beads coated with antibodies AB03 and

AB05, respectively.

Crude DNA Extraction from Bacteria by Boiling

Stock cultures frozen on Protect TM Bacterial Preservers were retrieved from frozen

(-700C) storage and allowed to thaw. One bead was aseptically transferred from the

Protect TM Bacterial Preserver into 10 ml TSB and grown overnight (37C). Overnight

cultures were plated for isolation on an appropriate selective and differential medium and

incubated overnight at 370C. Overnight plates were observed for typical colony

morphologies. One typical colony was transferred to 10 ml TSB and grown overnight

(37C).

A 1.0-ml aliquot of the overnight culture was transferred to a clean, sterile 1.5 ml

microcentrifuge tube and the bacterial cells were harvested by centrifugation (3,220 x g

for 10 min). The resulting supernatant was discarded and the pellet was re-suspended in

200 Cl de-ionized, sterilized water. Samples were then boiled for 10 min in a dry bath

incubator (Fisher Scientific, IsoTemp 125D). The supernatant (DNA template) was









aseptically transferred to a clean, sterile 1.5 ml microcentrifuge tube and stored at -200C.

The pellet was discarded.

DNA Extraction from Anti-.\/ltie/lt Beads using the DNeasy Kit

Extraction of DNA from bacterial cells captured on anti-.\/lge//lt

immunomagnetic beads was performed using a modified protocol with DNeasy spin

columns (Qiagen, Valencia, CA). A 100-tl aliquot of concentrated anti-.\/lge//At beads

was transferred to a clean, sterile 1.5 ml microcentrifuge tube and heated on a dry bath

incubator at 950C for 10 min to lyse bacterial cells. After incubation, 100 tl sterile water,

200 tl buffer AL (Qiagen) and 200 pl absolute ethanol was added to the sample. The

resulting mixture was vortexed for 5 sec and placed in a MPC-S for at least 1 min to draw

the magnetic particles to the side of the tube. Without disturbing the magnetic particles,

the mixture was transferred to a DNeasy spin column and centrifuged at 6,000 x g for 1

min. The DNeasy spin column was transferred to a clean collection tube and 500 pl

buffer AW1 (Qiagen) was passed through the column by centrifugation at 6,000 x g for 1

min. The DNeasy spin column was transferred to a clean collection tube and 500 pl

buffer AW2 (Qiagen) was passed through the column by centrifugation at 16,000 x g for

3 min. The DNeasy spin column was transferred to a clean, sterile 1.5 ml microcentrifuge

tube and 100 pl buffer AE (Qiagen) was added to elute bacterial DNA by centrifugation

at 6,000 x g for 1 min. The DNeasy spin column was discarded and the eluted DNA was

placed on ice or stored at -200C prior to analysis by real-time PCR.

Preparation of HeLa Cell Extracts

Fresh HeLa cells were obtained from the laboratory of Dr. F. Southwich,

University of Florida, resuspended in 1.0% Triton X-100 supplemented with Complete

Mini Protease Inhibitor cocktail tablets (Roche Applied Sciences, Indianapolis, IN).









HeLa cells were lysed using three cycles of freeze-thaw. One hundred [tl aliquots of the

lysed cells were transferred to clean, sterile 1.5 ml microcentrifuge tubes and proteins

were extracted using a modified method of Wessel and Fliugge (1984). Briefly, the

samples were vortexed and centrifuged at 5,000 x g for 2 min and the sample supernatant

was transferred to a clean, sterile 1.5 ml centrifuge tube and saved as 'HeLa cell extract

1'. The remaining pellet was resuspended in 400 ptl methanol (Fisherbrand), vortexed and

centrifuged at 9,000 x g for 10 sec. To the solution, 200 ptl chloroform (Fisherbrand) was

added and the resulting solution was vortexed and centrifuged at 9,000 x g for 10 sec. To

the solution, 300 ptl water was added and the resulting solution was vortexed vigorously

and centrifuged at 9,000 x g for 1 min. The upper phase in the resulting supernatant was

transferred to a clean, sterile 1.5 ml centrifuge tube and saved as 'HeLa cell extract 2'. To

the lower phase, 300 ptl methanol was added and the solution was vortexed and

centrifuged at 9,000 x g for 2 min. The resulting supernatant was transferred to a clean,

sterile 1.5 ml centrifuge tube and saved as 'HeLa cell extract 3'. The pellet was dried

using forced air (generated using a transfer pipette) and saved as 'HeLa cell extract 4'.

RNA Extraction Using the RNeasy Kit

Extraction of RNA from bacterial cells was performed using RNeasy spin columns

(Qiagen). A 250-1l aliquot of bacterial cells were transferred to clean, sterile 1.5 ml

microcentrifuge tubes containing 500 ptl RNAbacterial Protect Reagent (Qiagen),

incubated at room temperature for 5 min and centrifuged at 5,000 x g for 10 min. The

resulting supernatant was discarded and pellet was resuspended in 100 ptl TE buffer, pH

8.0, containing 1 mg/ml lysozyme (Sigma, St. Louis, MO) and incubated at room

temperature for 5 min. To the mixtures, 350 [tl buffer RLP (Qiagen), containing 10 itl/ml

2-mercaptoethanol (Sigma), were added and the solutions was vortexed vigorously for 5









sec. To the mixture, 200 [l absolute ethanol was added and the solution was mixed by

gentle action of the pipette. The resulting mixture (approx. 700 .il) was transferred to

clean, sterile RNeasy spin columns and centrifuged at 8,000 x g for 15 sec. In a fume

hood, the flow-through was discarded, 700 pl buffer RW1 (Qiagen) was added to the

RNeasy spin column and the sample was centrifuged at 8,000 x g for 15 sec. In the fume

hood, the collection vessel and flow-through were discarded and the RNeasy spin column

was transferred to a clean, sterile collection vessel. A 500 .l aliquot of buffer RPE

(Qiagen) was added to the RNeasy spin column, the sample was centrifuged at 8,000 x g

for 15 sec and the flow-through was discarded. A second 500 Cl aliquot of buffer RPE

(Qiagen) was added to the RNeasy spin column, the sample was centrifuged at 8,000 x g

for 2 min and the collection vessel and flow-through were discarded. The RNeasy spin

column was transferred to a clean, sterile 1.5 ml microcentrifuge tube and 50 Cl water

was added directly onto the RNeasy silica-gel membrane to elute RNA. The RNeasy spin

column inside of the 1.5 ml microcentrifuge tube was centrifuged at 8,000 x g for 2 min

and the RNeasy spin column was discarded. The collected RNA sample was placed on

ice or stored at -700C prior to further analysis.

DNase Treatment of RNA Extracts Prior to RT-PCR

DNase treatment of collected RNA samples was performed using the DNase I

Amplification Grade Kit (Invitrogen, Carlsbad, CA). To a clean, sterile 0.5 ml

microcentrifuge tube, 1 \l 10X DNase I Reaction Buffer (Invitrogen), 2 .l DNase I Amp

Grade (1 U/pl; Invitrogen) and 2 .l DEPC-treated water (ICN Biomedicals Inc.) was

added. A 5 Cl aliquot of RNA was added and the solution was mixed using the action of

the pipette and incubated for 15 min at room temperature. After incubation the reaction

was stopped by addition of 1 pl 25 mM EDTA Solution (Invitrogen) and the mixture was









heated at 650C for 10 min. DNase I-treated RNA samples were placed on ice prior to

further analysis by RT-PCR.

Induction and Expression of ipaH RNA in S. sonnei

A five-strain S. sonnei cocktail was prepared as described below in the section

'Inoculum Preparation'. The following test solutions were prepared in clean, sterile 1.5

ml microcentrifuge tubes: 900 Cl SB + 100 [l HeLa cell extract 1, 900 pl SB + 100 pl

HeLa cell extract 2, 900 pl SB + 100 pl HeLa cell extract 3, 900 pl SB + 100 pl HeLa

cell extract 4, 900 pl SB + 100 [l Congo red solution (see Appendix A) and 1000 pl SB

(control). A 10--l aliquot of the S. sonnei cocktail was added to each of the test solutions

and the solutions were vortexed. Total RNA was extracted from a 250-pl aliquot of each

test solution and analyzed by RT-PCR using ipaH gene-specific primers. The remaining

portions of the test solutions were incubated at 370C for 4 hr, after which the total RNA

was extracted from a second 250-[l aliquot of each test solution and analyzed by RT-

PCR using ipaH gene-specific primers.

Identification of '.\/,,e/hl i-Specific Genetic Loci

Homologous gene cluster tables were created using the Microbial Genome

Database for Comparative Analysis (MBGD; National Institute for Basic Biology,

National Institutes of Natural Sciences, Japan; available at: http://mbgd.genome.ad.jp/)

using the available genomes (as of December 2005) for Escherichia coli and .\//ge/ht

spp. (E. coli 0157:H7, E. coli CFT073, E. coli K-12 W3110, E. coli K-12 MG1655, E.

coli EDL933, S. flexneri 2a 2457T, S. flexneri 2a 301 and S. sonnei Ss046). Homologous

genes were analyzed using ClustalW (protein-protein alignment function included in the

MBGD) to identify potential genes for the detection of,//ige//t spp. or S. sonnei.

Specifically, those genes that contained sequences conserved among ,.\/igell/ spp. or









sequences unique to S. sonnei were identified as potential genetic targets. The nucleotide

sequences were obtained for the identified potential genetic targets and analyzed using

the nucleotide-nucleotide Basic Local Alignment Search Tool (BLASTn; available at the

National Center for Biotechnology Information website: http://www.ncbi.nlm.nih.gov/)

to further identify sequences specific to lgellt spp. or S. sonnei. The identified

sequences were then used for primer/probe development as described below.

Development of Primers/Probes for the Detection of,\ltige/h,

The primers/probes investigated in this study are listed in Table 3-2. All

primers/probes were developed using the Beacon Designer 5 software (PREMIER

Biosoft International, Palo Alto, CA). For design of the 01-023 primers and probe, the

conserved sequence from the chromosomally-located ipaH genes of S. sonnei Ss046 was

identified using ClustalW. Default primer settings were used with one exception; the 3'

maximum AG was adjusted from 10 -kcal/mol to 4.0 -kcal/mol. For primers designed for

the detection of all ./nge//At spp., regions with cross homology to the E. coli K-12

genome were avoided. For primers designed for the specific detection ofS. sonnei,

regions with cross homology to the S. flexneri 2a 2457T genome were avoided. All

hydrolysis probes were designed with the reporter dye FAM on the 5' end and the

quenching dye TAMRA on the 3' end. All primers/probes were purchased from Sigma-

Genosys (The Woodlands, TX).

Evaluation of Primer/Probe Specificity

All primers and probes were evaluated for specificity in silico using BLASTn in

addition to in-house testing against DNA from stock bacterial cultures (Table 3-3).

Primer specificity was initially tested using real-time PCR followed by melt curve









Table 3-2. Primers designed for the detection of. /nge//I All primers were designed
using Beacon Designer 5 software with gene sequences obtained from the
genome of S. sonnei Ss046 (accession number CP000038).


Designation
01-023F
01-023R
01-023P


01-024F
01-024R
01-024P

01-025F
01-025R

01-026F
01-026R

01-027F
01-027R

01-028F
01-028R

01-029F
01-029R

01-030F
01-030R

01-031F
01-031R

01-032F
01-032R


Gene ID
ipaH


SSO 0670




SSO 2067



SSO 2071



SSO 1019



SSO 2059



SSO 2863



SSO 2685



SSO 3247



SSO 0721


Sequence (5' -* 3')
GTGAAGGAAATGCGTTTCTATG


ACCAGTCCGTAAATTCATTCTC
AGTGACAGCAAATGACCTCCGCA

TTTCTAAAGTTGCAGTCACCTTTG
GGTGCCTAAAACGATATTGCTTTG
AGCTCAGCGAAAACCACCGGCG

GCCCCGCTACGCATGTC
GTGATCTCCAGTTCCGCAAATG

ACAATCGAAGACATCGCGTTTC
CCAATCACTTTCTTGCCACTTTTC

ACGCTTACAAGGCCATTATGAATG
CCTCAGCTTCAGATGCTTTATCAC

AAACCACTCATCAAATACGAGAG
TTCGCAATGACCAGACCTAC

CGGCTGGTTTGGCAAGTTAAG
TGGTTCACCCCATCAAGAACATC

TGAGCCCGGACAGTTTCAC
TTGTATGTTACGTCGCTGAACAC

CCCAACCATATTGACGTGTTCTTC
GCGTAGTTGCTGCCGTTAAC

TTTATGACAGTTGCTGATTTCAAAC
ATACTCTTTCTGAGGATGAATGTTC


analysis in 20 tl reaction mixtures consisting of: 10 pl IQTM Supermix with SYBR green

(Bio-Rad, Hercules, CA), 200 nM (final concentration) each of forward and reverse

primer, 2.0 ll DNA sample and purified water. The PCR cycling conditions were 95C


Product
length (bp)
106









Table 3-3. .\ lgel/lt and non-.\/hlge/l strains tested for specificity.
Designation Culture Origin
KRS101 .\iige/ll boydii serotype 18 ATCC 35966
KRS 102 .higell// boydii serotype 18 Outbreak isolate
KRS 103 liilge/ll sonnei Patient isolate
KRS 104 \lrnge/ll sonnei Patient isolate
KRS 105 \/lge//At sonnei Outbreak isolate
KRS107 .lge/ll flexneri Dr. K.A. Lampel, FDA
KRS 108 ,\/1ge/At dysenteriae serotype 1 ATCC 9361
KRS109 \hgell/t sonnei ATCC 25931
KRS 110 ,\/hgell/t sonnei ATCC 29930
KRS 111 \h/gell/t sonnei ATCC 29030
KRS 112 .\hgell/t sonnei ATCC 9290
KRS 113 hlgellt sonnei ATCC 29031
KRS201 Salmonella Typhimurium ATCC 15277
KRS202 Salmonella Agona ATCC BAA-707
KRS203 Salmonella Gaminara ATCC BAA-711
KRS204 Salmonella Poona ATCC BAA-709
KRS205 Salmonella Montevideo ATCC BAA-710
KRS206 Salmonella Enteritidis Dr. G.E. Rodrick,
University of Florida
KRS207 Salmonella Agona LJH617 Dr. L.J. Harris, University
of California, Davis
KRS208 Salmonella Gaminara LJH618 Dr. L.J. Harris, University
of California, Davis
KRS209 Salmonella Michigan LJH621 Dr. L.J. Harris, University
of California, Davis
KRS210 Salmonella Montevideo LJH619 Dr. L.J. Harris, University
of California, Davis
KRS211 Salmonella Poona LJH630 Dr. L.J. Harris, University
of California, Davis
KRS212 Salmonella Enteritidis Environmental isolate
KRS213 Salmonella Miami Environmental isolate
KRS214 Salmonella spp. Ground beef isolate
KRS215 Salmonella spp. Ground beef isolate
KRS216 Salmonella spp. Ground beef isolate
KRS217 Salmonella spp. Ground beef isolate
KRS301 Escherichia coli K-12 Dr. A.C. Wright,
University of Florida
KRS302 Escherichia coli JM104 Dr. A.C. Wright,
University of Florida
KRS303 Escherichia coli ATCC 25922
KRS304 Escherichia coli 0157:H7 FSIS 063-93
KRS305 Escherichia coli 0157:H7 FSIS 413-95
KRS306 Escherichia coli 0157:H7 GFP-85 Deibel Laboratories,
Gainesville, FL









Table 3-3. Continued.
Designation
KRS307
KRS308
KRS309
KRS310
KRS311
KRS312
KRS313
KRS314
KRS315
KRS316
KRS402
KRS403
KRS404
KRS405
KRS406
KRS407
KRS408
KRS409
KRS410
KRS411
KRS412
KRS413
KRS414


KRS415

KRS416
KRS417
KRS418
KRS419
KRS420
KRS421
KRS422
KRS423
KRS424
KRS425
KRS426
KRS427


Culture
Escherichia coli 0157:H7
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Acinetobacter lwoffii
Acinetobacter anitratus
Citrobacter feundii
Hafnia alvei
Klebsiella ozonae
Klebsiella pneumoniae
Enterobacter spp.
Enterobacter cloacae
Enterobacter u-,-i, i m.ei i//\
Serratia mercesens
Pseudomonas cepacia
Pseudomonas maltophila
Staphylococcus aureus 12293

Staphylococcus aureus 25293


Citrobacter feundii
Citrobacter feundii
Citrobacter feundii
Enterobacter spp.
Citrobacter feundii
Hafnia alvei
Serratia mercesens
Citrobacter feundii
Klebsiella pneumoniae
Hafnia alvei
Hafnia alvei
Enterobacter cloacae


Origin
ATCC 700599
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Tomato isolate
Tomato isolate
Cantaloupe isolate
Cantaloupe isolate
Cantaloupe isolate
Environmental isolate
Tomato isolate
Cantaloupe isolate
Tomato isolate
Cantaloupe isolate
Tomato isolate
Tomato isolate
Deibel Laboratories,
Gainesville, FL
Deibel Laboratories,
Gainesville, FL
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate
Tomato isolate
Ground beef isolate
Ground beef isolate
Ground beef isolate


for 1 min followed by 45 cycles of 95.00C for 10 sec then 60.0C for 30 sec. Following

PCR amplification a melt curve was performed from 60.0C to 950C with the

temperature increasing at a rate of 0.50C/10 sec. The annealing temperature was









optimized for primer pairs showing initial specificity and the specificity testing was

repeated using reaction mixtures similar to that described above except that the

hydrolysis probe was added at 200 nM (final concentration) and IQTM Supermix was used

in place of IQTM Supermix with SYBR green. All real-time PCR was performed using the

iCycler (BioRad).

Binding of Biotinylated Capture Probes to Streptavidin-Coated Paramagnetic Beads

The capture probe was designed using the conserved sequence of the

chromosomally-located ipaH genes of S. sonnei Ss046 and was purchased from Sigma-

Genosys (The Woodlands, TX). The sequence of the capture probe was tested for

specificity in silico using BLASTn. To reduce steric hindrance during hybridization, the

capture probe was constructed with a 12-carbon spacer between the 5' nucleotide and the

biotin molecule as shown in Figure 3-1.

A 200-tl aliquot of Dynabeads M-280 Streptavidin (Dynal Biotech; Oslo, Norway)

was transferred to a clean, sterile 1.5 ml microcentrifuge tube. The beads were washed as

follows: two times in 200 ptl 2X Binding/Washing buffer (B/W buffer; see Appendix A),

two times (minimum of 1-3 min each) in 200 ptl Dynabeads Solution A (see Appendix

A), two times in Dynabeads Solution B (see Appendix A), one time in 200 ptl B/W buffer

and finally resuspended in 400 ptl B/W buffer. A 40-1l aliquot of a 10 [iM capture probe

solution and 360 ptl sterile water were added to the beads and the resulting mixture was

incubated at room temperature with gentle end-over-end rotation for 10 min. The capture

probe-bead complexes (hereafter referred to as CP-.\/i/ge/lt beads) were then washed

three times in 400 p1l IX B/W buffer and resuspended in 400 p1l 1X B/W buffer

containing 1.5 tl ethanolamine. The CP-'\/ ge//lt beads were incubated at room












fiBiotin-CH,),2-CTCCAGCATCTCATATTTCTGC





Figure 3-1. Design of the capture probe and CP-.\/'ge//t beads. Biotinylated DNA probes
were used to label streptavidin (SA)-coated paramagnetic beads for use in
DNA sequence capture experiments. A 12-carbon spacer was inserted
between the biotin molecule and the DNA probe sequence to alleviate steric
hindrance during hybridization.

temperature with gentle end-over-end rotation for 1 hr, washed 3 times in 400.l low-salt

wash buffer (see Appendix A) and finally resuspended in 400 pl low-salt wash buffer.

CP-.\/igel//t beads were stored at 40C prior to use.

Inoculum Preparation

Three days prior to each experiment, S. sonnei cultures were individually

cultivated (37C, static incubation) in 10 ml tubes of TSCR and overnight transfers were

performed using 10 pl sterile, disposable loops (BD Diagnostics) each day. On the day of

the experiment, a five-strain S. sonnei cocktail was compiled by transferring 2.0 ml from

each of the five 18-hr S. sonnei cultures (late stationary phase) to a clean, sterile 15 ml

centrifuge tube. The cocktail was centrifuged (3,220 x g for 10 min at 40C) and the

resulting pellet was washed twice in 10 ml of 0.1% peptone. The final cell titer of the

cocktail inoculum was determined by pour plate using TSA.

Calculation of Generation Time of S. sonnei in \/l.'e//at Broth

Each S. sonnei strain was cultured and transferred in TSB (370C) overnight for

three days. On the third day, 10 pl of an 18-hr culture was used to inoculate 100 ml

sterile SB in a 250 ml Erlenmeyer flask. The SB flask was then swirled to disperse the









inoculum and a 1.0 ml aliquot was used to prepare dilutions in 0.1% peptone water.

Appropriate dilutions were analyzed by pour plate using TSB to estimate the initial

bacterial population. The SB flask was incubated aerobically at 440C without shaking and

1.0 ml samples were analyzed as described above every 30 min. Using a Microsoft Excel

spreadsheet, the log bacterial population vs. time was plotted to determine the length of

lag phase and the generation time during exponential growth.

Preliminary Experiments with Anti-.\/ li,//At Beads

Two whole tomatoes and two 50 g ground beef samples were inoculated with the

five-strain S. sonnei cocktail at approximately 4.3 x 105 CFU/tomato and 2.2 x 105

CFU/25 g, respectively. Two 25-g aliquots of ground beef were transferred to clean,

sterile filtered stomacher bags each containing 225 ml SB and homogenized for 30 sec.

The two tomato samples were transferred to clean, sterile stomacher bags each containing

250 ml SB and subjected to a 30 sec vigorous shake followed a 1 min hand manipulation.

The filtered SB from the ground beef samples and the SB rinse from the tomato samples

were transferred to clean, sterile stomacher bags and incubated at 440C for 18 hr. After

enrichment, the samples were analyzed by FTI using anti-.\/igell//A beads prepared using

either AB02 or AB04. The recovered anti-./Nlge//a, beads were analyzed by spread plate

using MAC (440C for 24 hr). The resulting colonies were identified based on

biochemical reactions using the BBL EnterotubeTM II (Becton Dickinson, Sparks, MD).

Separation of S. sonnei from Food Matrices Using Low-Speed Centrifugation

A five-strain cocktail of the rifampicin resistant S. sonnei was prepared as

described above and diluted in 0.1% peptone water to a final concentration of 3.3 x 105

CFU/ml. To test low-speed centrifugation for the separation of S. sonnei from food

matrices, five replicate 25-g samples of potato salad and ground beef were homogenized









with 25 ml SB and transferred to clean, sterile 50 ml centrifuge tubes. The SB

homogenates were then spiked with 1.0 ml of the diluted S. sonnei cocktail and the

centrifuge tubes were mixed by vigorous shaking for 30 sec. Prior to centrifugation, 1.0

ml aliquots of the homogenate were serially diluted using 9.0 ml 0.1% peptone tubes and

1.0 ml aliquots from appropriate dilutions were analyzed by pour-plate using TSA rif+.

The remaining SB homogenates were subjected to low-speed centrifugation (LSC; 100 x

g for 5 min). After centrifugation, 1.0 ml aliquots from the supernatant were serially

diluted using 9.0 ml 0.1% peptone tubes and 1.0 ml aliquots from appropriate dilutions

were analyzed by pour-plate using TSA rif+.

Survival Studies

Sample Inoculation and Subsequent Recovery

Tomatoes were placed in sterile fiberglass trays with the blossom scar faced up.

Smooth surfaces around the blossom scar of tomatoes were spot inoculated at 10 sites per

fruit with 10 tl per site using appropriate dilutions to obtain final inoculation levels of

approximately 5.0 x 105 CFU/tomato with five replicate tomatoes at each level. The

inoculated tomatoes were dried for at least 1 hr in a laminar flow hood at room

temperature. After drying, tomatoes were stored in humidity chamber at 130C with 85%

relative humidity (RH) and five inoculated tomatoes were transferred to sterile Stomacher

bags (Seward, Norfolk, UK) containing 100 ml of sterile PBS at each observation day.

For recovery of inocula, tomatoes were shaken vigorously for 30 sec, then massaged by

hand for 1 min in the stomach bag similar to that described by Zhuang et al. (1995).

For trials involving potato salad and ground beef, 50-g samples were aseptically

weighed into sterile 4-oz specimen cups and inoculated with 1.0 ml of an appropriate

dilution to obtain final inoculation levels of ca 1.0 x 106 CFU/g. Using sterile tongue









depressors, the inoculum was homogenized in each sample for a minimum of 30 sec,

after which the samples were allowed to sit at room temperature for 1 hr for bacterial

attachment. After attachment, each set of specimen cups with samples were stored at

2.5C and 80C, respectively. For recovery of inocula, sterile tongue depressors were used

to weigh 25 g aliquots from each sample to sterile, clean Stomacher bags after which 225

ml of PBS was added and the samples were placed in the stomacher (Tekmar Company,

Cincinnati, OH) for 30 sec.

Three-Tube Most Probable Number Estimation of Survivors

Recovered inocula in 100 ml PBS rinses of tomatoes represented a 0.01 dilution

of the tomatoes surface population. Recovered inocula in 225 ml PBS homogenates of

potato salad or ground beef samples represented a 0.1/g dilution of the surviving

population. Recovered inocula were serially diluted using 9.0 ml 0.1% peptone tubes and

1.0 ml aliquots from appropriate dilutions were used to inoculate each of three TSB rif+

tubes. To enumerate survivors at the 0.1/tomato level, three 10 ml aliquots from the 100

ml PBS tomato rinse were used to inoculate 10 ml double-strength TSB rif+ tubes. All

TSB rif+ tubes were incubated overnight (370C; static incubation) and scored as either

"positive" or "negative" for Salmonella or S. sonnei based on the presence/absence of

visible growth. Surviving populations were estimated using the three-tube most probable

number (MPN) table located in Appendix 2 of the FDA BAM (2003). Non-inoculated

samples of each food were analyzed to confirm the absence of indigenous microflora

with resistance to rifampicin at 100 ppm.

Evaluation of Detection Methods

A schematic representation of the experimental design is presented in Figure 3-2.

Tomato smooth surfaces, potato salad and ground beef samples were inoculated with a








Inoculated Tomatoes, Potato
Salad and Ground Beef


BAM ./hi,,e/,
Culture Method

I
.l/igel/t Broth
(44C, 18 hr, anaerobic)



I
MAC
(37C, 20 min)


TSI, LIA, MM
(37C, 24 hr)


Biochemical Identification
(37C, 24 hr)


Flow-Through
Immunocapture Method

I
./nige/l Broth
(44C, 18 hr)


Pathatrix
(37C, 20 min)


DNA
Extraction


Real-time
PCR


DNA Sequence
Capture Method

I
./nige/l Broth
(44C, 18 hr)

I
Centrifugation and
Cell Lysis

I
Hybridization
(40C, 1 hr)

I
Real-time
PCR


Figure 3-2. Flow diagram of the experiments involving inoculated food samples. Each
food samples was analyzed by the ./lge//lt culture method of the FDA
Bacteriological AnalyticalManual (BAM), by flow-through immunocapture
followed by analysis of recovered beads by spread plate using MacConkey
agar (MAC) or using real-time PCR and by DNA sequence capture. Suspected
isolates on MAC were further analyzed using Triple Sugar Iron (TSI) agar
slants, Lysine Iron Agar (LIA) slants and Motility Medium (MM).
five-strain S. sonnei cocktail and recovery of the inocula was investigated using the BAM


.\/nge/ll culture method, FTI-MAC, FTI-PCR and DSC.









Inoculation of Samples and Subsequent Recovery

For tomato samples, tomatoes were placed in sterile fiberglass trays with the

blossom scar faced up. Smooth surfaces around the blossom scar of tomatoes were spot

inoculated at 10 sites per fruit with 10 [il per site using appropriate dilutions of the S.

sonnei cocktail to obtain final inoculation levels from 104 to 100 CFU/tomato with ten

replicate tomatoes at each level. The inocula were allowed to dry completely at room

temperature. After drying, inoculated tomatoes were transferred to sterile Stomacher bags

(Seward) containing 250 ml of sterile SB pre-warmed to 440C. For recovery of S. sonnei,

the stomacher bags were sealed using stomacher bag clips and the tomatoes were shaken

vigorously for 30 sec then massaged by hand for 1 min, similar to the method described

by Zhuang et al. (1995).

For potato salad and ground beef samples, 50 g of sample was weighed into a

sterile 4-oz specimen cup and inoculated with appropriate dilutions of the S. sonnei

cocktail to obtain final inoculation levels from 102 to 100 CFU/25 g with ten replicate

samples at each level. Sterile wooden tongue depressors were used to homogenize the

inoculum in each specimen cup. The inoculated potato salad and ground beef samples

were allowed to sit for 1 hr at room temperature for bacterial attachment. A 25-g aliquot

from each sample was weighed into a sterile dual-chambered stomacher bag containing

225 ml of SB and homogenized for 30 sec.

Modified BAM Culture Method for S. sonnei

For tomato samples, the recovered inocula in SB was transferred to sterile 18 oz

Whirl-Pak bags (Nasco, Modesto, CA) and the top of the bag was folded over twice and

secured. For potato salad and ground beef samples, the internal filter of the dual-

chambered stomacher bag was used to transfer the SB supernatant only to sterile 18 oz









Whirl-Pak bags and the top of the bag was folded over twice and secured. Samples were

incubated at 440C for 18-24 hr under anaerobic conditions using 7.0 liter rectangular jars

(Mitsubishi Gas Chemical Company, Inc., Japan) with the Pack-Anaero anaerobic gas

generating system (Mitsubishi Gas Chemical Company). SB enrichments were

homogenized by hand and streaked for isolation on MAC and incubated at 370C for 18-

24 hr.

For confirmation, three isolates colonies demonstrating typical ,\/lngel/,

morphology were selected from each MAC plate. When three typical isolates were not

present, atypical isolates were selected for confirmation. Each selected isolate was used

to inoculate a TSI slant, a LIA slant and MM and incubated overnight (37C). From each

isolate that resulted in typical reactions for .\/ilge/l on TSI slants, LIA slants and MM,

growth from the TSI slant was cultivated overnight (37C) in TSB. Growth from the TSB

tubes was then streaked for isolation on MAC and incubated overnight (37C).

Biochemical reactions were tested using the BBL EnterotubeTM II (Becton Dickinson).

Enterotubes were incubated overnight (37C) and positive reactions were read according

to manufacturer's instructions.

Flow-Through Immunocapture (FTI) Using the Pathatrix

All samples were inoculated and enriched as described above for the modified

BAM method with the following exceptions. SB tomato rinses and the filtered SB

supernatant of homogenized potato salad and ground beef samples were transferred to

clean, sterile stomacher bags prior to overnight enrichment. Anaerobic conditions were

not generated for any samples analyzed by FTI. Instead Stomacher clips were used to seal

the Stomacher bags and samples were incubated at 440C (static).









All FTI experiments were performed using the Pathatrix system (Matrix

MicroScience, Golden, CO). The assembly/operation of the Pathatrix and the recovery of

anti-.\/ige/ll, beads were performed as per manufacturer instructions. Briefly, Stomacher

bags containing sample enrichments were placed into a Pathatrix incubation pot. After

the tubing was properly placed, a 50-tl aliquot of anti-.\/Nlgel / beads was injected to the

system per manufacturer's instructions. After 20 min circulation at 370C, the tubing

assembly was disconnected from the sample and the anti-.\/lge/Al, beads were washed

with 100 ml of PBS and recovered in the collection vessel suspended in 5 to 10 ml PBS.

Using a magnet, the anti-.\/ige/ll, beads were drawn to the side of the collection vessel

and the volume of PBS was reduced to approximately 250 tl using a sterile transfer

pipet. A 50-tl aliquot of the resuspended anti-.\/ige/ll, beads was analyzed by spread

plate using MAC. MAC plates were incubated at 370C for 24 hr. In addition, DNA was

extracted from a 100-tl aliquot of the resuspended anti-.\/Nlgel / beads as described above

and analyzed by real-time PCR.

Sequence Capture of.\/,ige/,t DNA

A schematic representation of the DSC method is presented in Figure 3-3.

Preliminary experiments were performed to evaluate hybridization buffers, hybridization

temperatures, type of streptavidin-coated paramagnetic beads and sensitivity.

All procedures for inoculation and recovery of food samples was followed as

described above for FTI samples. After 18-24 hr incubation, the samples were shaken

briefly to mix contents. For potato salad and ground beef samples, a 10 ml aliquot was

aseptically transferred to a clean, sterile 15 ml centrifuge tube and solid food material

was sedimented using low-speed centrifugation (100 x g for 5 min). For tomato samples,

no solid food material was present in the overnight enrichments, therefore no low-speed










Inoculated Potato Salad or
Ground Beef


,'\/gel//t Broth
(44C, 18 hr, Filtered stomacher bag)


Inoculated Tomatoes



/iiigel/,a Broth
(44C, 18 hr)


10 ml Filtrate


Low-Speed Centrifugation
(100 x g, 5 min)
---


1.0ml
High-Speed Centrifugation
(6,000 x g, 5 min)


Discard
Pellet


Discard
Supernatant


Resuspend Pellet in
Hybridization Buffer 1


Heat (100C, 10 min)
Cool on ice (10 min)


High-Speed Centrifugation
(6,000 x g, 5 min)


Hybridization
(40C, 1 hr)

I
Wash CP-.\'/gell/A Beads



Real-time PCR


Discard
Pellet


Figure 3-3. Flow diagram of the DNA sequence capture method. CP-.\/igell// beads were
prepared using Dynabeads M-280 Streptavidin coated with a 5' biotin-labeled
DNA probe with specificity for the ipaH gene of .\/igelt// and enteroinvasive
E. coli.









centrifugation was performed. A 1.0-ml aliquot of the resulting supernatant (potato salad

and ground beef samples) or a 1.0-ml aliquot of the overnight enrichment (tomato

samples) was transferred to a clean, sterile 1.5 ml microcentrifuge tube. The bacterial

cells were then sedimented using centrifugation (6,000 x g for 5 min) and the supernatant

was discarded. The pellet was resuspended in 530 .il hybridization buffer 1 (see

Appendix A), heated at 1000C for 10 min and finally cooled on ice for 10 min. Cellular

and solid material were sedimented by centrifugation (6,000 x g for 5 min) and the

supernatant was transferred to a clean, sterile 1.5 ml microcentrifuge tube containing 200

pl 3.75 M NaCl and 20 pl CP-./nge//'a beads. The contents were mixed using the action

of the pipette and heated at 400C for 5 min on a dry-bath incubator followed by end-over-

end rotation (hybridization) at 400C for 1 hr. Heated end-over-end rotation was achieved

using a Rugged Rotator (Glas-Col, Terre Haute, IN) inside of an environmental chamber

(Lab-Line, E2 series, Barnstead International, Melrose Park, IL). Following

hybridization, the beads were recovered using a magnetic rack and washed 2 times in 200

pl wash buffer (see Appendix A), 2 times in 200 pl low-salt wash buffer (see Appendix

A) and resuspended in 50 pl TE buffer. The resuspended beads were then heated at 75C

for 10 min to release the captured DNA from the probe and, using the magnetic rack, the

DNA-containing supernatant was transferred to a clean, sterile 1.5 ml microcentrifuge

tube. The DNA samples were placed on ice or frozen (-200C) prior to analysis by PCR.

Real-Time PCR and Reverse Transcriptase (RT) PCR

All real-time PCR and real-time RT-PCR analyses were performed using the

iCycler (Bio-Rad, Hercules, CA). For the analysis of FTI samples, real-time PCR was

performed using 50 pl reaction mixtures consisting of: 25 pl IQTM Supermix (Bio-Rad),

200 nM (final concentration) each of forward and reverse primer, 200 nm (final









concentration) hydrolysis probe, 20 pl DNA sample and purified water. The PCR

cycling conditions were 95C for 1 min followed by 40 cycles of 95C for 10 sec then

600C for 30 sec. For the analysis for DNA sequence capture samples, real-time PCR was

performed using 20 pl reaction mixtures consisting of: 10 pl IQTM Supermix (Bio-Rad),

200 nM (final concentration) each of forward and reverse primer, 200 nm (final

concentration) hydrolysis probe, 5 pl DNA sample and purified water. The PCR cycling

conditions were the same as for FTI samples.

Real-time RT-PCR was performed using 20 [l reaction mixtures consisting of: 10

pl 2X RT-PCR Reaction Mix for Probes (Bio-Rad), 200 nM (final concentration) each of

forward and reverse primer, 200 nM (final concentration) hydrolysis probe, 0.4 [il reverse

transcriptase (Bio-Rad), 0.3 .il RNasin (Promega, Madison, WI), 5 .il RNA sample and

purified water. The PCR cycling conditions were 500C for 10 min, 95C for 5 min

followed by 40 cycles of 95C for 10 sec then 600C for 30 sec. For all RNA samples

analyzed by RT-PCR, parallel reaction mixtures without the addition of reverse

transcriptase were prepared to verify complete digestion of DNA.

All PCR and RT-PCR reaction mixtures were prepared in a Labconco Purifier

Class II safety cabinet (Labconco Corporation, Kansas City, MO).

Recording of Data and Statistical Analysis

All statistical analyses of survival studies were performed using the Statview

statistical software package (SAS) version 9.1 (SAS Institute Inc., Cary, NC) using a

mixed model. Sample replications were treated as random variables within time.

Statistical analysis of population means in experiments involving LSC were performed

by hand using a two sample t-test. P values < 0.05 were considered significant.






76


All results from the evaluation of detection methods were recorded as "positive" or

"negative" for the detection of S. sonnei. Positive isolation on plating media resulted

from typical reactions for S. sonnei in all confirmation steps and positive biochemical

BBL EnterotubeTM II identification. Bias-reduced logistic regression (BRLR) models

were constructed using the R software (The R Foundation for Statistical Computing,

Version 2.2.0, http://cran.us.r-project.org/) to identify significant differences (P < 0.05)

among the detection methods tested.














CHAPTER 4
RESULTS

This study consisted of three phases of research. The first phase consisted of

preliminary trials involving the preparation of growth curves to calculate the generation

time of S. sonnei in SB, the development of primers/probes for the detection of S. sonnei

and testing the specificity of each set of primers/probes against a DNA library of positive

and negative controls. The second phase consisted of experiments designed to test the

ability of S. sonnei to survive in/on selected foods. The third phase of this study involved

the evaluation of newly developed sampling methods for the detection of artificially

inoculated S. sonnei on smooth tomato surfaces and in potato salad and ground beef.

Recovery/detection of the inocula was tested using the BAM .\//ge,//t culture method and

the newly developed methods: flow-through immunocapture (FTI) followed by analysis

of recovered anti-.\/ige,// beads by spread-plate using MAC (FTI-MAC), FTI followed

by analysis of recovered anti-.\/ilge,// beads by real-time PCR (FTI-PCR) and DNA

sequence capture (DSC).

Preliminary Studies

Calculation of Generation Time of S. sonnei in .\/,hle/l t Broth

Growth curves were prepared for S. sonnei ATCC 9290, S. sonnei ATCC 29031, S.

sonnei ATCC 29030, S. sonnei ATCC 25931 and S. sonnei ATCC 29930 (Figures 4-1, 4-

2, 4-3, 4-4 and 4-5, respectively) in \/lge//At broth (SB) incubated aerobically without

shaking at 440C. The length of the lag phase was identified and the exponential phase

was used to calculate the generation time for each strain. The average growth kinetics









among the five S. sonnei strains was a lag phase of approximately 2 hr and a generation

time of 18.8 + 0.6 min.

Growth curve of S. sonnei ATCC 9290

When the growth of S. sonnei ATCC 9290 was investigated in SB (44C), an initial

lag phase of 2 hr was observed prior to exponential growth (Figure 4-1). Logarithmic

regression used to analyze the exponential phase of growth showed linearity (R2 = 0.997)

and the equation of the line was used to calculate a generation time of 19.0 min. The

initial population ofS. sonnei ATCC 9290 was 4.3 x 104 CFU/ml and the final population

was 3.8 x 108 CFU/ml.

Growth curve of S. sonnei ATCC 29031

When the growth of S. sonnei ATCC 29031 was investigated in SB (44C), an

initial lag phase of 2 hr was observed prior to exponential growth (Figure 4-2).

Logarithmic regression used to analyze the exponential phase of growth showed linearity

(R2 = 0.994) and the equation of the line was used to calculate a generation time of 19.5

min. The initial population of S. sonnei ATCC 29031 was 3.8 x 104 CFU/ml and the final

population was 2.4 x 108 CFU/ml.

Growth curve of S. sonnei ATCC 29030

When the growth of S. sonnei ATCC 29030 was investigated in SB (44C), an

initial lag phase of 2 hr was observed prior to exponential growth (Figure 4-3).

Logarithmic regression used to analyze the exponential phase of growth showed linearity

(R2 = 0.996) and the equation of the line was used to calculate a generation time of 18.6

min. The initial population ofS. sonnei ATCC 29030 was 4.0 x 104 CFU/ml and the final

population was 5.0 x 108 CFU/ml.











9.00

8.00 y = 0.9511x + 2.8482
SR2 = 0.997
S7.00

a 6.00

5.00

4.00
0 1 2 3 4 5 6 7 8

Time (hr)

Figure 4-1. Growth curve: S. sonnei ATCC 9290 in .\l/lge//l broth. A 100 ml microcosm
was inoculated with a 10-tl aliquot of an 18-hr S. sonnei ATCC 9290 culture
and incubated (44C, static). At appropriate time intervals, a 1.0 ml aliquot
was serially diluted in 0.1% peptone and the population was estimated by pour
plate using tryptic soy agar. (o) lag/stationary phase growth; (m) exponential
phase growth. Error bars represent one standard deviation.


7

b 6
o
5

4


0 1 2 3 4 5 6 7 8


Time (hr)

Figure 4-2. Growth curve: S. sonnei ATCC 29031 in .\llg'e//l broth. A 100 ml microcosm
was inoculated with a 10-pl aliquot of an 18-hr S. sonnei ATCC 29031 culture
and incubated (44C, static). At appropriate time intervals, a 1.0 ml aliquot
was serially diluted in 0.1% peptone and the population was estimated by pour
plate in tryptic soy agar. (o) lag/stationary phase growth; (m) exponential
phase growth. Error bars represent one standard deviation.













9
y = 0.9734x + 2.5896
A R2 = 0.996
7

06



4
0 1 2 3 4 5 6 7 8

Time (hr)

Figure 4-3. Growth curve: S. sonnei ATCC 29030 in .h\lgel// broth. A 100 ml microcosm
was inoculated with a 10-Cl aliquot of an 18-hr S. sonnei ATCC 29030 culture
and incubated (44C, static). At appropriate time intervals, a 1.0 ml aliquot
was used to make serial dilutions in 0.1% peptone and the population was
estimated by pour plate using tryptic soy agar. (o) lag/stationary phase
growth; (m) exponential phase growth. Error bars represent one standard
deviation.

Growth curve of S. sonnei ATCC 25931

When the growth of S. sonnei ATCC 25931 was investigated in SB (44C), an

initial lag phase of 2 hr was observed prior to exponential growth (Figure 4-4).

Logarithmic regression used to analyze the exponential phase of growth showed linearity

(R2 = 0.990) and the equation of the line was used to calculate a generation time of 18.0

min. The initial population of S. sonnei ATCC 25931 was 2.9 x 104 CFU/ml and the final

population was 3.7 x 108 CFU/ml.

Growth curve of S. sonnei ATCC 29930

When the growth of S. sonnei ATCC 29930 was investigated in SB (44C), an

initial lag phase of 2 hr was observed prior to exponential growth (Figure 4-5).

Logarithmic regression used to analyze the exponential phase of growth showed linearity






















0 1 2 3 4 5 6 7 8


Time (hr)

Figure 4-4. Growth curve: S. sonnei ATCC 25931 in .\lnlgIe// broth. A 100 ml microcosm
was inoculated with a 10-pl aliquot of an 18-hour S. sonnei ATCC 25931
culture and incubated (44C, static). At appropriate time intervals, a 1.0 ml
aliquot was used to make serial dilutions in 0.1% peptone and the population
was estimated by pour plate using tryptic soy agar. (o) lag/stationary phase
growth; (m) exponential phase growth. Error bars represent one standard
deviation.


9.00

8.00

7.00

6.00

5.00

4.00


0 1 2 3 4 5 6 7 8


Time (hr)

Figure 4-5. Growth curve: S. sonnei ATCC 29930 in .\lNlgIe// broth. A 100 ml microcosm
was inoculated with a 10-Cl aliquot of an 18-hour S. sonnei ATCC 29930
culture and incubated (44C, static). At appropriate time intervals, a 1.0 ml
aliquot was used to make serial dilutions in 0.1% peptone and the population
was estimated by pour plate using tryptic soy agar. (o) lag/stationary phase
growth; (m) exponential phase growth. Error bars represent one standard
deviation.









(R2 = 0.998) and the equation of the line was used to calculate a generation time of 19.1

min. The initial population ofS. sonnei ATCC 29930 was 1.6 x 104 CFU/ml and the final

population was 4.3 x 108 CFU/ml.

Evaluation Anti-.\/lgle/ t Antibodies for Use with Flow-Through Immunocapture

Two polyclonal antibodies (AB01 and AB04) and one monoclonal antibody

(AB02) were evaluated for use in FTI of S. sonnei using the Pathatrix (Table 4-1). AB01

was generated using a mixture of S. boydii, S. flexneri and S. dysenteriae as the

immunogens. AB04 was generated using a membrane extract mixture of S. sonnei and S.

flexneri. In addition, a secondary capture in which S. sonnei pre-bound with AB01 or

AB02 in solution were separated from food matrices using FTI with paramagnetic beads

coated with goat anti-rabbit antibodies or rabbit anti-mouse antibodies (AB03 or AB05,

respectively).

The number of S. sonnei colonies which resulted from FTI followed by analysis of

recovered beads by spread plate using MAC using anti-.\/ lge//A beads prepared with the

various anti-.\s/l//ig/ antibodies is listed in Table 4-1. When AB01 was used for the

preparation of anti-.\slgl//A, beads, a population too numerous to count (TNTC) of S.

sonnei ATCC 25931, two colonies of S. sonnei ATCC 29930, 78 colonies of S. sonnei

ATCC 9290 and no colonies of S. sonnei ATCC 29030 or S. sonnei ATCC 29031 were

observed. When AB02 was used for the preparation of anti-.\/ilgl//A beads, a population

TNTC of S. sonnei ATCC 25931 and S. sonnei ATCC 9290, 147 colonies of S. sonnei

ATCC 29930, 112 colonies of S. sonnei ATCC 29030 and 87 colonies of S. sonnei ATCC

29031 was observed. When AB04 was used for the preparation of anti-.\/lgl//t beads,

261 colonies of S. sonnei ATCC 25931, 182 colonies of S. sonnei ATCC 29930, 15

colonies of S. sonnei ATCC 29030, 53 colonies of S. sonnei ATCC 9290 and










Table 4-1. Evaluation of anti-.\/igel // antibodies for flow-through immunocapture (FTI)
of S. sonnei. Microcosms of individual S. sonnei strains were prepared at
approximately 2.0 x 105 CFU/ml in 250 ml phosphate buffered saline (PBS).
The microcosms were then analyzed by FTI for 30 min at 370C using various
anti-.\/ilgel//t beads. After circulation, 100 pl aliquots of the recovered beads
were analyzed by spread plate using MacConkey agar (MAC; 370C for 24 hr).
The numbers of resulting colonies were enumerated in order to compare
immunocapture by the various antibodies. In the AB03-AB03 and AB05-
AB02 experiments, AB01 and AB02 were added directly to inoculated PBS
and allowed to bind for 5 min with shaking (60 rpm) followed by FTI using
beads coated with AB03 and AB05, respectively.

Plate counts of S. sonnei strains
ATCC ATCC ATCC ATCC ATCC
Antibodies 25931 29930 29030 9290 29031
AB01 TNTCa 2 NGb 78 NG
AB02 TNTC 147 112 TNTC 87
AB03-AB01 148 263 TNTC 68 TNTC
AB04 261 182 15 53 47
AB05-AB02 TNTC NG NG 10 NG
a too numerous to count
b no growth

47 colonies of S. sonnei ATCC 29031 was observed. These data suggest that anti-

.\/ige//lt beads prepared with AB02 or AB04, but not AB01, may be used for the

consistent immunocapture of S. sonnei by FTI.

When AB01 was used to bind S. sonnei prior to analysis by FTI using paramagnetic

beads coated with AB03 (secondary capture), populations TNTC of S. sonnei ATCC

29030 and S. sonnei ATCC 29031, 148 colonies of S. sonnei ATCC 25931, 263 colonies

of S. sonnei ATCC 29930 and 68 colonies of S. sonnei ATCC 9290 were observed. When

AB02 was used to bind S. sonnei prior to analysis by FTI using paramagnetic beads

coated with AB05 (secondary capture), populations TNTC of S. sonnei ATCC 25931, 10

colonies of ATCC 9290 and no colonies of S. sonnei ATCC 29930, ATCC 29031 or

ATCC 29030 were observed. These data suggest that secondary capture by FTI using









AB03 of S. sonnei which have been prior labeled with AB01 may be used for the

consistent immunocapture. Secondary capture by FTI using AB05 of S. sonnei that had

been prior labeled with AB02 did not provide consistent immunocapture under the

conditions investigated. Using the Pathatrix for secondary capture of S. sonnei that had

been prior labeled with anti-.\lhigel// antibodies required additional steps and time over

using the Pathatrix for primary capture.

Preliminary Experiments with Anti-.\l igel'// Beads

Tomato and ground beef samples were inoculated with a five-strain S. sonnei

cocktail and analyzed by FTI followed by analysis of the recovered beads by spread plate

using MAC. When anti-.\hlgell// beads were prepared with AB02, analysis of tomato and

ground beef samples resulted in MAC plates with only colonies exhibiting morphologies

typical for S. sonnei, and each colony selected (three colonies from each plate) for

confirmation was identified as S. sonnei. When anti-.\/1gel// beads were prepared with

AB04, analysis of tomato and ground beef samples resulted in MAC plates with colony

morphologies both typical and atypical for S. sonnei. Atypical colonies from tomato

samples were identified as Citrobacterfreundii, Enterobacter spp. and Klebsiella

pneumoniae. Atypical colonies from ground beef samples were identified as

Enterobacter cloacae and E. coli. Typical colonies from tomato samples were identified

as S. sonnei or E. coli. These data suggest that anti-.\/lge/ll beads prepared with AB02,

but not those prepared with AB04, may be used for the specific detection ofS. sonnei in

tomato and ground beef samples. Anti-.\lhige/la beads prepared with AB02 were used for

all subsequent experiments (hereafter referred to as anti-.\hlgel//, beads).

The manufacturer of AB02 reported no reactivity with S. boydii, S. flexneri or S.

dysenteriae, however when anti-.\hlgell// beads were tested against solutions of these









serogroups in PBS using the FTI-MAC method (as described for S. sonnei strains) the

anti-.\/igell//A beads reacted strongly with the S. flexneri strain (KRS106) and the S.

dysenteriae type 1 strain (KRS 108). When tested against two stock strains of S. boydii

type 18, anti-.\/ige,//A beads reacted weakly with one strain (KRS101) but did not react

with the other strain (KRS 102).

Optimization of Anti-.,\/lg,/ t Bead Concentration for Flow-Through Immunocapture of
S. sonnei

To optimize the concentration of anti-.,/Nigel//A beads required for detection of S.

sonnei by FTI, anti-.\/ige'// beads were diluted in PBS containing 0.1% sodium azide

and tested against various concentrations of S. sonnei ATCC 25931 or S. sonnei 29930

(Table 4-2). Twenty-five-il aliquots of the initial concentration of anti-.\/lgell// beads (50

mg/ml) and the following diluted concentrations were tested for FTI: 25 mg/ml (1:1

dilution), 16.7 mg/ml (1:2 dilution), 12.5 mg/ml (1:3 dilution) and 10 mg/ml (1:4

dilution). The number of colonies on MAC plates which resulted from the analysis of

various concentrations of S. sonnei ATCC 25931 and S. sonnei ATCC 29930 by FTI

using the various concentrations of anti-.\/igel/h beads are given in Table 4-2.

When S. sonnei ATCC 25931 microcosms of 2.8 x 103 CFU/ml and 3.7 x 105

CFU/ml were analyzed by FTI, all tested dilutions of anti-.\/ige/ll, beads resulted in

MAC plates with populations TNTC. For S. sonnei ATCC 25931 microcosms of 3.2 x

101 CFU/ml analyzed by FTI, undiluted anti-./iNge/lat beads and dilutions of 1:1, 1:2, 1:3

and 1:4 anti-.\///le, /t beads resulted in MAC plates with 27, 18, 13, 6 and 12 colonies,

respectively. These data suggest the sensitivity of FTI for S. sonnei ATCC 25931 is

approximately 3.0 x 101 CFU/ml and that any of the tested concentration of anti-.\/igell//A

beads may be used for detection.









Table 4-2. Optimization of anti-./Nigel//A bead concentration for flow-through
immunocapture (FTI) of S. sonnei. Microcosms of S. sonnei ATCC 25931 or
ATCC 29930 at various concentrations were prepared in phosphate buffered
saline (PBS). The microcosms were then analyzed by FTI for 30 min at 370C
using 25-pl aliquots of various dilutions of anti-.\/ige/ll, beads. After
circulation, 100-.l aliquots of the recovered beads were analyzed by spread
plate using MacConkey agar (MAC; 370C for 24 hr). The numbers of
resulting colonies were enumerated in order to determine the sensitivity of
FTI using various concentrations of anti-.\/igel/h beads.

S. sonnei titer Plate counts from dilution of anti-.\/,gel/ht beads
(CFU/ml) None 1:1 1:2 1:3 1:4
ATCC 25931
3.2 x 101 27 18 13 6 12
2.8 x 103 TNTCa TNTC TNTC TNTC TNTC
3.7 x 105 TNTC TNTC TNTC TNTC TNTC

ATCC 29930
2.1 x101 NGb NG NG NG NG
2.7 x 103 NG NG NG NG NG
2.8 x 105 11 7 13 12 12
a too numerous to count
b no growth

When S. sonnei ATCC 29930 microcosms of 2.1 x 101 CFU/ml and 2.7 x 103

CFU/ml were analyzed by FTI, all tested dilutions of anti-.\/igell//A beads resulted in

MAC plates with no colonies. For S. sonnei ATCC 29930 microcosms of 2.8 x 105

CFU/ml analyzed by FTI, undiluted anti-.\/iigell// beads and dilutions of 1:1, 1:2, 1:3 and

1:4 anti-./Nlge//a, beads resulted in MAC plates with 11, 7, 13, 12 and 12 colonies,

respectively. These data suggest the sensitivity of FTI for S. sonnei ATCC 29930 is

approximately 3.0 x 105 CFU/ml and that any of the tested concentration of anti-.\/ige/ll,

beads may be used for detection.

Although the tested concentrations of anti-.\/ige/ll, beads did not affect detection of

S. sonnei ATCC 29531 or ATCC 29930 in PBS, the addition of 25 pl anti-.\/lgell/A, bead

aliquots per FTI analysis of ground beef samples resulted in poor visual bead recovery.









To improve visual bead recovery, 50-rl aliquots of the 10 mg/ml anti-.\/Nigel//A bead

dilution were used for FTI in all inoculated trials.

Identification of Potentially .\/i.ge// t-Specific Genetic Loci

The MBGD and BLASTn were used to identify genetic loci potentially specific for

.\lge//At spp. or for S. sonnei alone. The genes identified using the MBGD are listed in

Table 4-3. Using BLASTn, none of the genes were identified with specificity for all five

of the .\/ige/lt genomes within the database (Table 4-4), however the genes SSO_0670,

SSO_2685, SSO_3247 and SSO_0721 were identified as potentially specific for some

species of.\/ige/llt When analyzed using BLASTn, the genes SSO_2067, SSO_2071,

SSO_1019, SSO_2059 and SSO_2863 were identified as potentially specific for S. sonnei

(Table 4-4). The nucleotide sequences of the identified genes were used to develop PCR

primers (Table 3-2).

Specificity of Primers Developed for Potentially .\li.ge/h t-Specific Genetic Loci

The developed primers were evaluated by real-time PCR against DNA extracted

from stock ./lge//lt spp. (Table 4-4) and closely-related microorganisms (data not

shown). Primer sets 01-025, 01-026, 01-028 and 01-029 amplified DNA from eight

strains of S. sonnei and one strain each of S. flexneri and S. dysenteriae, however DNA

from two strains of S. boydii were not amplified. The observed amplification of DNA

from the stock strains of S. flexneri and S. dysenteriae were not in agreement with the in

silico analysis using BLASTn. Primer set 01-027 amplified DNA from only three of five

S. sonnei strains tested; therefore it was not investigated against DNA from other .\/lge/ht