The Pennsylvania State University. The Graduate School. Department of Food Science VIRULENCE GENE AND CRISPR MULTILOCUS SEQUENCE TYPING

Transcription

1 The Pennsylvania State University The Graduate School Department of Food Science VIRULENCE GENE AND CRISPR MULTILOCUS SEQUENCE TYPING SCHEME FOR SUBTYPING THE MAJOR SEROVARS OF SALMONELLA ENTERICA SUBSPECIES ENTERICA A Thesis in Food Science by Fenyun Liu 2010 Fenyun Liu Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science December 2010

2 ii The thesis of Fenyun Liu was reviewed and approved* by the following: Stephen J. Knabel Professor of Food Science Thesis Co-Advisor Edward G. Dudley Assistant Professor of Food Science Thesis Co-Advisor Bhushan M. Jayarao Professor of Veterinary Science Rodolphe Barrangou Adjunct Professor of Food Science John D. Floros Professor of Food Science Head of the Department of Food Science *Signatures are on file in the Graduate School

3 iii ABSTRACT Salmonella enterica subsp. enterica is the leading cause of bacterial foodborne disease in the United States. Molecular subtyping methods are powerful tools for tracking the farm-to-fork spread of foodborne pathogens during outbreaks. In order to develop a novel multilocus sequence typing (MLST) scheme for subtyping the most prevalent serovars of Salmonella, the virulence genes fimh and ssel and Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR) regions were sequenced from 171 clinical isolates from serovars Typhimurium, Enteritidis, Newport, Heidelberg, Javiana, I 4, [5], 12; i: -, Montevideo, Muenchen and Saintpaul. Another 63 environmental isolates and 70 poultry isolates of S. Enteritidis from poultry industries in PA were also analyzed. The MLST scheme using only virulence genes was insufficient to separate all unrelated outbreak clones. However, the addition of CRISPR sequences dramatically improved discriminatory power of this MLST method. Moreover, the present MLST scheme provided better discrimination of S. Enteritidis strains than PFGE. Cluster analyses revealed the current MLST scheme is highly congruent with serotyping and epidemiological data. For the analyses with S. Enteritidis isolates, the current MLST scheme identified three persistent and predominant sequence types circulating among humans in the U.S. and poultry and hen house environments in PA. It also identified an environment-specific sequence type. Moreover, cluster analysis based on fimh and ssel identified three epidemic clones and one outbreak clone of S. Enteritidis. In conclusion, the novel MLST scheme described in the present study accurately differentiated outbreak clones of the major serovars of Salmonella, and therefore may be an excellent tool for subtyping this important foodborne pathogen during outbreak investigations. Furthermore, the MLST scheme may provide information about the ecological origin of S. Enteritidis isolates, potentially identifying strains that differ in virulence capacity.

4 iv TABLE OF CONTENTS LIST OF FIGURES vi LIST OF TABLES.vii LIST OF ABBREVIATIONS AND DEFINITIONS viii ACKNOWLEDGEMENTS.x Chapter 1 Statement of the problem... 1 Chapter 2 Literature review Salmonellosis Salmonella Salmonella taxonomy and serotyping Evolution of pathogenicity Salmonella reservoirs Salmonella association with foods Most common Salmonella serovars associated with human illnesses Subtyping of Salmonella Important definitions and performance criteria of subtyping methods Salmonella subtyping methods during epidemiologic investigations Phenotypic methods Serotyping Phage typing Multilocus enzyme electrophoresis (MLEE) Genotypic methods DNA-fragment-pattern-based methods Pulsed-Field Gel Electrophoresis (PFGE) Amplified Fragment Length Polymorphism (AFLP) Multiple Loci Variable number tandem repeat Analysis (MLVA) DNA-sequence-based methods Multilocus Sequence Typing (MLST) Multi-Virulence-Locus Sequence Typing (MVLST) Single Nucleotide Polymorphism (SNP) analysis Clustered Regularly Interspaced Palindromic Repeat (CRISPR) CRISPR in Salmonella Conclusions References Chapter 3 Novel virulence gene and CRISPR multilocus sequence typing scheme for subtyping the major serovars of Salmonella enterica subspecies enterica Abstract... 47

5 v 3.2 Introduction Materials and methods Results Discussion Acknowledgements References Chapter 4 Characterization of clinical, poultry and environmental Salmonella Enteritidis isolates using multilocus sequence typing based on virulence genes and CRISPRs Abstract Introduction Materials and methods Results Discussion Acknowledgements References Chapter 5 Conclusions and future research Conclusions Future research APPENDIX Supplemental materials

6 vi LIST OF FIGURES Figure 2.1 Model for the three-phase evolution of pathogenicity in Salmonella enterica subspecies enterica. The phylogenetic tree is not drawn to scale (7) Figure 2.2 Schematic view of the two CRISPR systems in Salmonella Typhimurium LT Figure 3.1. Schematic view of the two CRISPR systems in Salmonella Typhimurium LT Figure 3.2. (a) Cluster diagram based on only fimh and ssel. (b) Cluster diagram based on fimh, ssel and CRISPRs (combined allele of CRISPR1 and CRISPR2) Figure 4.1. Potential routes of transmission of S. Enteritidis contamination throughout the egg food system Figure 4.2. Schematic view of the two CRISPR systems in Salmonella Enteritidis strain P Figure 4.3. Frequency of the five predominant sequence types (E ST1, 3, 4, 8 and 10) in clinical, poultry and environmental isolates Figure 4.4. Cluster diagram based on only fimh and ssel for all 27 sequence types Figure 4.5. Cluster diagram based on virulence genes and CRISPRs for all 27 sequence types Figure 4.6. Graphic representation of spacer arrangements in CRISPR1 and CRISPR2 of the 27 S. Enteritidis sequence types Figure S1. Graphic representation of spacer arrangements in CRISPR1 and CRISPR

7 vii LIST OF TABLES Table 2.1 Top ten most frequently reported serovars from human sources in Table 2.2 Top ten most frequently reported serovars from human sources in Table 3.1. Top nine most frequently reported serovars from human sources in 2005 which were analyzed in the present study Table 3.2. Outbreak information, PFGE profile and MLST results for the 171 isolates analyzed in the present study Table 3.3. Size, function and nucleotide location of the four markers targeted in the present study Table 3.4. Primers used to amplify and sequence the four MLST markers Table 3.5. Number of isolates, allelic types and sequence types in each serovar Table 3.6. Allelic polymorphisms and nucleotide substitutions in the nucleotide sequences of fimh and ssel Table 3.7. Analysis of CRISPR repeat sequences Table 3.8. Analysis of CRISPR spacers in different serovars Table 3.9. Comparison of epidemiologic concordance 1 between PFGE and MLST based on virulence genes and CRISPRs for the selected strains analyzed in the present study Table 4.1. Sources, sample types and isolation information for the 167 S. Enteritidis isolates analyzed in the present study Table 4.2. Primers used to amplify and sequence the four MLST markers Table S1. Primers used to amplify and sequence other virulence genes Table S2. Source, isolate information and MLST results for the 167 isolates analyzed in the present study

8 viii LIST OF ABBREVIATIONS AND DEFINITIONS ADL AFLP bp C CDC Animal Diagnostic Lab Amplified Fragment Length Polymorphism Base Pair Cytosine Centers for Disease Control and Prevention C Degree Celsius Clone A group of isolates deriving from a common ancestor as part of a direct chain of replication and transmission from host to host or from the environment to host. CRISPR D DNA dntp DR E EC ml G MLEE MLST MLVA MVLST NCBI Clustered Regularly Interspaced Short Palindromic Repeats Discriminatory Power Deoxyribonucleic Acid Deoxyribonucleotide Triphosphate Direct Repeat Epidemiological Concordance Epidemic Clone milliliter Guanine Multi-Locus Enzyme Electrophoresis Multilocus Sequence Typing Multiple-Locus Variable-number tandem repeat Analysis Multi-Virulence-Locus Sequence Typing National Center for Biotechnology Information

9 ix PCR PFGE PEQAP RNA rrna SNP Strain Polymerase Chain Reaction Pulsed-Field Gel Electrophoresis Pennsylvania Egg Quality Assurance Program Ribonucleic Acid Ribosomal Ribonucleic Acid Single Nucleotide Polymorphism Isolate(s) that exhibit distinct phenotypic and/or genotypic characteristics from other isolates of the same species ST T USDA μl WGS Sequence Type Thymine United States Department of Agriculture microliter Whole Genome Shotgun Clone and strain were defined previously by Struelens et al. (101).

10 x ACKNOWLEDGEMENTS I thank my parents, Zijian Liu and Guixiang Liu, who support and encourage me to study in the US. I am also grateful for the support of my sister, Fenni Liu. I would like to give my sincere thanks my advisors, Dr. Stephen J. Knabel and Dr. Edward G. Dudley. I learned from them not only how to do research but also how to lead my life. I feel so grateful for the working experience with them. I also thank my committee members, Dr. Rodolphe Barrangou, and Dr. Bhushan M. Jayarao for their guidance and encouragement. Additionally, I thank Dr. Kariyawasam, Dr. Gerner-Smidt and Dr. Ribot for their help with the research. I thank my labmates, Jia Wen, Mei Lok, Gabari, Michelle, Carrie, and Mat for their help and encouragement. I also want to give special thanks to Dr. Bindhu Verghese for her guidance and help with my research. Furthermore, I want to thank all the faculty, graduate students and staff in the Department of Food Science for their support. At last, I thank USDA and the Department of Food Science for supporting my research.

11 1 Chapter 1 Statement of the problem Salmonella is one of the most common foodborne bacteria worldwide. In the United States alone, there were approximately 1.4 million cases of salmonellosis each year since 1996, which resulted in a heavy burden on public health and the economy. In order to develop effective intervention strategies to control salmonellosis during outbreaks, it is critical to rapidly and accurately track the farm-to-fork spread of Salmonella. Molecular subtyping methods are powerful tools for investigating the transmission of Salmonella by characterizing specific outbreak clones. Serotyping has been one of the major subtyping methods employed during outbreaks to provide base line information about the serovar involved. There are approximately 2,500 different serovars of Salmonella; however, the top ten serovars caused approximately 60% of all outbreak cases. Each of those top serovars is known to cause numerous outbreaks, each of which is typically caused by a specific outbreak clone. Therefore, molecular subtyping methods, which are generally more discriminatory than serotyping, are needed to further distinguish different strains of a particular serovar. Pulsed-field gel electrophoresis (PFGE) is currently CDC s gold standard approach for subtyping Salmonella. However, PFGE sometimes lacks discriminatory power and epidemiologic concordance for typing clonal serovars, such as S. Enteritidis and S. Montevideo. Many studies have been conducted to develop alternative subtyping methods, one of which is multi-locus sequence typing (MLST). Previous MLST schemes for Salmonella focused mainly on discriminatory power; however, none of the previous MLST studies examined the epidemiologic concordance of the MLST schemes or attempted to distinguish strains within highly clonal Salmonella serovars, such as S. Enteritidis and S.

12 2 Montevideo. Moreover, for S. Enteritidis, our knowledge of their epidemiology is hindered due to its clonal nature. Therefore, the main purpose of the present study was to enhance the molecular epidemiology of Salmonella by developing an MLST scheme that has both high discriminatory power and high epidemiologic concordance for subtyping the major serovars of Salmonella.

13 3 Chapter 2 Literature review 2.1 Salmonellosis Salmonella infections (salmonellosis) include three forms of disease: gastroenteritis, bacteremia and typhoid fever. After ingestion of Salmonella into the gastrointestinal system, gastroenteritis can develop, which is characterized by symptoms such as abdominal pain, nausea, vomiting and diarrhea. More severe manifestations of salmonellosis, such as bacteremia and typhoid fever can develop after the invasion of Salmonella into the bloodstream. Common symptoms of bacteremia are fever, focal infections, sepsis and meningitis. Typhoid fever is a deadly systemic infection for humans caused by S. Typhi. The incidence of typhoid fever has declined in the U.S. with approximately 400 cases annually (33). On the other hand, infections due to nontyphoidal Salmonella (mainly gastroenteritis) have increased dramatically during the last 3 to 4 decades (29, 53). The increased number of infections from nontyphoidal Salmonella may result from modern intensified farming and food production methods and global trade. Increased spread of Salmonella may also be promoted by the acquisition of genes for antibiotic resistance (102), and in the case of S. Enteritidis, genes permitting colonization of chicken ovaries (49). Globally, it is estimated that there are 93.8 million cases of gastroenteritis due to Salmonella annually, out of which 80.3 million (86%) cases are foodborne (76). In the United States, salmonellosis is the leading cause of foodborne bacterial disease, with approximately 1.4 million human cases each year, resulting in 17,000 hospitalizations, 585 deaths (28,116) and a cost of 2.6 billion dollars due to loss of work, medical care and loss of life (112). Therefore, it is

14 4 imperative to study the origins, transmission and epidemiology of this pathogen in order to control and prevent diseases in the future Salmonella Salmonella is one of the most well-known and frequent foodborne bacterial pathogens throughout the world (76). Salmonella is a genus of rod-shaped, gram negative, non-spore forming, facultative anaerobic and motile bacteria belonging to the family Enterobacteriaceae Salmonella taxonomy and serotyping The genus Salmonella is comprised of two species: S. enterica and S. bongori. The species S. bongori is rarely associated with human disease. The species S. enterica has six subspecies: enterica, salamae, arizonae, diarizonae, houtenae and indica (63, 107). S. enterica subspecies enterica is responsible for 99% of the human cases of salmonellosis, so it is of greatest clinical importance (2). Salmonella subspecies are further differentiated based on serotyping. Serotyping distinguishes Salmonella immunologically based upon O antigens (lipopolysaccharide) and H antigens (peritrichous flagella). There are more than 2,500 recognized S. enterica serovars, each with a unique combination of O and H antigens (54). Prior to 2000, serovars were sometimes used as species names (16). For example, the original S. typhimurium is now referred to as S. enterica subspecies enterica serovar Typhimurium or simply S. Typhimurium. The latter nomenclature is used more commonly in publications and public health surveillance programs such as those administrated by the Centers for Disease Control and Prevention (CDC).

15 Evolution of pathogenicity S. enterica subspecies enterica was proposed to evolve in 3 main steps (Fig. 2.1) (7). The first step involved acquisition of Salmonella pathogenicity island 1 (SPI1) which contributed to the divergence of Salmonella from E. coli and other related organisms. SPI1 is a 40 kb DNA region present in both S. enterica and S. bongori (78). It encodes a type III secretion system (T3SS) required for the intestinal phase of infection and promotes inflammation, the invasion of intestinal epithelial cells, and secretion of intestinal fluid (117). The second step of evolution was hypothesized to be the acquisition of a second pathogenicity island SPI2 in the species S. enterica but not in S. bongori (Fig. 2.1) (7). SPI2 encodes another T3SS and various effector proteins that are required for survival and replication inside host cells during systemic infection (86, 97). For example, one of the many SPI2 effector proteins, SseL, is involved in macrophage killing, thus promoting survival inside the host (95). Due to the presence of SPI2, S. enterica has increased capacity for systemic spread and is thus more virulent than S. bongori, which do not contain SPI2. Finally, the host range of S. enterica subspecies enterica expanded to warm-blooded animals, including humans (Fig. 2.1) (7). In contrast, the other five S. enterica subspecies and S. bongori are mainly associated with cold-blooded animals. The expansion of host range to warmblooded animals requires that bacteria recognize the new hosts for the first step of infection. Recognition and attachment to the host involves adherence and colonization factors called adhesins. For example, fimbrial adhesin encoded by the gene fimh allows Salmonella to recognize and adhere to different receptors on host cells (66, 99). Genetic changes of this gene by point mutation or recombination might allow the subspecies enterica to recognize new receptors in new hosts, thus helping to expand its host range. After recognition and attachment, other processes allowing the subspecies enterica to infect warm blooded animals may include the ability to survive the immune system and proliferating inside host cells (7). It is not clear which

16 6 genetic changes accounted for these processes during adaptation to new hosts because adaptation to a new animal host is a complex process that probably involves a large number of genes. In summary, acquisition of SPI1 separated the genus Salmonella from other related organisms like E. coli. Then, acquisition of SPI2 separated the genus Salmonella into two distinct lineages, S. bongori and S. enterica. Finally, the lineage of S. enterica branched into several distinct phylogenetic groups. This latter phase of evolution was characterized by host range expansion of the subspecies enterica to warm-blooded animals, including humans. Through all these evolutionary steps, Salmonella enterica subspecies enterica (hereafter referred to as Salmonella) became a highly successful human and animal pathogen. Figure 2.1 Model for the three-phase evolution of pathogenicity in Salmonella enterica subspecies enterica. The phylogenetic tree is not drawn to scale (7) Salmonella reservoirs Salmonella is mostly transmitted through the fecal-oral route. Salmonellosis occurs when humans consume foods or water contaminated by animal and human feces containing Salmonella during food-handling or harvesting. Therefore, foods serve as the main transmission vector for

17 7 Salmonella, which include animal foods that are not thoroughly cooked and contaminated uncooked vegetables and fruits (116). Generally speaking, transmission of Salmonella starts from its reservoirs, which are defined as any person, animal, plant, soil or substance (or combination of these) in which a microorganism normally lives and grows (67). Salmonella serovars have adapted to live in a variety of hosts. Many wild animals, such as gorillas (10), rhinoceros (68), lizards (88), reptiles and snakes (9) harbor Salmonella. More importantly, food animals including chickens, turkeys, cattle, swine and sheep have also been found to frequently carry Salmonella. Different serovars have different reservoirs and modes of pathogenesis. For example, S. Typhi, which causes the deadly disease typhoid fever, is a strict human pathogen. Some other serovars, such as S. Gallinarum in chickens, S. Choleraesuis in swine and S. Dublin in cattle, are known to be associated mainly with one animal, but rarely cause disease in humans. In contrast, other serovars like S. Typhimurium have adapted to a broad host range, including wild and domestic animals and humans. Moreover, different animals have different predominant serovars associated with them. Predominant serovars associated with poultry, cattle and swine will be reviewed here in brief because those animals are the primary vectors for transmitting Salmonella to humans and are the main focus of this study. The most prevalent and important reservoirs for Salmonella are poultry (23). The most common poultry-associated serovars, Enteritidis in eggs and Typhimurium in poultry, accounted for 33.3 % of the total human foodborne diseases in the U.S. (20). The top 5 most common serovars associated with broilers are Kentucky, Heidelberg, Enteritidis, Typhimurium and I 4, [5], 12: i: - (113). They represent 81% of all Salmonella isolates from broilers. Similarly, serovars Hadar, Heidelberg, Reading, Schwarzengrund, and Saintpaul account for 68% of all Salmonella isolates from turkeys (113).

18 8 Cattle are also frequently found to harbor Salmonella. They can carry many different serovars of Salmonella, with Montevideo, Anatum, Muenster, Newport, Mbandanka the most common serovars that account for 47 % of Salmonella isolates from cattle (114). As for swine, another important reservoir for Salmonella, the 5 most frequent serovars are Derby, Typhimurium, Infantis, Anatum and Saintpaul. These 5 serovars comprise 60% of all isolates from swine (114). It is noteworthy that most of these serovars found predominantly in food animals are the same serovars that are frequently associated with human diseases. Given this fact, it is of great importance to control and monitor levels of the most common serovars in animals and subsequently prevent their transmission to humans Salmonella association with foods Another important vehicle for transmitting Salmonella to humans is produce. Salmonella can cycle through the food chain and the environment in soil, water, manure, and insects. Therefore, contamination of produce can occur by various ways throughout the food system. Like predominant serovars in animals, there are also predominant produce-associated serovars, which include Enteritidis, Newport, Poona, Typhimurium, Braenderup, Javiana, Montevideo and Muenchen (60). The overlap between serovars most commonly associated with animals and those associated with produce suggests contamination of produce during growing or harvesting processes directly or indirectly by animals containing Salmonella. Moreover, evidence is accumulating that enteric bacteria have the ability to grow and persist on and in plants, such as tomatoes, radish sprouts, bean sprouts, barley, and lettuce (15, 47, 62). Contamination and persistence of Salmonella on produce promote the transmission of this pathogen to humans. Salmonella outbreaks associated with fresh produce have increased in the U.S in recent years (98). Many kinds of produce have been linked to Salmonella outbreaks,

19 9 such as tomatoes, sprouts, melons, cantaloupe, lettuce, peppers and mangos (98). Produce causes the highest number of human diseases and second highest number of outbreaks among various food vehicles in the U.S. (3). For example, the largest Salmonella outbreak to date occurred in 2008 and was caused by consumption of Jalapeño and Serrano peppers that were contaminated with S. Saintpaul (22). Besides foods of animal origin and produce, there has been an increase in Salmonella outbreaks caused by new food vehicles, such as salami, peanut butter, veggie booty, pot pies, and dry cereals. For instance, in 2010, Italian-style salami and its ingredients (red and black peppers containing S. Montevideo) caused a multistate outbreak which infected 252 people from 44 states (27). As a result, approximately 1,378,754 pounds of Italian sausage products were recalled by Daniele International, Inc. (27). Another recent outbreak caused by a new food vehicle is the peanut butter outbreak, which infected 714 people from 46 states and caused 6 deaths (24). As a result, more than 2,100 peanut-containing products were recalled by over 200 companies. Outbreaks due to those new food vehicles were not expected because they are more or less processed foods which do not possess conditions that permit the growth of Salmonella. For example, peanut butter is a dry food with an a w below the minimum level for growth (0.94). Moreover, Salmonella can be inhibited or killed by heat, acid, high salt concentration, etc. during food manufacturing processes (38). Persistence of Salmonella in processed foods might be due to 1) high levels of Salmonella in food ingredients; 2) inadequate sanitary practices; 3) and the ubiquity of Salmonella in animals, produce and the environment Most common Salmonella serovars associated with human illnesses Although there are over 2,500 Salmonella serovars, only a handful of Salmonella serovars caused most human illnesses (Tables 2.1 and 2.2) (20, 21).

20 10 Table 2.1 Top ten most frequently reported serovars from human sources in 2005 Rank Serovar No. of laboratory-confirmed cases % of total cases 1 Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12: i : Montevideo Muenchen Saintpaul Braenderup total 66 Laboratory-confirmed cases include both outbreak cases and sporadic cases. Source: 2005 Salmonella annual review (20). Table 2.2 Top ten most frequently reported serovars from human sources in 2006 Rank Serovar No. of laboratory-confirmed cases % of total cases 1 Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12: i : Montevideo Muenchen Oranienburg Mississippi total 60 Laboratory-confirmed cases include both outbreak cases and sporadic cases. Source: 2006 Salmonella annual review (21).

21 11 Compared to all the other serovars of Salmonella, S. Typhimurium caused the highest number of human illnesses and was associated with a broad range of foods (Table 2.3). As mentioned before, S. Typhimurium has adapted to various hosts, including birds, amphibians, and all food animals, especially poultry, cattle and swine. Not only can S. Typhimurium reside in so many animals, but it can also be found in them at high frequency (114). The ubiquity and relatively high numbers of S. Typhimurium might explain why it has caused so many outbreaks via so many kinds of foods (Table 2.3). The second most common serovar is S. Enteritidis, which caused nearly as many human cases as S. Typhimurium (Tables 2.1 and 2.2). The major food vehicles for S. Enteritidis are shell eggs, as 80% of the S. Enteritidis outbreaks were egg-associated (89). S. Enteritidis contaminates eggs either through horizontal transmission, by which eggs are externally contaminated by feces containing S. Enteritidis (36), or by vertical transmission, where the inside of the eggs is contaminated by infected ovaries before the laying of the egg (50, 87). Vertical transmission is believed to be the more important route because eggs contaminated by vertical transmission produce a new generation of infected broilers or layers after hatching (50, 57, 79). In order to control S. Enteritidis in poultry, one of the interventions employed in the U.S. is egg quality assurance programs on farms. These voluntary programs involve acquisition of S. Enteritidis free chicks, control of pests (including rodents and flies), use of S. Enteritidis-free feeds, and routine microbiologic testing for S. Enteritidis in the farm environment (14). The third most commonly reported serovar causing salmonellosis is S. Newport (Tables 2.1 and 2.2). S. Newport can be detected in many food animals, but is most frequently isolated from cattle (113). S. Newport has been implicated in many outbreaks via a variety of food vehicles, such as beef, chicken, pork, tomatoes, cantaloupes, melons, avocadoes and guacamole

22 12 (23). In 2010, S. Newport caused a multistate outbreak due to contaminated alfalfa sprouts, in which 35 people became ill (26). Cases of illness caused by S. Newport have increased in recent years, which might be due to the emerging multidrug-resistant S. Newport isolates (19). The fourth most common serovar is S. Heidelberg (Tables 2.1 and 2.2). It is often isolated from commercial broilers and ground chicken (113). As a result, poultry and eggs have been identified as the major food vehicles for this serovar (32). The largest outbreak caused by S. Heidelberg occurred in 2007, when 802 people became infected via contaminated hummus (Table 2.3). Following S. Heidelberg, S. Javiana caused the fifth most human infections (Tables 2.1 and 2.2). Unlike other serovars, S. Javiana is rarely isolated from poultry, cattle or swine (113). The major reservoirs for S. Javiana were considered to be amphibians, as direct contact with amphibians has been associated with outbreaks. Amphibian feces-contaminated tomatoes were identified to be the main food vehicles for S. Javiana (34). For example, tomatoes were identified to be the food source of S. Javiana for a multistate outbreak in 2002, which resulted in 159 cases (Table 2.3). The sixth most common serovar I 4, [5], 12: i :-, a variant of serovar S. Typhimurium, is antigenically similar to S. Typhimurium, but lacks the second-phase flagella antigens (39). It is also one of the most commonly identified serovar in broilers and ground chicken (113). I 4, [5], 12: i :- contaminated pot pies caused a multistate outbreak in 2007 (Table 2.3). S. Montevideo is the next most commonly reported serovar. S. Montevideo is frequently isolated from cattle and ground beef (113). Food vehicles of S. Montevideo include beef, turkey, pork and sprouts (22). The most recent outbreak caused by S. Montevideo occurred in 2010 due to contaminated Italian-style meats (27). The eighth most common serovar is S. Muenchen. S. Muenchen can be detected in swine, cattle, chicken etc. It has been associated with outbreaks due to multiple food vehicles, such as

23 13 chicken, sprouts, tomato, and cantaloupe (22). In 1999, a multistate outbreak was caused by S. Muenchen in orange juice, which infected 398 people. S. Saintpaul ranks as the ninth most common serovar in 2005, but dropped to eleventh in 2006 (20, 21). However, its ranking might have risen higher since then, because it caused the largest Salmonella outbreak in 2008 due to contaminated peppers. S. Saintpaul is frequently isolated from swine and has caused outbreaks due to foods like sprouts, tomatoes, mangoes, orange juice, turkey etc. The importance of the above top serovars is reflected by the high number of salmonellosis cases they cause. Their success as human pathogens might be largely due to adaptation to food animals. For example, 4 of the top 8 serovars are frequently found in poultry, namely Typhimurium, Enteritidis, Heidelberg and I 4, [5], 12: i :-. Two other serovars, Newport and Montevideo, are mainly found in cattle.

24 14 Table 2.3 Salmonella outbreaks caused by the top 8 serovars in the United States from Year Serovar Ill Hospitalizations Deaths Food vehicle 2008 Typhimurium peanut butter 2001 Typhimurium unidentified 2006 Typhimurium deli meat 2006 Typhimurium tomato 2005 Typhimurium sauces; fajita 2006 Typhimurium chicken 1998 Typhimurium multiple foods 2002 Typhimurium unidentified 2002 Typhimurium milk 1999 Typhimurium clover sprouts 2002 Typhimurium milk 2007 Typhimurium Veggie Booty 2007 Typhimurium lettuce; spinach 2003 Typhimurium eggs 2007 Typhimurium pork 2003 Typhimurium beef 2005 Typhimurium cake 2003 Typhimurium ground beef 1998 Typhimurium smoked fish 2003 Typhimurium queso fresco 2002 Enteritidis salsa 2005 Enteritidis turkey 1999 Enteritidis ice cream 2001 Enteritidis egg-based sauce 2002 Enteritidis cake 2005 Enteritidis cantaloupe 2006 Enteritidis oil; chicken 2001 Enteritidis eggs 2000 Enteritidis macaroni cheese 2007 Enteritidis chicken 2003 Enteritidis crab cakes 2001 Enteritidis eggs 2002 Enteritidis beef; pork 2000 Enteritidis 88 orange juice 1999 Enteritidis honeydew melon 2002 Newport 510 tomato 2006 Newport tomato 2004 Newport milk 2004 Newport 97 lettuce 2000 Newport pico de gallo 1999 Newport 79 mango 2006 Newport turkey 2003 Newport honeydew melon 2007 Newport pork 2007 Newport tomato 2004 Newport turkey and gravy

25 Newport ground beef 2007 Newport 46 tomato; avocado 2007 Heidelberg hummus 2003 Heidelberg 517 chicken 2002 Heidelberg beef 1998 Heidelberg cake 2002 Heidelberg macaroni cheese 2007 Heidelberg 79 mashed potato 2004 Heidelberg turkey 2005 Heidelberg sandwich; vanilla cake 2003 Heidelberg Swiss cheese 2003 Heidelberg eggs; pancakes 2000 Heidelberg macaroni salad 1999 Heidelberg 41 chicken 2003 Javiana fajita, chicken 2002 Javiana tomato 2004 Javiana beans 2000 Javiana bread; chicken 2007 I 4,[5],12:i : pot pie 2010 Montevideo 252 Italian-style meats 2006 Montevideo sandwich, beef 2002 Montevideo beef 1999 Muenchen 398 orange juice 1999 Muenchen alfalfa sprouts 2003 Muenchen cantaloupe 2002 Muenchen pasta salad 2005 Saintpaul ; Typhimurium 157 orange juice 2008 Saintpaul peppers 2009 Saintpaul 235 alfalfa sprouts Source: CDC foodborne outbreak database (23). 2.2 Subtyping of Salmonella In order to control Salmonella outbreaks, it is important to trace back the sources and identify the routes by which Salmonella are transmitted to foods. However, trace-back investigation of outbreaks can be hindered due to the complexity of the food chain and the limitations of traditional epidemiologic investigations. The limitations of traditional epidemiologic investigations include 1) Only a limited number of cases are reported; 2) People tend not to recall the foods that were eaten before disease onset; 3) Cases are often spread out in

26 16 time and space; and 4) Investigations can be hindered if the food source is not listed on the investigation questionnaire (60). Based on the reasons above, another trace-back method called subtyping is carried out along with traditional epidemiologic investigations. Subtyping characterizes bacteria at the strain level (101). By characterizing the outbreak-related strains and separating them from non-related strains, subtyping can play an essential role in investigating Salmonella outbreaks. Besides tracking pathogens in epidemiologic investigations, the other use of subtyping methods is to study the population structure, evolution and diversity of bacteria on a long-term scale. For example, one subtyping method called multilocus enzyme electrophoresis (MLEE) has been used to study the genetic diversity of Salmonella populations (8). Studies like this can provide insight into the evolutionary history and emergence of Salmonella serovars. However, the focus of this review is on the short-term epidemiologic applications of subtyping methods Important definitions and performance criteria of subtyping methods Before considering the epidemiology of Salmonella, it is important to first clarify the definitions for outbreak, epidemic, strain, epidemic clone (EC), and outbreak clone (OC) used frequently in epidemiologic studies. These definitions were previously compiled by Chen and Knabel (30). Outbreak is an acute appearance of a cluster of an illness that occurs in numbers in excess of what is expected for that time and place. Epidemic is defined as one or more outbreaks that spread widely over a long period of time. Strain is defined as isolates that have distinct phenotypic and genotypic characteristics from other isolates from the same species. Epidemic clone is a strain or group of strains descended asexually from a single ancestral cell (source strain) that is involved in one epidemic, and can often include several outbreaks. Outbreak clone is a strain or group of strains descended asexually from a single ancestral cell (source strain) that is involved in one outbreak (30).

27 17 To evaluate and compare different subtyping schemes, there are several performance criteria, which include typeability, reproducibility, discriminatory power and epidemiologic concordance. Typeability is the capability of a method to generate an interpretable result for each strain typed. For example, strains that do not have plasmids cannot be typed by plasmid profiles. Reproducibility is the ability of a subtyping method to generate the same result each time the sample is tested. Discriminatory power is the ability of a subtyping method to differentiate between unrelated epidemic or outbreak clones. Epidemiologic concordance is the capacity of a typing method to correctly cluster epidemic and outbreak clones, and separate them from clones that are not epidemiologically related (101). Many studies of subtyping methods focused on the discriminatory power of the subtyping system. On the other hand, few studies have examined the epidemiologic concordance of a particular subtyping method. The reason for the lack of studies examining epidemiologic concordance might be that most studies did not utilize well-defined strains from multiple outbreaks. The choice of strain collection is critical when developing and evaluating a new subtyping system for outbreak investigations. As mentioned before, an ideal strain collection should include well-defined strains from multiple common-source outbreaks in order to access both discriminatory power and epidemiologic concordance. A good subtyping system should separate strains from different outbreaks, but not separate strains within the same outbreak/outbreak clone Salmonella subtyping methods during epidemiologic investigations Subtyping methods can be either phenotypic or genotypic approaches. Phenotypic methods include screening for antibiotic resistance, bacteriophage susceptibility and surface antigens, such as the H and O antigens. Genotypic methods differentiate strains based on differences in genome sequence and/or structure. Major phenotypic and genotypic subtyping

28 18 methods available for Salmonella will be briefly discussed here with the primary focus on genotypic methods Phenotypic methods Before the advent of genotypic methods, many phenotypic methods were widely used for typing Salmonella strains. Common phenotypic methods for Salmonella include serotyping, phage typing and MLEE. In general, although phenotypic methods provide useful information about the strains, they often lack enough discriminatory power Serotyping As mentioned in the taxonomy section, serotyping distinguishes Salmonella based on immunological classification of the H and O antigens (54). Serotyping is one of the most important phenotypic methods for Salmonella, which provides baseline information before other typing methods can be carried out to further separate strains in a particular serovar. Serotyping is very useful because the serovar name often points to the specific reservoir and mode of pathogenesis. However, serotyping alone is not suit for molecular epidemiology, because individual serovars are responsible for multiple outbreaks (20, 21). As a result, other subtyping methods with more resolution need to be carried out after serotyping Phage typing Phage typing utilizes the selective capacity of individual bacteriophage to infect bacterial cells. During phage typing, a panel of bacteriophages is used to infect bacteria and phage types are assigned according to the patterns of lysis. Phage typing has been shown to be a good

29 19 indicator for pandemic clones of Salmonella. For instance, S. Enteritidis phage type (PT) 4 is the most common PT in Europe, while PT8 is the most common PT in the U.S. Another example is S. Typhimurium definitive type 104 (DT104), which is typically resistant to a number of antibiotics and has had a major impact on global health (106). However, phage typing sometimes suffers from low typeability in that many strains are resistant to all typing phages (1). Moreover, it requires maintenance of the typing phage stocks and specially trained personnel (45) Multilocus enzyme electrophoresis (MLEE) MLEE differentiates strains based on the relative electrophoretic mobility of cellular enzymes. The variation in amino acid sequences of the enzymes from different strains results in differences in electrostatic charges. This leads to different migrations of the enzymes in an electric field. By comparing the electrophoretic profiles, genetic relatedness of strains can then be determined. MLEE has been carried out to analyze the population structure of Salmonella serovars and the relatedness of strains within a serovar (8). Population studies by MLEE subtyping revealed that while many serovars have similar electrophoretic types (ETs) that form a single cluster, other serovars like S. Newport have divergent ETs clustered distantly in MLEE trees. Using MLEE to determine phylogenetic relationships of bacteria is generally accepted. However, MLEE has been replaced by a more reproducible and portable method called multilocus sequence typing (MLST), which looks directly at DNA sequences of several genes (75). MLST will be introduced later as one of the genotypic methods Genotypic methods Genotypic methods target genetic differences between different strains of bacteria. Generally speaking, genotypic methods have better reproducibility and increased discriminatory

30 20 power than phenotypic methods. Because of these advantages, genotypic methods are often carried out after serotyping during Salmonella outbreak investigations. Two categories of genotypic methods, DNA-fragment-pattern-based methods and DNA-sequence-based methods, will be discussed DNA-fragment-pattern-based methods Three DNA-fragment-pattern-based subtyping methods have been extensively studied for subtyping Salmonella, which are pulsed-field gel electrophoresis (PFGE), amplified fragment length polymorphism (AFLP) and multiple loci variable number tandem repeat analysis (MLVA) Pulsed-Field Gel Electrophoresis (PFGE) PFGE is currently the gold standard method for subtyping Salmonella and is used by public health surveillance systems such as the PulseNet program of CDC. During PFGE procedures, bacterial cells are first immobilized in agarose plugs to avoid mechanical shearing of the long genomic DNA. Cells in agarose plugs are then lysed and genomic DNA is digested by a rare-cutting restriction endonuclease. Next, agarose plugs containing digested genomic DNA are put into wells of an agarose gel. The agarose gel is then subjected to an electric field whose orientation is periodically changing. This pulsed electrical field can resolve large DNA fragments that could not be separated by a constant unidirectional electrical field. The standardized PFGE protocol of Salmonella uses two restriction endonucleases XbaI and BlnI in separate reactions (40). PFGE has been used in detection, investigation and control of numerous outbreaks and is generally very successful (51). The main advantage of PFGE is its comparatively high discriminatory power for subtyping most serovars of Salmonella. However, PFGE lacks

31 21 discriminatory power for clonal serovars like Enteritidis (25, 120) and Montevideo (27), or clonal phage types like S. Typhimurium DT104 (51). This is reflected by low PFGE pattern diversity for those serovars and clonal phage types in the PulseNet database (51). In the cases of such low discriminatory power, outbreak clones cannot be separated from sporadic isolates and other nonoutbreak related isolates, which can hinder epidemiologic detection and investigation. For example, during the recent Italian-style meat outbreak, the outbreak clone of S. Montevideo had the most common PFGE pattern in PulseNet database, which made it difficult to detect the outbreak (27). Besides low discriminatory power for clonal serovars, another limitation of PFGE is the ambiguous interpretation of banding patterns. Banding patterns can change due to insertions, deletions and point mutations. For instance, a single nucleotide mutation might cause up to 3- fragment changes in the PFGE banding pattern. Because of this difficulty, interpretation of PFGE banding patterns has been proposed to follow several guidelines: 1) strains showing no fragment differences with the outbreak strain are part of the outbreak; 2) strains showing 1 fragment difference with the outbreak strain are probably part of the outbreak; 3) strains showing 2-3 fragment differences with the outbreak strain are possibly part of the outbreak; 4) strains showing more than 3-fragment differences with the outbreak strain are not part of the outbreak (105). More recommendations for interpretation of PFGE patterns have been published recently. The recommendations include taking into account the quality of the PFGE gel, the diversity of the organism and the temporal and geographical information during analysis of PFGE patterns (40). Although those suggestions helped standardize the interpretation of PFGE patterns, these recommendations are still not completely objective. Another drawback of PFGE is low reproducibility if the standardized protocol is not strictly followed. As a result, subsequent comparison of PFGE banding patterns cannot be carried out, especially when comparing PFGE patterns between different laboratories. To overcome this limitation, PulseNet implemented an extensive quality assurance system (51). This system

32 22 requires laboratories to obtain PFGE gel preparation and gel analysis certification and participate in the annual proficiency testing program. All these steps help ensure comparability and reproducibility, but at the same time it requires personnel specially trained by the quality assurance system. To sum up, although it is the current gold standard subtyping method, PFGE suffers from several drawbacks which limit its performance for subtyping Salmonella Amplified Fragment Length Polymorphism (AFLP) AFLP is a method that employs both restriction digestion and polymerase chain reaction (PCR) techniques. In AFLP, genomic DNA is digested with one or more restriction enzymes. The ends of the digested DNA fragments are then ligated to adaptors that are complementary to the restriction sites. The digested and ligated DNA fragments are then selectively amplified using PCR primers targeting the adaptor sequences. PCR primers typically contain one to three additional nucleotides on their 3 -end to reduce the number of amplified fragments to a manageable number. PCR products are then subjected to electrophoresis and characteristic banding patterns are then produced. AFLP is a relatively simple and fast approach. The discriminatory power of AFLP is equal to that of PFGE for subtyping S. Typhimurium (73, 103), but higher than that of PFGE for subtyping S. Enteritidis (52) and other serovars (109). However, its discriminatory power has been reported to be insufficient to separate all epidemiologically unrelated S. Typhimurium strains (92). Like PFGE, the reproducibility of AFLP among different laboratories is problematic since comparing AFLP results among different laboratories is difficult (48). Variability in the AFLP profile can be generated by minor changes in the amplification conditions. Therefore, replicates of the sample could be identified as different strains (45). To enhance reproducibility,

33 23 PCR should be performed under highly stringent conditions (84) and gel electrophoresis should be standardized Multiple Loci Variable number tandem repeat Analysis (MLVA) MLVA targets tandem repeats of short DNA sequences in bacterial genomes. The difference in the number of repeated DNA motifs is employed to differentiate strains. In a MLVA assay, a number of well-selected and characterized loci are amplified by PCR using primers targeting the flanking regions of the repeated loci. PCR products are then separated and the number of repeat units at each locus can be measured according to the size of the PCR products. Differences in the number of repeats in each locus are used to distinguish different strains. Since this method is based on PCR, MLVA has the advantage of being easy to perform and rapid. Moreover, MLVA yields discreet and unambiguous data, reported as the number of repeat units at each locus. Comparison of MLVA profiles between laboratories can be made with a simple nomenclature recently proposed (70). The discriminatory power of MLVA was reported to be higher than PFGE and AFLP for subtyping S. Typhimurium (72, 108) and higher than PFGE for S. Enteritidis (11, 93). However, in some circumstances, strains that have the same MLVA type were separated by PFGE profiles (13). This indicates that strains of same MLVA type might not be closely related. However, the reproducibility of MLVA is a potential problem. The instability of MLVA alleles has been observed for subtyping S. Newport and S. Typhimurium (18, 35). Replicates of the same strains have been shown to have different number of repeat units at a specific locus (35). The instability of the MLVA loci is probably due to DNA polymerase slippage during genome replication (110). This instability might make interpretation difficult when strains have slightly different MLVA types.

34 24 To conclude, by providing improved discriminatory power and having a short turnaround time, MLVA can be used as a complementary method to PFGE in epidemiologic investigations of Salmonella. MLVA has been used successfully along with other subtyping methods in outbreak investigations to track Salmonella (12, 83, 85). However, MLVA also suffered from some drawbacks and thus it has not been widely used for this purpose DNA-sequence-based methods DNA-sequence-based methods differentiate strains by the detection of polymorphic DNA sequences. Multilocus sequence typing (MLST) and single nucleotide polymorphism (SNP) analysis are both DNA-sequence-based methods and will be briefly reviewed here Multilocus Sequence Typing (MLST) MLST discriminates among bacterial strains by comparing nucleotide sequences of several DNA loci in bacteria chromosomes. For each locus in the MLST scheme, every new allele is assigned a unique number in order of discovery and is designated an allelic type. The collective allelic types make up the allelic profile or sequence type, which may also be assigned a unique and arbitrary number. For example, in the MLST database ( based on the seven loci: aroc, dnan, hemd, hisd, pure, suca, and thra, one of the strains in the database has an allelic profile of (1, 1, 2, 1, 1, 1, 9) for each of the seven genes, and was assigned sequence type 3 (80). The collective allelic types and sequence types are compared among bacterial strains and then cluster analysis can be carried out. Compared to PFGE, MLST is a less labor-intensive method and involves common techniques including primer design, PCR amplification and DNA sequencing. Furthermore, DNA sequence represents discreet, unambiguous, highly informative, highly portable and

35 25 reproducible data. Many MLST data sets are available over the internet ( so that a uniform nomenclature is ensured and comparison of results among laboratories can be conducted rapidly. The application of MLST is promoted due to the increased speed and reduced cost of nucleotide sequencing and improved internet database and tools (74). These advantages make MLST an attractive subtyping approach. MLST schemes originally target housekeeping genes, which are genes required for fundamental metabolic functions and are found within all members of a given species (75). For example, 7 housekeeping genes were targeted in the first MLST scheme for Neisseria meningitidis (75). Housekeeping genes are excellent genetic markers for studying the population structure, long-term evolution and diversity of bacteria. A good overview of Salmonella diversity and evolution is provided by the internet-based MLST data. Based upon MLST data, Salmonella, especially S. enterica subspecies enterica, is highly clonal (69). Moreover, the data suggest that many serovars including Typhimurium, Enteritidis, Newport and Saintpaul may have more than one origin (69). However, MLST schemes based on housekeeping genes for typing Salmonella usually have much lower discriminatory power than that of PFGE (43, 61, 109). The results of those studies suggested that housekeeping genes do not provide sufficient resolution to distinguish closely related strains. Therefore, MLST schemes based on housekeeping genes are not suitable for outbreak investigations. To conclude, MLST possesses many attractive advantages. It is an excellent tool for global phylogenetic studies. However, housekeeping genes selected in previous MLST studies lacked sequence variation and thus were ineffective for subtyping Salmonella for epidemiologic purposes. To track strains of this important pathogen during outbreaks, genetic markers that give sufficient DNA sequence variations need to be identified.

36 Multi-Virulence-Locus Sequence Typing (MVLST) Besides housekeeping genes, virulence genes which are responsible for pathogenesis, have been selected as genetic markers for MLST schemes. MLST schemes that only target virulence genes have been referred to as multi-virulence-locus sequence typing (MVLST) (31, 119). Unlike housekeeping genes, virulence genes are commonly under positive selection (41). As a result, DNA sequences of virulence genes tend to be more variable than housekeeping genes and thus are able to provide increased discrimination. It is also speculated that virulence genes can provide high epidemiologic concordance because they are responsible for causing diseases and thus outbreaks. For example, six virulence genes were targeted in an MVLST scheme for subtyping Listeria monocytogenes, which showed very high discriminatory power (0.99) and perfect epidemiologic concordance (1.0) (31). No MVLST scheme has yet been developed for subtyping Salmonella. However, MLST based on both virulence genes and housekeeping genes has been published for typing Salmonella enterica subspecies enterica serovars, which targeted flagellin genes flic and fljb along with two housekeeping genes, gyrb and atpd (104). This study included several strains from all subspecies and 22 of the more prevalent Salmonella enterica subspecies enterica serovars attempting to develop a DNA-based assay for serotype identification. However, the use of this MLST scheme to further characterize strains under serovar level was not tested. Another MLST based on both virulence genes and housekeeping genes has been developed for subtyping S. Typhimurium and showed high discriminatory power (0.98), which was slightly higher than that of PFGE (0.96) (46). In that MLST scheme, three virulence genes were included together with the 16S rrna gene and three housekeeping genes. One of the virulence genes in that MLST scheme is hila which regulates transcription of invasion proteins (4). The other two virulence genes, pefb and fimh, encode different fimbriae and both mediate adherence to host cells (6, 66). Although this MLST scheme seems to have adequate discriminatory power for subtyping S.

37 27 Typhimurium, its capacity to discriminate strains from more clonal serovars such as S. Enteritidis has not yet been tested. Currently, there is no published MLST study for differentiating strains within S. Enteritidis. In the SNP database of NCBI (National Center for Biotechnology Information), two strains of S. Enteritidis were compared side by side to examine their SNPs (56). Nearly all virulence genes were identical between the two, suggesting that MVLST might not be discriminatory enough for differentiating strains of S. Enteritidis. In summary, although MVLST has higher discriminatory power than MLST using housekeeping genes, it may not provide enough discrimination for clonal serovars like Enteritidis. In order to develop an MLST scheme for outbreak investigations, additional genetic markers with even higher sequence variability need to be identified Single Nucleotide Polymorphism (SNP) analysis SNP analysis differentiates strains by nucleotide substitutions at specific sites in the bacterial genome. SNP analysis often involves three steps: 1) Select SNP sites that are variable to provide discrimination among strains; 2) Determine the nucleotide bases at the selected sites of different strains; and 3) Compare the SNPs among strains. Selection of the SNP sites is often based on previous knowledge of specific polymorphic genes (42, 71) or comparative genomic studies (118). To determine the nucleotide base (adenine, guanine, cytosine, and thymine) at a defined SNP site, multiple methods can be used, such as pyrosequencing or realtime PCR (82, 91, 111). Because SNP analysis targets SNPs in the bacterial genome, it has the potential to be more rapid and cost efficient than MLST. However, there are very few SNP analysis studies for subtyping Salmonella. SNP analysis targeting genes associated with quinolone resistance has been used to study the antibiotic resistance of Salmonella (42, 71). Another SNP analysis study targeted SNPs in flagella antigens in order to develop a SNP typing method to replace serotyping

38 28 (82). No SNP typing methods have been developed for differentiating Salmonella strains for outbreak investigations. The reason might be that the SNP loci of Salmonella that could provide the desirable discrimination have not been identified. In conclusion, although SNP analysis has the potential to be rapid, cost efficient and high-throughput, the lack of information about SNP sites suitable for subtyping Salmonella make it difficult to develop a SNP typing protocol for epidemiologic purposes. 2.3 Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR) Since virulence genes alone might not provide enough discrimination for subtyping clonal Salmonella serovars, additional genetic elements that are evolving faster than virulence genes are needed. One of the fastest evolving genetic elements in bacteria genomes are CRISPRs (Clustered Regularly Interspaced Palindromic Repeat) (100). CRISPRs are regions of direct repeats (DRs) and spacers in the chromosomes of archaea and bacteria, including S. enterica (Fig. 2.2) (65, 100). DRs are bp long, separated by spacers of similar size (Fig. 2.2). Sequences of DRs are generally conserved, except the repeat at one end of the CRISPR is not totally conserved, and is thus called a degenerate direct repeat (Fig. 2.2). On the other hand, sequences of spacers are quite variable from each other. It was recently demonstrated that CRISPR spacers are derived from phages or plasmids, which when inserted into the CRISPR of a bacterial cell help protect that cell from subsequent infection by those same phages and plasmids (5). CRISPR is generally flanked at one end by a common leader sequence of bp, which is believed to act as a promoter to transcribe CRISPR into small RNAs (77). Immediately upstream from the CRISPR there are CRISPR-associated (Cas) proteins that carry functional domains of nucleases, helicases, polymerases and polynucleotide-binding proteins (58). Some Cas proteins can recognize foreign DNA invading the bacteria, and then integrate a new repeat-spacer unit into CRISPR at the leader end. Therefore, when the same exogenous nucleic acid invades next time,

39 29 the CRISPR transcribed crrnas (CRISPR RNAs) can recognize the foreign nucleic acid and lead the Cas proteins to degrade these invading nucleic acid (17, 59, 65). In this way, CRISPR along with Cas proteins can block foreign sequences, such as sequences of phages and plasmids. CRISPR1 CRISPR2 Figure 2.2 Schematic view of the two CRISPR systems in Salmonella Typhimurium LT2. Direct repeats and spacers are represented by black diamonds and white rectangles, respectively. The degenerate direct repeats are represented by white diamonds. Numbers of direct repeats and spacers are represented by the numbers of diamonds and white rectangles, respectively. L stands for leader sequence. cas genes are in grey while other core flanking genes (ygcf, iap and ptps) are in white. The graph is not drawn to scale. As a bacterial immune system against phages and plasmids, CRISPRs evolve rapidly and adaptively (115). As mentioned before, new spacers could be added when foreign DNA invades the bacteria. Besides addition of new spacers, deletion of spacers is also frequently observed (37, 90). However, the mechanism of deletion of spacers is not clear. The addition of new spacers and deletion of one or several spacers make CRISPR one of the most variable DNA loci in bacteria and form a high degree of polymorphism among strains (90). CRISPRs have been used for subtyping Mycobacterium tuberculosis, this subtyping method is called Spacer oligotyping or spoligotyping (55). In this method, PCR is carried out using primers designed according to the sequence of the DR so that each spacer can be amplified. The PCR products are then hybridized to a membrane containing probes for specific spacers. The hybridization patterns showing the presence or absence of spacers are then compared among strains. Spoligotyping is now the standard method for subtyping M. tuberculosis for outbreak

40 30 investigations. It has also been used in subtyping Corynebacterium diphtheria (81). Other than spoligotyping, CRISPR sequence analysis has also been used for other bacteria, such as Yersinia pestis (90), Streptococcus (64), and Campylobacter jejuni (96). As for Salmonella, although CRISPRs have the potential to be excellent markers for separating Salmonella strains, they have not been widely used for subtyping purposes CRISPR in Salmonella CRISPR can be found in multiple numbers in bacteria. Two CRISPR loci are found in all Salmonella serovars in the CRISPRs database ( CRISPR direct repeats in Salmonella are bp long. Salmonella CRISPRs have great polymorphism even among strains belonging to the same serovar. Therefore, CRISPRs might serve as good markers for subtyping Salmonella during epidemiologic investigations. 2.4 Conclusions Salmonella is the leading cause of foodborne bacterial disease in the U.S. Most human illnesses are caused by a handful of serovars, such as Typhimurium, Enteritidis, Newport, Heidelberg, I 4, [5], 12; i: -, Montevideo, Muenchen and Saintpaul. Salmonella can reside in many wild and domestic animals and can spread from numerous reservoirs to contaminate numerous kinds of foods, which makes it especially challenging to track this pathogen during outbreaks. Therefore, to reduce outbreaks caused by the most common serovars of Salmonella, it is critical to employ a subtyping method that can accurately identify its sources and pathways of transmission. Many subtyping methods have been developed for differentiating Salmonella strains, such as PFGE, AFLP and MLVA. Each method has its own advantages and drawbacks. PFGE is currently the gold standard method for outbreak investigations. However, PFGE

41 31 produces ambiguous data that are hard to interpret and more importantly PFGE often lacks discriminatory power for subtyping clonal serovars such as Enteritidis. In contrast, MLST generates highly informative and discreet data consisting of nucleotide sequences that can be easily interpreted and rapidly compared on internet databases. Previous MLST schemes targeting housekeeping genes were not very successful largely due to low discriminatory power associated with conserved housekeeping genes. Unlike housekeeping genes, virulence genes can provide important information about the pathogenesis of strains and improve the discriminatory power of MLST. However, the discriminatory power of virulence genes may still not be enough for subtyping clonal serovars of Salmonella. CRISPRs are one of the fastest evolving genetic elements that could be implemented in an MLST scheme to provide increased discrimination. In order to develop an MLST scheme for outbreak investigation, virulence genes and CRISPRs were targeted in the present study to subtype the top 10 serovars of Salmonella. This MLST scheme was speculated to provide high discriminatory power and epidemiologic concordance for subtyping Salmonella for epidemiologic purposes. 2.5 References 1. Amavisit, P., P. F. Markham, D. Lightfoot, K. G. Whithear, and G. F. Browning Molecular epidemiology of Salmonella Heidelberg in an equine hospital. Vet. Microbiol. 80: Anjum, M. F., C. Marooney, M. Fookes, S. Baker, G. Dougan, A. Ivens, and M. J. Woodward Identification of Core and Variable Components of the Salmonella entericas subspecies I genome by microarray. Infect. Immun. 73: Anonymous Outbreak alert! Closing the gaps in our federal food safety net.

42 32 4. Bajaj, V., R. L. Lucas, C. Hwang, and C. A. Lee Co-ordinate regulation of Salmonella Typhimurium invasion genes by environmental and regulatory factors is mediated by control of hila expression. Mol. Microbiol.. 22: Barrangou, R., C. Fremaux, H. Deveau, M. Richards, P. Boyaval, S. Moineau, D. A. Romero, and P. Horvath CRISPR provides acquired resistance against viruses in prokaryotes. Science. 315: Baumler, A., R. Tsolis, F. Bowe, J. Kusters, S. Hoffmann, and F. Heffron The pef fimbrial operon of Salmonella Typhimurium mediates adhesion to murine small intestine and is necessary for fluid accumulation in the infant mouse. Infect. Immun. 64: Baumler, A. J., R. M. Tsolis, T. A. Ficht, and L. G. Adams Evolution of host adaptation in Salmonella enterica. Infect. Immun. 66: Beltran, P., J. M. Musser, R. Helmuth, J. J. Farmer, W. M. Frerichs, I. K. Wachsmuth, K. Ferris, A. C. McWhorter, J. G. Wells, and A. Cravioto Toward a population genetic analysis of Salmonella: genetic diversity and relationships among strains of serotypes S. Choleraesuis, S. Derby, S. Dublin, S. Enteritidis, S. Heidelberg, S. Infantis, S. Newport, and S. Typhimurium. Proc. Natl. Acad. Sci. U.S.A. 85: Bemis, D. A., L. M. Grupka, S. Liamthong, D. W. Fooland, J. M. Sykesiv, and E. C. Ramsay Clonal relatedness of Salmonella isolates associated with captive and wildcaught rattlesnakes. Vet. Microbiol Benirschke, K., and F. D. Adams Gorilla diseases and causes of death. J. Reprod. Fertil. 28: Beranek, A., C. Mikula, P. Rabold, D. Arnhold, C. Berghold, I. Lederer, F. Allerberger, and C. Kornschober Multiple-locus variable-number tandem repeat analysis for subtyping of Salmonella enterica subsp. enterica serovar Enteritidis. Int. J. Med. Microbiol. 299:43-51.

43 Best, E. L., M. D. Hampton, S. Ethelberg, E. Liebana, F. A. Clifton-Hadley, and E. J. Threlfall Drug-resistant Salmonella Typhimurium DT 120: use of PFGE and MLVA in a putative international outbreak investigation. Microb. Drug Resist.15: Boxrud, D., K. Pederson-Gulrud, J. Wotton, C. Medus, E. Lyszkowicz, J. Besser, and J. M. Bartkus Comparison of multiple-locus variable-number tandem repeat analysis, pulsed-field gel electrophoresis, and phage typing for subtype analysis of Salmonella enterica serotype Enteritidis. J. Clin. Microbiol. 45: Braden, C. R Salmonella enterica serotype Enteritidis and eggs: a national epidemic in the United States. Clin. Infect. Dis. 43: Brandl, M. T Fitness of human enteric pathogens on plants and implications for food safety. Annu. Rev. Phytopathol. 44: Brenner, F. W., R. G. Villar, F. J. Angulo, R. Tauxe, and B. Swaminathan Salmonella Nomenclature. J. Clin. Microbiol. 38: Brouns, S. J. J., M. M. Jore, M. Lundgren, E. R. Westra, R. J. H. Slijkhuis, A. P. L. Snijders, M. J. Dickman, K. S. Makarova, E. V. Koonin, and J. van der Oost Small CRISPR RNAs guide antiviral defense in prokaryotes. Science. 321: Call, D., L. Orfe, M. Davis, S. Lafrentz, and M. Kang Impact of compounding error on strategies for subtyping pathogenic bacteria. Foodborne Pathog. Dis.5: CDC Outbreak of multidrug-resistant Salmonella Newport--United States, January- April JAMA. 288: CDC Salmonella Annual Summary df

44 CDC Salmonella Annual Summary y2006.pdf 22. CDC Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce items --- United States, MMWR Morb Mortal Wkly Rep. 57: CDC OutbreakNet Foodborne Outbreak Online Database CDC Multistate outbreak of Salmonella infections associated with peanut butter and peanut butter--containing products --- United States, MMWR Morb Mortal Wkly Rep. 58: CDC Investigation update: multistate outbreak of human Salmonella Enteritidis infections associated with shell eggs CDC Investigation announcement: multistate outbreak of human Salmonella Newport infections linked to raw alfalfa sprouts CDC Investigation update: multistate outbreak of human Salmonella Montevideo infections CDC Preliminary FoodNet data on the incidence of infection with pathogens transmitted commonly through food States, MMWR Morb Mortal Wkly Rep. 58: Chalker, R. B., and M. J. Blaser A review of human salmonellosis: III. magnitude of Salmonella infection in the United States. Rev. Infect. Dis. 10: Chen, Y., and S. Knabel Strain typing, p In D. Liu (ed.), Handbook of Listeria monocytogenes.

45 Chen, Y., W. Zhang, and S. J. Knabel Multi-virulence-locus sequence typing identifies single nucleotide polymorphisms which differentiate epidemic clones and outbreak strains of Listeria monocytogenes. J. Clin. Microbiol. 45: Chittick, P., A. Sulka, R. V. Tauxe, and A. M. Fry A summary of national reports of foodborne outbreaks of Salmonella Heidelberg infections in the United States. J. Food Prot. 69: Christie, A. B Infectious diseases: epidemiologic and clinical practice. Churchill Livingstone, New York. 34. Clarkson, L. S., M. Tobin-D angelo, C. Shuler, S. Hanna, J. Benson, and A. Voetsch Sporadic Salmonella enterica serotype Javiana infections in Georgia and Tennessee: a hypothesis-generating study. Epidemiol. Infect. 138: Davis, M. A., K. N. K. Baker, D. R. Call, L. D. Warnick, Y. Soyer, M. Wiedmann, Y. Grohn, P. L. McDonough, D. D. Hancock, and T. E. Besser Multilocus variablenumber tandem-repeat method for typing Salmonella enterica serovar Newport. J. Clin. Microbiol. 47: De Reu, K., K. Grijspeerdt, W. Messens, M. Heyndrickx, M. Uyttendaele, J. Debevere, and L. Herman Eggshell factors influencing eggshell penetration and whole egg contamination by different bacteria, including Salmonella Enteritidis. Int. J. Food Microbiol. 112: Deveau, H., R. Barrangou, J. E. Garneau, J. Labonte, C. Fremaux, P. Boyaval, D. A. Romero, P. Horvath, and S. Moineau Phage response to CRISPR-encoded resistance in Streptococcus thermophilus. J. Bacteriol. 190: Doyle, M. P., and L. R. Beuchat eds Food microbiology: fundamentals and frontiers. ASM Press, Washington, DC.

46 Echeita, M. A., S. Herrera, and M. A. Usera Atypical, fljb-negative Salmonella enterica subsp. enterica strain of serovar 4,5,12:i:- appears to be a monophasic variant of serovar Typhimurium. J. Clin. Microbiol. 39: Efrain, M. R., M. A. Fair, R. Gautom, D. N. Cameron, S. B. Hunter, B. Swaminathan, and T. J. Barrett Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 3: Endo, T., K. Ikeo, and T. Gojobori Large-scale search for genes on which positive selection may operate. Mol. Biol. Evol. 13: Esaki, H., K. Noda, N. Otsuki, A. Kojima, T. Asai, Y. Tamura, and T. Takahashi Rapid detection of quinolone-resistant Salmonella by real time SNP genotyping. J. Microbiol. Methods. 58: Fakhr, M. K., L. K. Nolan, and C. M. Logue Multilocus sequence typing lacks the discriminatory ability of pulsed-field gel electrophoresis for typing Salmonella enterica serovar Typhimurium. J. Clin. Microbiol. 43: Fakhr, M. K., J. S. Sherwood, J. Thorsness, and C. M. Logue Molecular characterization and antibiotic resistance profiling of Salmonella isolated from retail turkey meat products. Foodborne Pathog. Dis. 3: Foley, S. L., S. Zhao, and R. D. Walker Comparison of molecular typing methods for the differentiation of Salmonella foodborne pathogens. Foodborne Pathog. Dis. 4: Foley, S. L., D. G. White, P. F. McDermott, R. D. Walker, B. Rhodes, P. J. Fedorka- Cray, S. Simjee, and S. Zhao Comparison of subtyping methods for differentiating Salmonella enterica serovar Typhimurium isolates obtained from food animal sources. J. Clin. Microbiol. 44: Franz, E., and A. H. C. van Bruggen Ecology of E. coli O157:H7 and Salmonella enterica in the primary vegetable production chain. Crit. Rev. Microbiol. 34:

47 Fry, N. K., B. Afshar, P. Visca, D. Jonas, J. Duncan, E. Nebuloso, A. Underwood, and T. G. Harrison Assessment of fluorescent amplified fragment length polymorphism analysis for epidemiological genotyping of Legionella pneumophila serogroup 1. Clin. Microbiol. Infect. 11: Gantois, I., R. Ducatelle, F. Pasmans, F. Haesebrouck, R. Gast, T. J. Humphrey, and F. Van Immerseel Mechanisms of egg contamination by Salmonella Enteritidis. FEMS Microbiology Reviews. 33: Gast, R. K., and C. W. Beard Production of Salmonella Enteritidis-contaminated eggs by experimentally infected hens. Avian Dis. 34: Gerner-Smidt, P., K. Hise, J. Kincaid, S. Hunter, S. Rolando, E. Hyytiä-Trees, E. M. Ribot, B. Swaminathan, and Pulsenet Taskforce Pulsenet USA: a five-year update. Foodborne Pathog. Dis. 3: Giammanco, G. M., C. Mammina, C. Romani, I. Luzzi, A. M. Dionisi, and A. Nastasi Evaluation of a modified single-enzyme amplified fragment length polymorphism (SE- AFLP) technique for subtyping Salmonella enterica serotype Enteritidis. Res. Microbiol. 158: Gomez, T. M., Y. Motarjemi, S. Miyagawa, F. K. Käferstein, and K. Stöhr Foodborne salmonellosis. World Health Stat Q. 50: Grimont, P. A. D., and F. X. Weill Antigenic formulae of the Salmonella serovars, 9th ed. WHO Collaborating Centre for Reference and Research on Salmonella. Institute Pasteur, Paris, France. 55. Groenen, P. M., A. E. Bunschoten, D. Soolingen, and J. D. Errtbden Nature of DNA polymorphism in the direct repeat cluster of Mycobacterium tuberculosis; application for strain differentiation by a novel typing method. Mol. Microbiol.10:

48 Guard, J Evolutionary trends in two strains of Salmonella enterica subsp. I serovar Enteritidis PT13a that vary in virulence potential Guard-Petter, J The chicken, the egg and Salmonella Enteritidis. Environ. Microbiol. 3: Haft, D. H., J. Selengut, E. F. Mongodin, and K. Nelson A Guild of 45 CRISPRassociated (Cas) protein families and multiple CRISPR/cas subtypes exist in prokaryotic genomes. PLoS Comput Biol. 1:e Hale, C. R., P. Zhao, S. Olson, M. O. Duff, B. R. Graveley, L. Wells, R. M. Terns, and M. P. Terns RNA-guided RNA cleavage by a CRISPR RNA-cas protein complex. Cell. 139: Hanning B. Irene, J.D. Nutt, Steven C. Ricke Salmonellosis outbreaks in the United States due to fresh produce: sources and potential intervention measures. Foodborne Pathog. Dis. 6: Harbottle, H., D. G. White, P. F. McDermott, R. D. Walker, and S. Zhao Comparison of multilocus sequence typing, pulsed-field gel electrophoresis, and antimicrobial susceptibility typing for characterization of Salmonella enterica serotype Newport isolates. J. Clin. Microbiol. 62. Harris, L. J., J. N. Farber, L. R. Beuchat, M. E. Parish, T. V. Suslow, E. H. Garrett, and F. F. Busta Outbreaks associated with fresh produce: incidence, growth, and survival of pathogens in fresh and fresh-cut produce. Compr Rev Food Sci Food Saf. 2: Heyndrickx, M., F. Pasmans, R. Ducatelle, A. Decostere, and F. Haesebrouck Recent changes in Salmonella nomenclature: The need for clarification. Vet. J. 170: Hoe, N., K. Nakashima, D. Grigsby, X. Pan, S. J. Dou, S. Naidich, M. Garcia, E. Kahn, D. Bergmire-Sweat, and J. M. Musser Rapid molecular genetic subtyping of serotype M1 group A Streptococcus strains. Emerg Infect Dis. 5:

49 Horvath, P., and R. Barrangou CRISPR/cas, the immune system of bacteria and archaea. Science. 327: Humphries, A. D., S. M. Townsend, R. A. Kingsley, T. L. Nicholson, R. M. Tsolis, and A. J. Baumler Role of fimbriae as antigens and intestinal colonization factors of Salmonella serovars. FEMS Microbiology Letters. 201: Jay, J. M Modern Food Microbiology. Springer Science, New York. 68. kenny, D. E Salmonella spp. survey of captive rhinoceroses in U.S. zoological institutions and private ranches. J. Zoo Wildl. Med. 30: Lan, R., P. R. Reeves, and S. Octavia Population structure, origins and evolution of major Salmonella enterica clones. Infect. Genet. Evol. 9: Larsson, J. T., M. Torpdahl, R. F. Petersen, G. Sørensen, B. A. Lindstedt, and E. M. Nielsen Development of a new nomenclature for Salmonella Typhimurium multilocus variable number of tandem repeats analysis (MLVA). Euro Surveill. 14: Levy, D. D., B. Sharma, and T. A. Cebula Single-nucleotide polymorphism mutation spectra and resistance to quinolones in Salmonella enterica serovar Enteritidis with a mutator phenotype. Antimicrob. Agents Chemother. 48: Lindstedt, B., E. Heir, E. Gjernes, and G. Kapperud DNA fingerprinting of Salmonella enterica subsp. enterica serovar Typhimurium with emphasis on phage type DT104 based on variable number of tandem repeat loci. J. Clin. Microbiol. 41: Lindstedt, B., E. Heir, T. Vardund, and G. Kapperud Fluorescent amplifiedfragment length polymorphism genotyping of Salmonella enterica subsp. enterica serovars and comparison with pulsed-field gel electrophoresis typing. J. Clin. Microbiol. 38: Maiden, M. C. J Multilocus sequence typing of bacteria. Annu. Rev. Microbiol. 60:

50 Maiden, M. C. J., J. A. Bygraves, E. Feil, G. Morelli, J. E. Russell, R. Urwin, Q. Zhang, J. Zhou, K. Zurth, D. A. Caugant, I. M. Feavers, M. Achtman, and B. G. Spratt Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc. Natl. Acad. Sci. U.S.A. 95: Majowicz, S., J. Musto, E. Scallan, F. Angulo, M. Kirk, S. O Brien, T. Jones, A. Fazil, and R. Hoekstra Food safety: The global burden of nontyphoidal Salmonella gastroenteritis. Clin. Infect. Dis.50: Marraffini, L. A., and E. J. Sontheimer. Self versus non-self discrimination during CRISPR RNA-directed immunity. Nature. 463: Mills, D. M., V. Bajaj, and C. A. Lee A 40 kb chromosomal fragment encoding Salmonella typhimurium invasion genes is absent from the corresponding region of the Escherichia coli K-12 chromosome. Mol. Microbiol. 15:749-v. 79. Miyamoto, T., E. Baba, T. Tanaka, K. Sasai, T. Fukata, and A. Arakawa Salmonella Enteritidis contamination of eggs from hens inoculated by vaginal, cloacal, and intravenous routes. Avian Dis. 41: MLST Mokrousov, I., E. Limeschenko, A. Vyazovaya, and O. Narvskaya Corynebacterium diphtheriae spoligotyping based on combined use of two CRISPR loci. Biotechnol J. 2: Mortimer, C., T. Peters, S. Gharbia, J. Logan, and C. Arnold Towards the development of a DNA-sequence based approach to serotyping of Salmonella enterica. BMC Microbiol. 4: Much, P., J. Pichler, S. Kasper, H. Lassnig, C. Kornschober, A. Buchner, C. König, and F. Allerberger A foodborne outbreak of Salmonella Enteritidis phage type 6 in Austria, Wien. Klin. Wochenschr. 121:

51 Mueller, U. G., and L. L. Wolfenbarger AFLP genotyping and fingerprinting. Trends Ecol Evol. 14: Munnoch, S., K. Ward, S. Sheridan, G. Fitzsimmons, C. Shadbolt, J. Piispanen, Q. Wang, T. Ward, T. Worgan, C. Oxenford, J. Musto, J. McAnulty, and D. Durrheim A multi-state outbreak of Salmonella Saintpaul in Australia associated with cantaloupe consumption. Epidemiol Infect. 137: Ochman, H., F. C. Soncini, F. Solomon, and E. A. Groisman Identification of a pathogenicity island required for Salmonella survival in host cells. Proc. Natl. Acad. Sci. U.S.A. 93: Okamura, M., Y. Kamijima, T. Miyamoto, H. Tani, K. Sasai, and E. Baba Differences among six Salmonella serovars in abilities to colonize reproductive organs and to contaminate eggs in laying hens. Avian Dis. 45: Pasmans, F., A. Martel, F. Boyen, D. Vandekerchove, I. Wybo, F. V. Immerseel, M. Heyndrickx, J. M. Collard, R. Ducatelle, and F. Haesebrouck Characterization of Salmonella isolates from captive lizards. Vet. Microbiol. 110: Patrick, M. E., P. M. Adcock, T. M. Gomez, S. F. Altekruse, B. H. Holland, R. V. Tauxe, and D. L. Swerdlow Salmonella Enteritidis infections, United States, Emerging Infect. Dis. 10: Pourcel, C., G. Salvignol, and G. Vergnaud CRISPR elements in Yersinia pestis acquire new repeats by preferential uptake of bacteriophage DNA, and provide additional tools for evolutionary studies. Microbiology. 151: Roos, A., P. Dieltjes, R. H. A. M. Vossen, M. R. Daha, and P. de Knijff Detection of three single nucleotide polymorphisms in the gene encoding mannose-binding lectin in a single pyrosequencing reaction. J. Immunol. Methods. 309:

52 Ross, I. L., and M. W. Heuzenroeder Use of AFLP and PFGE to discriminate between Salmonella enterica serovar Typhimurium DT126 isolates from separate foodrelated outbreaks in Australia. Epidemiol. Infect. 133: Ross, I. L., and M. W. Heuzenroeder A comparison of two PCR-based typing methods with pulsed-field gel electrophoresis in Salmonella enterica serovar Enteritidis. Int. J. Med. Microbiol. 299: Rychlik, I., D. Gregorova, and H. Hradecka Distribution and function of plasmids in Salmonella enterica. Vet. Microbiol. 112: Rytkönen, A., J. Poh, J. Garmendia, C. Boyle, A. Thompson, M. Liu, P. Freemont, J. C. D. Hinton, and D. W. Holden SseL, a Salmonella deubiquitinase required for macrophage killing and virulence. Proc. Natl. Acad. Sci. U.S.A. 104: Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. L. Willems, K. E. Dingle, F. M. Colles, and J. D. A. Van Embden Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41: Shea, J. E., M. Hensel, C. Gleeson, and D. W. Holden Identification of a virulence locus encoding a second type III secretion system in Salmonella Typhimurium. Proc. Natl. Acad. Sci. U.S.A. 93: Sivapalasingam, S., C. R. Friedman, L. Cohen, and Robert V. T Fresh produce: a growing cause of outbreaks of foodborne illness in the United States, 1973 through J. Food Prot. 67: Sokurenko, E. V., H. S. Courtney, D. E. Ohman, P. Klemm, and D. L. Hasty FimH family of type 1 fimbrial adhesins: functional heterogeneity due to minor sequence variations among fimh genes. J. Bacteriol. 176:

53 Sorek, R., V. Kunin, and P. Hugenholtz CRISPR a widespread system that provides acquired resistance against phages in bacteria and archaea. Nat. Rev. Microbiol. 6: Struelens, M. J Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin. Microbiol. Infect. 2: Swartz, M Human diseases caused by foodborne pathogens of animal origin. Clin. Infect. Dis. 34:S111-S Tamada, Y., Y. Nakaoka, K. Nishimori, A. Doi, T. Kumaki, N. Uemura, K. Tanaka, S. Makino, T. Sameshima, M. Akiba, M. Nakazawa, and I. Uchida Molecular typing and epidemiological study of Salmonella enterica serotype Typhimurium isolates from cattle by fluorescent amplified-fragment length polymorphism fingerprinting and pulsed-field gel electrophoresis. J. Clin. Microbiol. 39: Tankouo-Sandjong, B., A. Sessitsch, E. Liebana, C. Kornschober, F. Allerberger, H. Hächler, and L. Bodrossy MLST-v, multilocus sequence typing based on virulence genes, for molecular typing of Salmonella enterica subsp. enterica serovars. J. Microbiol. Methods. 69: Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J Clin Microbiol. 33: Threlfall, E. J Epidemic Salmonella Typhimurium DT 104--a truly international multiresistant clone. J. Antimicrob. Chemother. 46: Tindall, B. J., P. A. D. Grimont, G. M. Garrity, and J. P. Euzeby Nomenclature and taxonomy of the genus Salmonella. Int. J. Syst. Evol. Microbiol. 55:

54 Torpdahl, M., G. Sørensen, B. Lindstedt, and E. M. Nielsen Tandem repeat analysis for surveillance of human Salmonella Typhimurium infections. Emerg Infect Dis. 13: Torpdahl, M., M. N. Skov, D. Sandvang, and D. L. Baggesen Genotypic characterization of Salmonella by multilocus sequence typing, pulsed-field gel electrophoresis and amplified fragment length polymorphism. J. Microbiol. Methods. 63: Torres-Cruz, J., and M. W. van der Woude Slipped-strand mispairing can function as a phase variation mechanism in Escherichia coli. J. Bacteriol. 185: Tyagi, S., D. P. Bratu, and F. R. Kramer Multicolor molecular beacons for allele discrimination. Nat. Biotechnol. 16: USDA-ERS Foodborne illness cost calculator: Salmonella USDA-FSIS Serotypes profile of Salmonella isolates from meat and poultry products January 1998 through December USDA-FSIS Progress report on Salmonella testing of raw meat and poultry products, Vale, P. F., and T. J. Little CRISPR-mediated phage resistance and the ghost of coevolution past. Proc. R. Soc. Lond., B, Biol. Sci. 277: Voetsch, A., T. Van Gilder, F. Angulo, M. Farley, S. Shallow, R. Marcus, P. Cieslak, V. Deneen, and R. Tauxe FoodNet estimate of the burden of illness caused by nontyphoidal Salmonella infections in the United States. Clin. Infect. Dis. 38:S127-S Wallis, T. S., and E. E. Galyov Molecular basis of Salmonella-induced Enteritis. Mol. Microbiol. 36:

55 Zhang, W., W. Qi, T. J. Albert, A. S. Motiwala, D. Alland, E. K. Hyytia-Trees, E. M. Ribot, P. I. Fields, T. S. Whittam, and B. Swaminathan Probing genomic diversity and evolution of Escherichia coli O157 by single nucleotide polymorphisms. Genome Res. 16: Zhang, W., B. M. Jayarao, and S. J. Knabel Multi-virulence-locus sequence typing of Listeria monocytogenes. Appl. Environ. Microbiol. 70: Zheng, J., C. E. Keys, S. Zhao, J. Meng, and E. W. Brown Enhanced subtyping scheme for Salmonella Enteritidis. Emerging Infect. Dis. 13:

56 46 Chapter 3 Novel virulence gene and CRISPR multilocus sequence typing scheme for subtyping the major serovars of Salmonella enterica subspecies enterica Fenyun Liu 1, Rodolphe Barrangou 2, Peter Gerner-Smidt 3, Efrain Ribot 3, Stephen Knabel 1, and Edward Dudley 1* 1 Department of Food Science, the Pennsylvania State University, University Park, Pennsylvania 16802; 2 Danisco USA Incorporation, 3329 Agriculture Drive, Madison, Wisconsin Centers for Disease Control and Prevention, Atlanta, Georgia * Corresponding author. Mailing address: 326 Food Science Building, The Pennsylvania State University, University Park, PA 16802, US. Phone: egd100@psu.edu

57 Abstract Salmonella enterica subsp. enterica is the leading cause of bacterial foodborne disease in the United States. Molecular subtyping methods are powerful tools for tracking the farm-to-fork spread of foodborne pathogens during outbreaks. In order to develop a novel multilocus sequence typing (MLST) scheme for subtyping the major serovars of S. enterica subspecies enterica, the virulence genes ssel and fimh and Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR) regions were sequenced from 171 clinical isolates from serovars Typhimurium, Enteritidis, Newport, Heidelberg, Javiana, I 4, [5], 12; i: -, Montevideo, Muenchen and Saintpaul. The MLST scheme using only virulence genes identified epidemic clones, but was insufficient to separate outbreak clones. However, the addition of CRISPR sequences dramatically improved discriminatory power of this MLST method by accurately differentiating individual outbreak clones. Moreover, the present MLST scheme provided better discrimination of S. Enteritidis strains than PFGE. Cluster analyses also revealed the current MLST scheme is highly congruent with serotyping. In conclusion, the novel MLST scheme described in the present study accurately differentiated outbreak clones of the major serovars of Salmonella, and therefore maybe an excellent method for subtyping this important foodborne pathogen during outbreak investigations.

58 Introduction Salmonella enterica subsp. enterica (Salmonella) is the leading cause of bacterial foodborne disease in the United States, with approximately 1.4 million human cases each year since 1996, resulting in an estimated 17,000 hospitalizations, more than 500 deaths (10, 53) and a cost of 2.6 billion dollars (51). The nine most common serovars, S. Typhimurium, S. Enteritidis, S. Newport, S. Heidelberg, S. Javiana, S. I 4, [5], 12; i: -, S. Montevideo, S. Muenchen and S. Saintpaul, were responsible for more than 60% of human illnesses based on the Centers for Disease Control and Prevention s (CDC s) annual summaries of 2005 and 2006 (Table 3.1) (4, 5). Salmonella has been isolated from a broad range of foods, including raw animal foods (poultry, eggs, pork, beef, mutton and seafood), produce (sprouts, lettuce, spinach, tomatoes and peppers) and various processed foods including Italian-style salami, peanut butter, veggie booty and dry cereal (7). Widespread distribution of these foods makes tracking the transmission of Salmonella difficult during outbreak investigations. Outbreak investigations can also be hindered if the food source is not listed on the investigation questionnaire (27). For example, for the 2008 Salmonella outbreak due to consumption of Jalapeño and Serrano peppers, initial questionnaires did not ask if peppers were recently eaten (6). In order to define the routes of transmission of Salmonella within the food system, molecular subtyping methods have been employed to distinguish outbreak clones from non-related clones (18). Serotyping is one of the most common molecular subtyping methods for Salmonella. Serotyping distinguishes Salmonella based on immunological classification of the H and O antigens (21) and is typically the first subtyping method utilized during an outbreak. However, serotyping alone cannot distinguish outbreak clones of Salmonella. Several nucleic acids-based molecular subtyping methods have been used to subtype Salmonella, including amplified fragment length polymorphism (AFLP) (20, 36, 39,45, 49),

59 49 multiple loci variable number tandem repeat analysis (MLVA) (2, 34, 35, 40), and pulsed-field gel electrophoresis (PFGE) (13). PFGE is currently considered the gold standard method for subtyping foodborne pathogens and is the subtyping method used by PulseNet, the molecular surveillance network in the U.S. and throughout the world to investigate foodborne illnesses and outbreaks (19). The main advantage of PFGE is its high discriminatory power (i.e. ability to separate unrelated strains) for subtyping foodborne pathogens, including many of the major serovars of Salmonella (31). However, PFGE lacks discriminatory power for highly clonal serovars of Salmonella, such as S. Enteritidis (19, 54) and S. Montevideo (9), or highly clonal phage types like S. Typhimurium DT104 (19). The multistate S. Enteritidis outbreak associated with shell eggs in 2010 was caused by the most common PFGE-XbaI pattern (JEGX ) for S. Enteritidis in the PulseNet database (8). A similar scenario was also observed recently during the 2010 Italian-style salami outbreak, when the outbreak clone of S. Montevideo had the most common PFGE pattern in the PulseNet database (9). Besides inadequate discriminatory power, PFGE sometimes produces ambiguous data that are hard to compare and interpret between different laboratories. To enhance comparability and interpretation, a standardized PFGE protocol and an extensive quality assurance system were established by CDC (13, 19). Compared to PFGE, multilocus sequence typing (MLST), which targets nucleotide sequence differences of several DNA loci, has the potential to be a less labor-intensive method. Moreover, DNA sequence data are discreet, unambiguous, highly informative, portable and reproducible. Although MLST is an attractive subtyping approach, no satisfactory MLST scheme has yet been developed for subtyping Salmonella during outbreak investigations. MLST schemes targeting housekeeping genes have been developed; however, these schemes usually have much lower discriminatory power than that of PFGE (16, 28, 33, 49). This suggests that housekeeping genes do not provide sufficient resolution for investigating Salmonella outbreaks. In order to increase discriminatory power, virulence genes have been included in MLST schemes for subtyping Salmonella (17). Unlike housekeeping genes, virulence genes are

60 50 commonly under positive, diversifying selection (15). As a result, DNA sequences of virulence genes tend to be more variable than housekeeping genes, and thus able to provide increased discrimination (11, 17). Virulence genes have also been shown to provide high epidemiologic concordance (i.e. able to group related strains together). For example, six virulence genes were targeted in an MVLST (multi-virulence-locus sequence typing) scheme for subtyping Listeria monocytogenes, which showed very high discriminatory power (0.99) and perfect epidemiologic concordance (1.0) (11). Tankouo-Sandjong et al. (47) developed an MLST scheme based on both virulence genes and housekeeping genes to identify serovars of Salmonella. This scheme targeted the virulence genes flic and fljb along with two housekeeping genes, gyrb and atpd (47). However, the use of this MLST scheme to further characterize strains below serovar level was not tested. In another MLST study, virulence genes and housekeeping genes showed high discriminatory power (0.98) for subtyping S. Typhimurium, which was slightly higher than that of PFGE (0.96) (17). In that MLST scheme three virulence genes, hila, pefb and fimh, were included together with the 16S rrna gene and three housekeeping genes. Although this MLST scheme appeared to have adequate discriminatory power for subtyping S. Typhimurium, its capacity to discriminate strains from more clonal serovars such as S. Enteritidis was not tested. Comparative genomic analysis (25) suggested that virulence genes alone are not discriminatory enough for differentiating outbreak clones of S. Enteritidis. Therefore, additional genome targets with greater sequence diversity than virulence genes are needed in order to create an effective MLST scheme for Salmonella. One of the fastest evolving genetic elements in bacteria genomes is CRISPRs (Clustered Regularly Interspaced Palindromic Repeats) (43). CRISPRs have been identified within the genomes of many archaeal and bacterial species, including Salmonella (30, 43, 50). CRISPRs encode tandem sequences containing bp direct repeats (DRs) separated by spacers of similar size (Fig. 3.1). Spacers are derived from foreign nucleic acids such as phage or plasmids and can protect bacteria from subsequent infection by homologous phage and plasmids (1).

61 51 Many CRISPR loci are flanked at the 3 end by an AT-rich leader sequence and CRISPRassociated (Cas) genes (Fig. 3.1) (1, 3, 26). As a bacterial immune system against foreign DNA, CRISPRs must evolve rapidly to adapt to different phage pools (52). Besides addition of new spacers, deletion of spacers is also frequently observed (12, 38). Because of the high polymorphism of CRISPRs, they have been successfully used to subtype M. tuberculosis during outbreak investigations (24). CRISPR sequence analysis has also characterized a number of other bacteria, including Yersinia pestis (38), serotype M1 group A Streptococcus strains (29), and Campylobacter jejuni (42). Two CRISPR loci are found in all Salmonella serovars in the CRISPR database ( (22, 50). Generally, the two CRISPR loci have different number of repeats/spacers and different set of spacers. There have been no reports of CRISPRs being used as markers in an MLST scheme for subtyping Salmonella. Therefore, the purpose of the present study was to investigate whether MLST based on both virulence genes and CRISPRs can accurately differentiate outbreak clones of the major serovars of Salmonella.

62 Materials and methods Bacterial isolates and DNA extraction. All 171 Salmonella isolates used in this study (Table 3.2) were from culture collections maintained by the Centers for Disease Control and Prevention (CDC) in Atlanta, GA, USA. This set of isolates represents the 9 serovars most commonly associated with human disease and includes isolates involved in multiple outbreaks, with 2 to 3 isolates per outbreak. In some cases, isolates obtained from the same outbreak which had different PFGE patterns (had poor epidemiologic concordance by PFGE) were deliberately included. All isolates were previously analyzed by serotyping and most isolates were analyzed by PFGE by CDC. Bacterial isolates were stored at -80 C in 20% glycerol. When needed, isolates were grown overnight in Tryptic Soy Broth (TSB) (Difco Laboratories, Becton Dickinson, Sparks, MD) at 37 C. For all isolates, DNA was extracted using the UltraClean Microbial DNA extraction kit (Mo Bio Laboratories, Solana Beach, CA) and stored at -20 C before use. Selection of virulence genes. Two virulence genes (fimh and ssel) and two Clustered Regularly Interspaced Short Palindromic Repeats regions (CRISPR1 and CRISPR2) were selected as markers for MLST. The lengths and functions of these MLST markers are listed in Table 3.3. Additionally, 12 other virulence genes (hila, fimh2, pipb, sope, ssef, ssej, siia, sifb, stda, fima, bcfc and phoq) (Table S1) were initially investigated, but were excluded from the MLST scheme due to inadequate sequence variation. PCR amplification. Primers were designed using Primer 3.0 ( and are listed in Tables 3.4 and S1. Primers for CRISPR1 were designed based upon consensus alignments of the published S. Typhimurium LT2 (accession number AE006468) and S. Newport str. SL254 genomes (accession number CP001113), and the S. Javiana str. GA_MM (accession number ABEH ) whole genome shotgun sequence (Table 3.4). Primers for the other three markers were designed based on the published S.

63 53 Typhimurium LT2 genome. PCR amplifications were performed using a Taq PCR Master Mix Kit (Qiagen Inc., Balencia, CA) and a Mastercycler PCR thermocycler (Eppendorf Scientific, Hamburg, Germany). A 25 µl PCR reaction system contained 12.5 µl Taq PCR 2 master mix, 9.5 µl PCR-grade water, 1.0 µl DNA template, 1.0 µl forward primer (final concentration, 0.4 µm) and 1.0 µl reverse primer (final concentration, 0.4 µm). A single PCR cycling condition was used for separately amplifying all four markers (initial denaturation at 94 C for 10 min; 28 cycles of 94 C for 1 min, 55 C for 1 min,72 C for 1 min; final extension at 72 C for 10 min). DNA sequencing. After PCR, products for sequencing were treated with 1/20 volume of shrimp alkaline phosphatase (1 U/µl, USB Corp. Cleveland, OH) and 1/20 volume of exonuclease I (10 U/µl, USB Corp). The mixture was then incubated at 37 C for 45 min to degrade remaining primers and unincorporated dntps. After that, the mixture was incubated at 80 C for 15 min to inactivate the added enzymes. PCR products were sent to the Genomics Core Facility at the Pennsylvania State University for sequencing using the ABI Data 3730XL DNA Analyzer. In order to obtain complete DNA sequences of fimh and ssel, two more primers targeting the internal regions of these two genes were used together with the forward and reverse primers (Table 3.4). Both DNA strands of the amplicons were sequenced. Sequence analysis and sequence type assignment. For fimh and ssel, sequences were aligned and single nucleotide polymorphisms (SNPs) were identified using MEGA 4.0 (46). For CRISPR1 and CRISPR2, analyses of the spacer arrangements were performed using CRISPRcompar (23) and spacers were visualized as described by Deveau et al. (12). Different allelic types (ATs) (sequences with at least one-nucleotide difference or one-spacer difference in the case of CRISPRs) were assigned arbitrary numbers. The combination of 4 alleles (fimh, ssel, CRISPR1 and CRISPR2) determined its allelic profile and each unique allelic profile was designated as a unique sequence type (ST).

64 54 Calculation of epidemiologic concordance (E). Epidemiologic concordance (E) was calculated using the equation developed by the European Study Group on Epidemiologic Markers (44). Cluster analysis. Cluster analyses were performed based on allelic profile data and results were visualized using the tree drawing tool on PubMLST ( CRISPR1 and CRISPR2 were combined into one allele for a more accurate cluster analysis, because CRISPR1 and CRISPR2 might be spatially linked (50). Nucleotide sequence accession number. DNA sequences of the four genetic MLST markers were deposited in GenBank under accession numbers HQ to HQ

65 Results Results of MVLST. We began this study by sequencing 14 virulence genes (fimh, ssel, hila, fimh2, pipb, sope, ssef, ssej, siia, sifb, stda, fima, bcfc and phoq) from 20 S. Typhimurium, 15 S. Newport, and 15 S. Enteritidis isolates. Two virulence genes, fimh and ssel, were found to provide discrimination equal to the combined discrimination of all 14 virulence genes (data not shown), therefore, the other 12 virulence genes were excluded from the rest of the study. fimh and ssel were sequenced from the remaining isolates, and the total number of allelic types was about the same for fimh (17 allelic types) and ssel (16 allelic types) (Table 3.5). The total number of polymorphic sites and percentage of polymorphic sites for fimh was 48 and 4.76% and for ssel it was 69 and 7.23%, respectively (Table 3.6). For sequence variations of fimh within each serovar, the percentage of polymorphic sites ranged from 0% to 1.79%. For ssel, the percentage of polymorphic sites ranged from 0% to 3.88%. For both fimh and ssel, less polymorphism was observed for serovars Typhimurium, Enteritidis, Heidelberg, Javiana and I 4, [5], 12: i :-, compared to serovars Newport, Montevideo, Muenchen and Saintpaul (Table 3.6). Sequences of ssel were especially conserved in serovars Typhimurium, Heidelberg, Javiana and I 4, [5], 12: i :-, with no SNPs observed within each serovar. For all serovars, a total of 39 polymorphic sites in ssel were nonsynonymous, and 13 polymorphic sites in fimh were nonsynonymous (Table 3.6). Addition of CRISPR1 and CRISPR2 in the MLST scheme. Since the discrimination provided by virulence genes was limited (separation to outbreak level was not achieved), addition of CRISPR1 and CRISPR2 into the MLST scheme was investigated. The number of allelic types for CRISPR1 (49 allelic types) and CRISPR2 (53 allelic types) were significantly greater than those for virulence genes (Table 3.5). In total, there were 69 sequence types based on both virulence genes and CRISPRs for all 171 isolates (Table 3.5). An equal number of allelic types

66 56 was observed in both CRISPR1 and CRISPR2 for serovars Javiana and Montevideo (Table 3.5). However, for serovars Typhimurium, Enteritidis, Newport, Heidelberg and Saintpaul, CRISPR2 contained more allelic types than CRISPR1. In contrast, for serovar Muenchen, CRISPR1 contained more allelic types than CRISPR2 (Table 3.5). Repeat sequences of the two CRISPRs were generally conserved as shown by the typical repeat in Table 3.7. However, SNPs were sometimes observed in the repeat sequences in both CRISPRs and we define these as repeat variants (Table 3.7). The repeat variant of CRISPR1 had one SNP at the first nucleotide, which is A instead of C (Table 3.7). Terminal repeat sequences which are located furthest from the leader sequence (Fig. 3.1) had more SNPs than the repeat variants sequences when compared to the typical repeat sequence (Table 3.7). The total numbers of unique spacers in CRISPR1 and CRISPR2 for all 171 isolates analyzed were 166 and 182, respectively (Table 3.8). The number of spacers in CRISPR1 ranged from 3 spacers to 24, while the number of spacers in CRISPR2 ranged from 2 to 25 (Table 3.8 and Fig. S1). CRISPR2 had more spacers than CRISPR1 for all serovars except S. Muenchen, in which CRISPR1 had more spacers than CRISPR2 (Table 3.8 and Fig. S1). The number of spacers also varied between different serovars. For example, the average number of spacers in CRISPR2 of S. Muenchen was 2.5, while the average number of spacers in CRISPR2 of S. Typhimurium was 19.6 (Table 3.8). Cluster analyses. Cluster diagrams based on allelic profiles were constructed using only the two virulence genes (Fig. 3.2a) and also using virulence genes combined with CRISPRs (Fig. 3.2b), respectively. Again, significantly greater separation was provided by the addition of CRISPR1 and CRISPR2 for all serotypes, compared to the separation provided by virulence genes alone. MLST results showed high congruence with serotypes of Salmonella. On both cluster trees, different serovars occupied distinct branches, except serovars Typhimurium and I 4, [5], 12: i :-, which were clustered together. Also, three singletons Mvo ST3, Mcn ST12 and S ST4 were observed on both cluster trees (Fig. 3.2a and ab).

67 57 Comparison of MLST with PFGE. Compared to PFGE, the addition of CRISPRs into the present MLST scheme provided greater discrimination of outbreak clones of S. Enteritidis (Table 3.2 and 3.5). Most isolates of S. Enteritidis (25 out of 34) had either XbaI and BlnI PFGE profile (JEGX , JEGA ) or (JEGX , JEGA ) (Table 3.2). Isolates SE1, SE2, SE23, SE18, SE17, SE20, SE32 and SE33 had the same PFGE profile (JEGX , JEGA ), but had two MLST sequence types (E ST1 and E ST 9) (Table 3.2). Also, the PFGE profile (JEGX , JEGA ), which included isolates SE6, SE7, SE8, SE9, SE15, SE16, SE19, SE30, SE12, SE13, SE14, SE26, SE31, SE28, SE29, SE24 and SE34, were further separated into five sequence types (E ST3, E ST4, E ST6, E ST7, E ST8) by MLST (Table 3.2). However, in the case of some serovars (S. Newport, S. Typhimurium, S. I 4, [5], 12: i :-, S. Montevideo) PFGE provided greater separation than MLST for strains associated with different outbreaks. For example, PFGE separated S. I 4, [5], 12: i :- isolates (ST1, ST2 and ST3) of the turkey potpie outbreak (cluster 0706PAJPX-1c) from isolates (ST14 and ST15) of cluster 0607INjpx-1c, while these isolates could not be distinguished by MLST (Table 3.2). This was also the case for S. Typhimurium isolates from the Noble Farm raw milk outbreak and outbreak cluster 0309ORJPX-1c (Table 3.2). Another example when PFGE was more discriminatory than MLST was the raw chicken outbreak (cluster 0807AZJIX-1c) and the salami/pepper outbreak (cluster 0908ORJIX-1) of S. Montevideo (Table 3.2). For S. Heidelberg, the most accurate outbreak identification was achieved by combining MLST and PFGE. MLST provided separation for the cruise ship outbreak (cluster 0607NYJF6-1c) and religious camp outbreak (cluster 0607PAJF6-1c), and for the hummus outbreak (cluster JF6X ) and outbreak cluster 0702TNJF6-1c, which could not be distinguished by PFGE (Table 3.2). However, PFGE separated the cruise ship outbreak from the hummus outbreak, which were indistinguishable by MLST (Table 3.2). For S. Saintpaul, both methods allowed accurate separation and identification of all outbreaks due to this serovar (Table 3.2).

68 58 Comparison of MLST results with epidemiologic data. Isolates with the same cluster code had identical MLST sequence types for serovars S. Typhimurium, S. Newport, S. I 4, [5], 12: i :-, S. Saintpaul and S. Montevideo (Table 3.2). MLST sequence types were the same among isolates with the same cluster code for S. Enteritidis, except clusters 0505GAJEC-1c and 0612MEJEC-1c, and S. Heidelberg, except for 0704AZJPX-1c. Isolates SE10 and SE11, which have the same cluster code (0505GAJEC-1c), had different sequence types and also different PFGE patterns (Table 3.2). Three isolates of S. Enteritidis (SE12, SE13 and SE14) and three isolates of S. Heidelberg (SH18, SH19 and SH20) which have the same cluster code (0612MEJEC-1c and 0704AZJPX-1c), respectively, also had different sequence types (Table 3.2). For S. Muenchen, almost all isolates within each of the six cluster codes had different sequence types (Table 3.2). We could not perform a similar analysis with S. Javiana, because it did not contain any isolates with a cluster code identified by PulseNet. Epidemiologic concordance of MLST. Values of epidemiologic concordance of MLST and PFGE for each serovar are listed in Table 3.9, except for the serovar Javiana which didn t contain any isolates with a cluster code identified by PulseNet which prevented calculation of an epidemiologic concordance value for this serovar. Epidemiologic concordance values were calculated based on isolates from well-defined outbreaks (isolates with cluster codes), so sporadic isolates and isolates without cluster codes were excluded from epidemiologic concordance calculations. Values of epidemiologic concordance were biased against PFGE, because isolates from outbreaks with variations in PFGE patterns were deliberately included in this study, which reduced the epidemiologic concordance of PFGE. For instance, isolates ST6, ST7 and ST8 were all from the 2008 peanut butter outbreak, but each of them had a distinct PFGE pattern (Table 3.2). Generally speaking, MLST showed high epidemiologic concordance for subtyping all serovars included in this study, except for S. Muenchen (epidemiologic concordance= 0.39) (Table 3.9). MLST showed higher epidemiologic concordance than PFGE for serovars Enteritidis, Typhimurium, Newport and Montevideo, equal epidemiologic concordance for

69 59 serovar Saintpaul, but lower epidemiologic concordance for serovars Heidelberg and Muenchen (Table 3.9).

70 60 Table 3.1. Top nine most frequently reported serovars from human sources in 2005 which were analyzed in the present study Rank Serovar No. of laboratory-confirmed cases 1 % of total cases 1 Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12: i : Montevideo Muenchen Saintpaul total Laboratory-confirmed cases include both outbreak cases and sporadic cases. Data reproduced from CDC s Salmonella annual review (

71 61 Table 3.2. Outbreak information, PFGE profile and MLST results for the 171 isolates analyzed in the present study CDC MLST Code 1 Source State Food vehicle Cluster PFGE XbaI PFGE BlnI ST 2 ST29 Water filter UT Frog 0909MAJPX-1 JPXX JPXA T ST1 ST30 Human /Stool MD Frog 0909MAJPX-1 JPXX JPXA T ST1 ST31 Human /Stool OH Frog 0909MAJPX-1 JPXX JPXA T ST1 ST4 Human /Stool CO Water 0803COJPX-1c JPXX JPXA T ST2 ST5 Water CO Water 0803COJPX-1c JPXX JPXA T ST2 ST6 Human /Stool OH peanut butter 0811MLJPX-1c JPXX JPXA T ST3 ST7 Human /Stool OH peanut butter 0811MLJPX-1c JPXX JPXA T ST3 ST8 Food/peanut butter MN peanut butter 0811SDCJPX-1c JPXX JPXA T ST3 ST9 Stool MA Raw milk Noble Farm outbreak JPXX JPXA T ST4 ST10 Raw milk MA Raw milk Noble Farm outbreak JPXX JPXA T ST4 ST17 NA OR NA 0309ORJPX-1c JPXX JPXA T ST4 ST18 NA OR NA 0309ORJPX-1c JPXX JPXA T ST4 ST11 Stool NM NA Santa Fe JPXX JPXA T ST5 ST12 Stool NM NA Santa Fe JPXX JPXA T ST5 ST13 Stool NM NA Santa Fe JPXX JPXA T ST5 ST26 Human /Stool OR Snake / mouse 0908ORJPX-1 JPXX JPXA T ST5 ST27 Human /Stool OR Snake / mouse 0908ORJPX-1 JPXX JPXA T ST5 ST28 Animal OR Snake / mouse 0908ORJPX-1 JPXX JPXA T ST5 ST39 Human /Stool VA Sporadic Sporadic JPXX JPXA T ST5 ST16 Stool MA Veggie booty 0704WIWWS-c JPXX JPXA T ST6 ST19 Stool VT Veggie booty 0704WIWWS-1c JPXX JPXA T ST6 ST20 Stool VT Veggie booty 0704WIWWS-1c JPXX JPXA T ST6 ST32 Human /Stool AR Day care 0602ARJPX-2c JPXX JPXA T ST7 ST33 Human /Stool AR Day care 0602ARJPX-2c JPXX JPXA T ST7 ST34 Human /Stool AR Day care 0602ARJPX-2c JPXX JPXA T ST7 ST40 Human /Stool NY Sporadic Sporadic JPXX JPXA T ST8 SE1 Human/stool MN Stuffed chicken 0603MNJEG-1c JEGX JEGA E ST1 SE2 Human/stool MN Stuffed chicken 0603MNJEG-1c JEGX JEGA E ST1 SE23 Human/stool MN NA 0603MNJEG-1c JEGX JEGA E ST1 SE18 Human/Stool MN NA 0803MNJEG-1 JEGX JEGA E ST1 SE3 Environment CA Almonds Almonds 2001 JEGX NA E ST2 SE4 Food/raw almonds CA Almonds Almonds 2001 JEGX NA E ST2 SE5 Environment CA Almonds Almonds 2001 JEGX NA E ST2 SE21 Environment NA NA Almonds 2001 JEGX NA E ST2 SE25 Environment NA Prison Almonds 2001 JEGX NA E ST2 SE6 Human/stool ME NA 0612MEJEG-1c JEGX JEGA E ST3 SE7 Human/stool ME NA 0612MEJEG-1c JEGX JEGA E ST3 SE26 Human/stool CO NA NA JEGX JEGA E ST3 SE31 Human/stool CO NA NA JEGX JEGA E ST3 SE24 Human/Stool WV NA NA JEGX JEGA E ST3 SE8 Human/Stool PA Egg 0801PAJEG-1 JEGX JEGA E ST4 SE9 Human/Stool PA Egg 0801PAJEG-1 JEGX JEGA E ST4 SE15 Human/Stool PA NA 0801PAJEG-1 JEGX JEGA E ST4 SE34 Human/Stool CT NA NA JEGX JEGA E ST4 SE11 Human/stool GA Hospital eggs 0505GAJEG-1c JEGX JEGA E ST4 SE10 Human/stool GA Hospital eggs 0505GAJEG-1c JEGX JEGA E ST5 SE12 NA ME NA 0612MEJEG-1c JEGX JEGA E ST6 SE13 Human/stool ME NA 0612MEJEG-1c JEGX JEGA E ST6 SE14 Human/stool ME NA 0612MEJEG-1c JEGX JEGA E ST7 SE16 Human/stool GA NA 0506GAJEG-1c JEGX JEGA E ST8 SE19 Human/stool GA NA 0506GAJEG-1c JEGX JEGA E ST8 SE30 Human/stool GA Prison 0506GAJEG-1c JEGX JEGA E ST8 SE22 Human/stool OR NA 0509ORJEG-1c JEGX JEGA E ST8 SE27 Human/stool OR NA 0509ORJEG-1c JEGX JEGA E ST8 SE28 Human SC NA 0504SCJEG-1c JEGX JEGA E ST8 SE29 Human/stool ID NA 0504CAOCJEG-1c JEGX JEGA E ST8

72 SE17 NA OH Frozen chicken Outbreak JEGX JEGA E ST9 SE20 NA OH NA Outbreak JEGX JEGA E ST9 SE32 Human/Stool MI NA 0708MIJEG-1c JEGX JEGA E ST9 SE33 Human/Stool MI NA 0708MIJEG-1c JEGX JEGA E ST9 SN1 NA IL NA NA JJPX NA N ST1 SN2 NA IL NA NA JJPX NA N ST1 SN3 NA NA NA 0509NHJJP-1c. JJPX JJPA N ST2 SN4 NA NA NA 0509NHJJP-1c. JJPX JJPA N ST2 SN5 NA NA NA NA JJPX NA N ST3 SN6 NA NA NA NA JJPX NA N ST3 SN7 Human/Stool CA NA 0710CAJJP-1c JJPX JJPA N ST4 SN8 Human/Stool CA NA 0710CAJJP-1c JJPX JJPA N ST4 SN11 Human/Stool SD NA 0712SDJJP-1c JJPX JJPA N ST4 SN12 Human/Stool SD NA 0712SDJJP-1c JJPX JJPA N ST4 SN9 Human/Stool AZ NA 0802AZJJP-1c JJPX JJPA N ST5 SN10 Human/Stool AZ NA 0802AZJJP-1c JJPX JJPA N ST5 SN13 Human/Stool GA NA 0711GAJJP-1c JJPX JJPA N ST6 SN14 Human/Stool GA NA 0711GAJJP-1c JJPX JJPA N ST6 SN15 Human/Stool GA NA 0711GAJJP-1c JJPX JJPA N ST6 SH1 Human DE cruise ship 0607NYJF6-1c JF6X NA H ST1 SH2 Human NY cruise ship 0607NYJF6-1c JF6X NA H ST1 SH3 Human NY cruise ship 0607NYJF6-1c JF6X NA H ST1 SH8 Human IL hummus 0707ILJF6-1c JF6X JF6A H ST1 SH9 Human IL hummus 0707ILJF6-1c JF6X JF6A H ST1 SH10 Human IL hummus 0707ILJF6-1c JF6X JF6A H ST1 SH11 Human IL hummus 0707ILJF6-1c JF6X JF6A H ST1 SH16 NA NA Sporadic Sporadic JF6X NA H ST1 SH17 NA NA Sporadic Sporadic JF6X NA H ST1 SH18 Human NA NA 0704AZJPX-1c JF6X NA H ST1 SH4 Human PA a religious camp 0607PAJF6-1c JF6X NA H ST2 SH5 Human PA a religious camp 0607PAJF6-1c JF6X NA H ST2 SH6 Human PA a religious camp 0607PAJF6-1c JF6X NA H ST2 SH7 Human PA a religious camp 0607PAJF6-1c JF6X NA H ST2 SH15 NA NA Sporadic Sporadic JF6X NA H ST2 SH12 Human TN NA 0702TNJF6-1c JF6X JF6A H ST3 SH13 Human TN NA 0702TNJF6-1c JF6X JF6A H ST3 SH14 NA NA Sporadic Sporadic JF6X NA H ST4 SH19 Human NA NA 0704AZJPX-1c JF6X NA H ST5 SH20 Human NA NA 0704AZJPX-1c JF6X NA H ST6 SJ1 NA AL NA NA JGGX NA J ST1 SJ5 NA AR NA NA JGGX NA J ST1 SJ13 NA LA NA NA NA NA J ST1 SJ15 NA outbreak NA JGGX JGGA J ST1 SJ2 NA TX NA NA JGGX NA J ST2 SJ3 NA LA NA NA NA NA J ST3 SJ8 NA LA NA NA NA NA J ST3 SJ4 NA TX NA NA NA NA J ST4 SJ6 NA AR NA NA JGGX NA J ST5 SJ9 NA AR NA NA JGGX NA J ST5 SJ7 NA TX NA NA JGGX NA J ST6 SJ10 NA HU NA NA NA NA J ST7 SJ11 NA MD NA NA JGGX NA J ST8 SJ12 NA IL NA NA JGGX NA J ST9 SJ14 NA NV NA NA NA NA J ST10 ST1 5 Stool CA Turkey potpie 0706PAJPX-1c JPXX JPXA I ST1 ST2 5 Stool GA Turkey potpie 0706PAJPX-1c JPXX JPXA I ST1 ST3 5 Food/Turkey potpie WI Turkey potpie 0706PAJPX-1c JPXX JPXA I ST1 ST14 5 Stool IN NA INjpx-1c JPXX JPXA I ST1 ST15 5 Stool IN NA 0607INjpx-1c JPXX JPXA I ST1 ST21 5 Human /Stool OH Snake 0806OHJPX-1c JPXX JPXA I ST2 ST22 5 Human /Stool OH Snake 0806OHJPX-1c JPXX JPXA I ST2 ST23 5 Human /Stool OH Snake 0806OHJPX-1c JPXX JPXA I ST2 62

73 63 ST24 5 Food/Egg wash ME Egg 0404PAJPX-1c JPXX JPXA I ST3 ST25 5 NA VT Egg 0404PAJPX-1c JPXX JPXA I ST3 ST35 5 Human /Stool OH Sporadic Sporadic JPXX JPXA I ST4 ST36 5 Human /Stool MA Sporadic Sporadic JPXX JPXA I ST4 ST37 5 Human /Stool MO Sporadic Sporadic JPXX JPXA I ST4 SMvo1 Blood TX NA NA NA NA Mvo ST1 SMvo2 Stool TX NA NA NA NA Mvo ST2 SMvo7 Human MD NA NA JIXX NA Mvo ST3 SMvo3 Human/Rectal swab AZ Raw chicken 0807AZJIX-1c JIXX NA Mvo ST3 SMvo8 Human/Stool AZ Raw chicken 0807AZJIX-1c JIXX JIXA Mvo ST3 SMvo9 Human/Swab AZ Raw chicken 0807AZJIX-1c JIXX JIXA Mvo ST3 SMvo10 Human/Stool AZ Raw chicken 0807AZJIX-1c JIXX JIXA Mvo ST3 SMvo11 Human/Stool UT Salami/pepper 0908ORJIX-1 JIXX JIXA Mvo ST3 SMvo12 Human/Urine OR Salami/pepper 0908ORJIX-1 JIXX JIXA Mvo ST3 SMvo13 Human/Stool AZ Salami/pepper 0908ORJIX-1 JIXX NA Mvo ST3 SMvo15 Human/Stool TN Salami/pepper 0908ORJIX-1 JIXX NA Mvo ST3 SMvo14 NA AZ NA NA NA NA Mvo ST3 SMvo4 NA TX NA NA JIXX NA Mvo ST4 SMvo5 Human/Stool TX NA NA JIXX NA Mvo ST5 SMvo6 Human/Stool TN NA NA JIXX NA Mvo ST6 SMcn1 NA TX NA outbreak JJPX NA Mcn ST1 SMcn2 NA NYC NA outbreak JJPX NA Mcn ST2 SMcn3 Human/Stool LA NA 0509NHJJP-1c. JJPX JJPA Mcn ST3 SMcn4 NA TX NA 0509NHJJP-1c. JJPX JJPA Mcn ST4 SMcn5 NA TX NA NA NA NA Mcn ST5 SMcn6 Human/Stool TX NA NA NA NA Mcn ST6 SMcn7 NA TX NA 0710CAJJP-1c JJPX JJPA Mcn ST7 SMcn8 NA TX NA 0710CAJJP-1c JJPX JJPA Mcn ST8 SMcn9 Human/Stool TX NA 0802AZJJP-1c JJPX JJPA Mcn ST9 SMcn10 NA TX NA 0802AZJJP-1c JJPX JJPA Mcn ST10 SMcn11 Human/Stool TX NA 0712SDJJP-1c JJPX JJPA Mcn ST11 SMcn12 Human MD NA 0712SDJJP-1c JJPX JJPA Mcn ST12 SMcn13 NA OR Orange Juice 0711GAJJP-1c JJPX JJPA Mcn ST13 SMcn15 NA WA Orange Juice 0711GAJJP-1c JJPX JJPA Mcn ST13 SMcn14 NA WA Orange Juice 0711GAJJP-1c JJPX JJPA Mcn ST14 SS10 Human MA NA 0806MAJN6-1c JN6X JN6A S ST1 SS11 Human MA NA 0806MAJN6-1c JN6X JN6A S ST1 SS12 Human MA NA 0806MAJN6-1c JN6X JN6A S ST1 SS6 Human NE sprouts 0902NEJN6-1 JN6X NA S ST2 SS7 Human NE sprouts 0902NEJN6-1 JN6X NA S ST2 SS8 Human NE sprouts 0902NEJN6-1 JN6X NA S ST2 SS9 Human NE sprouts 0902NEJN6-1 JN6X NA S ST2 SS1 NA MN jalapeños 0805NMJN6-1c JN6X JN6A S ST3 SS2 Human TX jalapeños 0805NMJN6-1c JN6X JN6A S ST3 SS3 Human NM jalapeños 0805NMJN6-1c JN6X JN6A S ST3 SS4 Human AZ jalapeños 0805NMJN6-1c JN6X JN6A S ST3 SS18 NA NE Sporadic Sporadic JN6X NA S ST3 SS19 NA TX Sporadic Sporadic JN6X JN6A S ST3 SS16 NA CA Sporadic Sporadic JN6A NA S ST4 SS13 NA CA NA 0807LACJN6-1c JN6X JN6A S ST5 SS14 NA CA NA 0807LACJN6-1c JN6X JN6A S ST5 SS15 NA MD Sporadic Sporadic JN6X NA S ST5 SS20 NA NV Sporadic Sporadic JN6X JN6A S ST6 1 ST: S. Typhimurium (ST are isolates of S. Typhimurium var Copenhagen). SE: S. Enteritidis. SN: S. Newport. SH: S. Heidelberg. SJ: S. Javiana. SI: S. I 4, [5], 12; i: -. SMvo: S. Montevideo. SMcn: S. Muenchen. SS: S. Saintpaul.

74 64 2 ST: sequence type. T: S. Typhimurium. E: S. Enteritidis. N: S. Newport. H: S. Heidelberg. J: S. Javiana. Mvo: S. Montevideo. Mcn: S. Muenchen. S: S. Saintpaul. For instance, T ST1 stands for Typhimurium sequence type 1. 4 NA: Not available. 5 ST1-3, 14-15,21-25 and are isolates of S. I 4, [5], 12; i: -.

75 65 Table 3.3. Size, function and nucleotide location of the four markers targeted in the present study Marker Size (bps) Function Nucleotide location in S. Typhimurium LT2 fimh 1008 Host-cell-specific recognition 28,425-29,432 ssel 954 Inflammation and macrophage killing 2,394,795-2,395,748 CRISPR Defense against phage 3,076,611-3,077,006 CRISPR Defense against phage 3,094,279-3,096,260 1 Length of CRISPRs varied because the number of repeats/spacers changed among the different strains analyzed.

76 66 Table 3.4. Primers used to amplify and sequence the four MLST markers Marker Primer sequence (5'-3') Note fimh CGTCGTCATAAAAGGAAAAA Forward primer for both amplification and sequencing GAACAAAACACAACCAATAGC Reverse primer for both amplification and sequencing CTCGCCAGACAATGTTTACT Reverse primer for sequencing internal region CATTCACTTCGCAGTTTTG Forward primer for sequencing internal region ssel AGGAAACAGAGCAAAATGAA Forward primer for both amplification and sequencing TAAATTCTTCGCAGAGCATC Reverse primer for both amplification and sequencing GGAGTTGAAAATCTTTGGTG Reverse primer for sequencing internal region TTTACCGAGAGAAAAGGTGA Forward primer for sequencing internal region CRISPR1 GATGTAGTGCGGATAATGCT Forward primer for both amplification and sequencing GGTTTCTTTTCTTCCTGTTG 1 Reverse primer for both amplification and sequencing GATGATATGGCAACAGGTTT 1 Reverse primer for both amplification and sequencing TATTGACTGCGATGAGATGA 2 Reverse primer for both amplification and sequencing CRISPR2 ACCAGCCATTACTGGTACAC Forward primer for both amplification and sequencing ATTGTTGCGATTATGTTGGT Reverse primer for both amplification and sequencing 1 The 2 reverse primers (reverse 1 and reverse 2) of CRISPR1 were added together with forward primer to amplify CRISPR1 in all serovars except S. Javiana. 2 Reverse primer for SJ (S. Javiana) was needed for amplification and sequencing of CRISPR1 in S. Javiana isolates.

77 67 Serovar Table 3.5. Number of isolates, allelic types and sequence types in each serovar No. of Isolates No. of allelic types fimh ssel CRISPR1 CRISPR2 No. of MLST STs No. of PFGE patterns Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12; i: Montevideo Muenchen Saintpaul Total Total number of allelic types for fimh does not equal the sum of allelic types in each serovar, because the same allelic type was sometimes present in more than one serovar.

78 68 Gene Table 3.6. Allelic polymorphisms and nucleotide substitutions in the nucleotide sequences of fimh and ssel Serovar No. of polymorphic sites % of polymorphic sites No. of synonymous substitutions No. of nonsynonymous substitutions fimh Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12; i: Montevideo Muenchen Saintpaul Total ssel Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12; i: Montevideo Muenchen Saintpaul Total

79 69 Table 3.7. Analysis of CRISPR repeat sequences CRISPR Type Repeat sequences (5-3 ) 1 CRISPR1 Typical repeat CGGTTTATCCCCGCTGGCGCGGGGAACAC Repeat variant AGGTTTATCCCCGCTGGCGCGGGGAACAC Terminal repeats GTGTTTATCCCCGCTGACGCGGGGAACAC GTGTTTATCCCCGCTGGCGCGGGGAACAT CRISPR2 Typical repeat Same as the typical repeat in CRISPR1 Repeat variants GGGTTTATCCCCGCTGGCGCGGGGAACAC CAGTTTATCCCCGCTGGCGCGGGGAACAC CGGTTTATCCCCGCTGACGCGGGGAACAT CGGTTTATCCCCGCTAGCGCGGGGAACAC CGGTTTATCCCCGCTGACGCGGGGAACAC TGGTTTATCCCCGCTGGCGCGGGGAACAC CGGTTTATCCCCGCTGGCACGGGGAACAC CGATTTATCCCTGCTGGCGCGGGGAACAC CGGTTTATCCCTGCTGGCGCGGGGAACAC Terminal repeats ACGGCTATCCTTGTTGGCGCGGGGAACAC CGGTTTATCCCCGCTGCGCGGGGAACACT 1 Underscored nucleotides are SNPs, compared to the typical repeat.

80 70 Serovar Table 3.8. Analysis of CRISPR spacers in different serovars No. of unique spacers Avg no. of spacers + SD 1 Minimum no. of spacers Maximum no. of spacers CRISPR1 CRISPR2 CRISPR1 CRISPR2 CRISPR1 CRISPR2 CRISPR1 CRISPR2 Typhimurium Enteritidis Newport Heidelberg Javiana I 4, [5], 12; i: Montevideo Muenchen Saintpaul Total SD: value of standard deviation. 2 Number of total unique spacers does not equal the sum of unique spacers in each serovar, because a unique spacer was sometimes present in more than one serovar.

81 71 Table 3.9. Comparison of epidemiologic concordance 1 between PFGE and MLST based on virulence genes and CRISPRs for the selected strains analyzed in the present study Subtyping method Enteritidis Typhimurium Newport Heidelberg I 4, [5], 12; i: - Saintpaul Montevideo Muenchen MLST PFGE Values for epidemiologic concordance were calculated based on isolates with cluster codes identified by PulseNet. 2 The above values for epidemiologic concordance are biased against PFGE, because in some cases outbreaks that contained strains with variations in PFGE patterns (had poor epidemiologic concordance by PFGE) were deliberately selected in the present study.

82 72 CRISPR1 CRISPR2 Figure 3.1. Schematic view of the two CRISPR systems in Salmonella Typhimurium LT2. Direct repeats and spacers are represented by black diamonds and white rectangles, respectively. The terminal direct repeats are represented by white diamonds. L stands for leader sequence. cas genes are in grey while other core flanking genes (ygcf, iap and ptps) are in white. The figure is not drawn to scale.

83 73 (a) (b) Figure 3.2. (a) Cluster diagram based on only fimh and ssel. (b) Cluster diagram based on fimh, ssel and CRISPRs (combined allele of CRISPR1 and CRISPR2). ST: sequence type. T: Typhimurium. E: Enteritidis. N: Newport. H: Heidelberg. J: Javiana. I: I 4, [5], 12: i :-. Mvo: Montevideo. Mcn: Muenchen. S: Saintpaul. CRISPR1 and CRISPR2 were combined into one allele for the cluster analysis because CRISPR1 and CRISPR2 are spatially linked (50).

84 Discussion There are several important criteria to follow when selecting genetic markers to use in an MLST scheme. First, the selected genetic markers should exhibit adequate sequence variations to provide separation of unrelated strains (37). Secondly, genetic markers which provide epidemiologically meaningful information should be selected so that the MLST scheme can exhibit high epidemiologic concordance. Last but not least, genetic markers should be present in the genome within all strains of the species of interest. Previous studies demonstrated that MLST schemes based on Salmonella housekeeping genes showed poor discriminatory power when compared to PFGE (16, 28, 49). Inclusion of virulence genes into one published MLST scheme for subtyping S. Typhimurium increased discriminatory power to 0.98, which was comparable to that of PFGE (0.96) (17). Virulence genes provided epidemiologically meaningful separation and clustering of strains of Listeria monocytogenes (11). Besides virulence genes, CRISPRs were selected as markers in the current MLST scheme because they were found to be one of the fastest evolving genetic elements in bacterial genomes (43). In the present study, cluster analyses based on the two virulence genes and two CRISPRs accurately grouped isolates according to their specific serovars, except serovar Typhimurium and I 4, [5], 12: i :-, which were clustered together. As serovar I 4, [5], 12: i:- is a monophasic variant of serovar S. Typhimurium (14), our result is not unexpected. Virulence genes were previously found to provide accurate identification of different serovars of Salmonella in other studies as well (41, 47, 48). Addition of CRISPRs significantly increased discriminatory power (Fig. 3.2) compared to previously published MLST schemes, and the identification of individual outbreak clones was achieved (Table 3.2). For example, one MLST scheme based on three housekeeping genes

85 75 (manb, pduf, and glna) genes and one virulence gene (spam) identified one sequence type among 85 S. Typhimurium isolates and discriminatory power for the MLST scheme was 0 (16). Another MLST scheme targeted seven housekeeping genes, aroc, dnan, hemd, hisd, pure, suca, and thra, and identified 12 sequence types among a total of 81 S. Newport isolates, which also resulted in poor discriminator y power (0.61) (28). One MLST study based on virulence genes (hila, pefb and fimh), 16S rrna gene and housekeeping genes showed high discriminatory power (0.98) for subtyping S. Typhimurium (17); however, its capacity to discriminate strains from more clonal serovars such as S. Enteritidis was not tested. In conclusion, the MLST scheme described in the present study has superior discriminatory power, compared to previously published MLST schemes for subtyping the major serovars of Salmonella, especially for the highly clonal serovar S. Enteritidis. As mentioned previously, the isolates selected for this study were biased towards those that had poor epidemiologic concordance of PFGE, so future studies comparing of MLST and PFGE need to be performed using a nonbiased strain collection. Generally speaking though, the current MLST scheme showed high epidemiologic concordance for subtyping the major serovars of Salmonella, except Muenchen (E=0.39) (Table 3.9). All S. Muenchen isolates had different sequence types, except the two isolates, SMcn13 and SMcn15 from the orange juice outbreak (Table 3.2). Interestingly, the allelic types of fimh and ssel were the same for all the S. Muenchen isolates except isolate SMcn12 (Fig. 3.2a), which means CRISPR1 and CRISPR2 provided almost all of the discriminatory power in the case of S. Muenchen isolates (Fig. 3.2b). Perhaps PFGE lacks pattern diversity for S. Muenchen because it cannot detect the subtle, but epidemiologically important changes, detected by CRISPRs. Alternatively, CRISPRs may be evolving too fast for S. Muenchen outbreak investigations, either because the specific niche where S. Muenchen resides harbors a large number of different phage, and/or phage pools of S. Muenchen are very dynamic. Dramatic differences have been observed in the rate of spacer

86 76 acquisition between different eubacteria. In Streptococcus thermophilus, CRISPRs are very active and new spacer acquisition appear to be the primary mechanism of this species to defend against phage (12); however, the rate of new spacer acquisition in other bacteria such as E. coli appear to be much slower (50). The current MLST scheme provided greater separation of S. Enteritidis isolates than PFGE (Table 3.2). The predominant PFGE XbaI patterns for S. Enteritidis in the PulseNet database are JEGX and JEGX , which is problematic because this lack of PFGE pattern diversity sometimes makes it difficult to separate potential outbreak-related isolates from sporadic isolates (19). The discriminatory power of PFGE has been increased by the combination of multiple restriction enzymes (54). However, whether the increased discrimination caused potential loss of epidemiologic concordance was not addressed in that study. The present MLST scheme allowed separation of the two predominant PFGE patterns of S. Enteritidis isolates (Table 3.2) and resulted in high epidemiologic concordance (Table 3.9). CRISPRs provided most of the discrimination (Fig. 3.2b). CRISPRs in S. Enteritidis are evolving due to plasmids and/or phage present in the environment (52). Fortunately, the rate of spacer insertion and deletion in CRISPRs is slow enough such that they do not appear to change during an outbreak (Table 3.2). CRISPRs may also reflect the specific phage and plasmid pool in the environment and hence contain ecologically and geographically meaningful information for bacteria (32, 52). As a result, CRISPRs may be useful for tracing an outbreak clone of Salmonella to the specific farm or food processing plant which serves as the reservoir for the source strain of an outbreak. In conclusion, the current MLST scheme effectively subtyped the two most common PFGE patterns of S. Enteritidis and thus could enhance cluster detection and outbreak investigation capabilities. This MLST method has the potential to be integrated into public surveillance laboratories to complement PFGE for S. Enteritidis outbreak investigations.

87 77 It has been previously suggested that CRISPRs are poor epidemiological markers in enterobacteria due to the slow rate of spacer acquisition (50). However, that study only analyzed 16 complete Salmonella genomes for CRISPRs, and only four of them were from the same serovar as strains analyzed in the current study. Additionally, Touchon et al. only included in their study one isolate of serovars Typhimurium, Enteritidis, Newport, and Heidelberg, so the true value of CRISPRs for epidemiologic investigations could not be fully appreciated. Our study analyzed 26, 34, 15 and 20 isolates from these serovars, respectively, and demonstrated that CRISPR sequences may be implemented for epidemiologic investigations. We are currently testing this hypothesis using larger numbers of isolates obtained from current and past Salmonella outbreaks. This MLST scheme has several other advantages that make it a potential subtyping method for routine surveillance of Salmonella. First, the primers in this MLST scheme were designed to have the same annealing temperature for all four markers so that it can be conveniently performed in large-scale epidemiologic investigations. Second, the number of the markers targeted was minimized to two virulence genes and two CRISPRs so that time and expense can be saved during routine typing of Salmonella strains (37). Third, all four markers, fimh, ssel, CRISPR1 and CRISPR2, are present in the major serovars of Salmonella and also in all published genomes of Salmonella serovars, so the current MLST scheme is widely applicable. Although this MLST scheme shows great promise, future research is needed to further validate it for molecular epidemiologic purposes. For example, future research is needed using a random collection of isolates representing a larger number of outbreaks, or otherwise epidemiologically related isolates, to accurately compare the epidemiologic concordance of the present MLST scheme with PFGE. In conclusion, the MLST scheme described in the current study maybe an excellent subtyping method for tracking the farm-to-fork spread of the most prevalent serovars of Salmonella during outbreaks.

88 Acknowledgements We thank Dr. Bindhu Verghese for technical guidance throughout the study, especially for the idea of combining CRISPRs into one allele to construct the cluster analysis. We also acknowledge the Penn State Genomics Core Facility - University Park, PA for DNA sequencing. This study was supported by a U.S. Department of Agriculture Special Milk Safety grant to the Pennsylvania State University (contract: ).

89 References 1. Barrangou, R., C. Fremaux, H. Deveau, M. Richards, P. Boyaval, S. Moineau, D. A. Romero, and P. Horvath CRISPR provides acquired resistance against viruses in prokaryotes. Science. 315: Beranek, A., C. Mikula, P. Rabold, D. Arnhold, C. Berghold, I. Lederer, F. Allerberger, and C. Kornschober Multiple-locus variable-number tandem repeat analysis for subtyping of Salmonella enterica subsp. enterica serovar Enteritidis. Int. J. Med. Microbiol. 299: Brouns, S. J. J., M. M. Jore, M. Lundgren, E. R. Westra, R. J. H. Slijkhuis, A. P. L. Snijders, M. J. Dickman, K. S. Makarova, E. V. Koonin, and J. van der Oost Small CRISPR RNAs guide antiviral defense in prokaryotes. Science. 321: CDC Salmonella Annual Summary df 5. CDC Salmonella Annual Summary df 6. CDC Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce items --- United States, MMWR Morb. Mortal. Wkly. Rep. 57: CDC OutbreakNet Foodborne Outbreak Online Database CDC Investigation update: Multistate Outbreak of Human Salmonella Enteritidis Infections Associated with Shell Eggs.

90 80 9. CDC Investigation update: multistate outbreak of human Salmonella Montevideo infections CDC Preliminary FoodNet data on the incidence of infection with pathogens transmitted commonly through food States, Chen, Y., W. Zhang, and S. J. Knabel Multi-virulence-locus sequence typing identifies single nucleotide polymorphisms which differentiate epidemic clones and outbreak strains of Listeria monocytogenes. J. Clin. Microbiol. 45: Deveau, H., R. Barrangou, J. E. Garneau, J. Labonte, C. Fremaux, P. Boyaval, D. A. Romero, P. Horvath, and S. Moineau Phage response to CRISPR-encoded resistance in Streptococcus thermophilus. J. Bacteriol. 190: Efrain, M. R., M. A. Fair, R. Gautom, D. N. Cameron, S. B. Hunter, B. Swaminathan, and T. J. Barrett Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 3: Echeita, M. A., S. Herrera, and M. A. Usera Atypical, fljb-negative Salmonella enterica subsp. enterica strain of serovar 4,5,12:i:- appears to be a monophasic variant of serovar Typhimurium. J. Clin. Microbiol. 39: Endo, T., K. Ikeo, and T. Gojobori Large-scale search for genes on which positive selection may operate. Mol. Biol. Evol. 13: Fakhr, M. K., L. K. Nolan, and C. M. Logue Multilocus sequence typing lacks the discriminatory ability of pulsed-field gel electrophoresis for typing Salmonella enterica serovar Typhimurium. J. Clin. Microbiol. 43: Foley, S. L., D. G. White, P. F. McDermott, R. D. Walker, B. Rhodes, P. J. Fedorka- Cray, S. Simjee, and S. Zhao Comparison of subtyping methods for differentiating

91 81 Salmonella enterica serovar Typhimurium isolates obtained from food animal sources. J. Clin. Microbiol. 44: Foley, S. L., S. Zhao, and R. D. Walker Comparison of molecular typing methods for the differentiation of Salmonella foodborne pathogens. Foodborne Pathog. Dis. 4: Gerner-Smidt, P., K. Hise, J. Kincaid, S. Hunter, S. Rolando, E. Hyytiä-Trees, E. M. Ribot, B. Swaminathan, and Pulsenet Taskforce PulseNet USA: a five-year update. Foodborne Pathog. Dis. 3: Giammanco, G. M., C. Mammina, C. Romani, I. Luzzi, A. M. Dionisi, and A. Nastasi Evaluation of a modified single-enzyme amplified fragment length polymorphism (SE- AFLP) technique for subtyping Salmonella enterica serotype Enteritidis. Res. Microbiol. 158: Grimont, P. A. D., and F. Weill Antigenic formulae of the Salmonella serovars, 9th ed. WHO Collaborating Centre for Reference and Research on Salmonella. Institut Pasteur, Paris, France. 22. Grissa, I., and C. Drevet CRISPRs web-service Grissa, I., G. Vergnaud, and C. Pourcel CRISPRcompar: a website to compare clustered regularly interspaced short palindromic repeats. Nucl. Acids Res. 36:W Groenen, P. M., A. E. Bunschoten, D. Soolingen, and J. D. Errtbden Nature of DNA polymorphism in the direct repeat cluster of Mycobacterium tuberculosis; application for strain differentiation by a novel typing method. Mol. Microbiol. 10: Guard, J Evolutionary trends in two strains of Salmonella enterica subsp. I serovar Enteritidis PT13a that vary in virulence potential.

92 Hale, C. R., P. Zhao, S. Olson, M. O. Duff, B. R. Graveley, L. Wells, R. M. Terns, and M. P. Terns RNA-Guided RNA cleavage by a CRISPR RNA-Cas protein complex. Cell. 139: Hanning B. Irene, J.D. Nutt, Steven C. Ricke Salmonellosis outbreaks in the United States due to fresh produce: sources and potential intervention measures. Foodborne Pathog. Dis. 6: Harbottle, H., D. G. White, P. F. McDermott, R. D. Walker, and S. Zhao Comparison of multilocus sequence typing, pulsed-field gel electrophoresis, and antimicrobial susceptibility typing for characterization of Salmonella enterica serotype Newport isolates. J. Clin. Microbiol. 44: Hoe, N., K. Nakashima, D. Grigsby, X. Pan, S. J. Dou, S. Naidich, M. Garcia, E. Kahn, D. Bergmire-Sweat, and J. M. Musser Rapid molecular genetic subtyping of serotype M1 group A Streptococcus strains. Emerg. Infect. Dis. 5: Horvath, P., and R. Barrangou CRISPR/Cas, the immune system of bacteria and archaea. Science. 327: Hunter, P. R., and M. A. Gaston Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J. Clin. Microbiol. 26: Kunin, V., S. He, F. Warnecke, B. S. Peterson, M. Haynes, N. Ivanova, L. L. Blackall, M. Breitbart, F. Rohwer, D. Mcmahon, and P. Hugenholtz A bacterial metapopulation adapts locally to phage predation despite global dispersal. Genome Res. 18: Lan, R., P. R. Reeves, and S. Octavia Population structure, origins and evolution of major Salmonella enterica clones. Infect. Genet. Evol. 9: Lindstedt, B. -., M. Torpdahl, E. M. Nielsen, T. Vardund, L. Aas, and G. Kapperud Harmonization of the multiple-locus variable-number tandem repeat analysis method

93 83 between Denmark and Norway for typing Salmonella Typhimurium isolates and closer examination of the VNTR loci. J. Appl. Microbiol. 102: Lindstedt, B., E. Heir, E. Gjernes, and G. Kapperud DNA fingerprinting of Salmonella enterica subsp. enterica serovar typhimurium with emphasis on phage type DT104 based on variable number of tandem repeat Loci. J. Clin. Microbiol. 41: Lindstedt, B., E. Heir, T. Vardund, and G. Kapperud Fluorescent amplifiedfragment length polymorphism genotyping of Salmonella enterica subsp. enterica serovars and comparison with pulsed-field gel electrophoresis Typing. J. Clin. Microbiol. 38: Maiden, M. C. J Multilocus sequence typing of bacteria. Annu. Rev. Microbiol. 60: Pourcel, C., G. Salvignol, and G. Vergnaud CRISPR elements in Yersinia pestis acquire new repeats by preferential uptake of bacteriophage DNA, and provide additional tools for evolutionary studies. Microbiology. 151: Ross, I. L., and M. W. Heuzenroeder Use of AFLP and PFGE to discriminate between Salmonella enterica serovar Typhimurium DT126 isolates from separate foodrelated outbreaks in Australia. Epidemiol. Infect. 133: Ross, I. L., and M. W. Heuzenroeder A comparison of two PCR-based typing methods with pulsed-field gel electrophoresis in Salmonella enterica serovar Enteritidis. Int. J. Med. Microbiol. 299: Scaria, J., R. U. M. Palaniappan, D. Chiu, J. A. Phan, L. Ponnala, P. McDonough, Y. T. Grohn, S. Porwollik, M. McClelland, C. Chiou, C. Chu, and Y. Chang Microarray for molecular typing of Salmonella enterica serovars. Mol. Cell. Probes. 22: Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. L. Willems, K. E. Dingle, F. M. Colles, and J. D. A. Van Embden Comparative genotyping of Campylobacter jejuni

94 84 by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41: Sorek, R., V. Kunin, and P. Hugenholtz CRISPR a widespread system that provides acquired resistance against phage in bacteria and archaea. Nat. Rev. Microbiol. 6: Struelens, M. J Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin. Microbiol. Infect. 2: Tamada, Y., Y. Nakaoka, K. Nishimori, A. Doi, T. Kumaki, N. Uemura, K. Tanaka, S. Makino, T. Sameshima, M. Akiba, M. Nakazawa, and I. Uchida Molecular typing and epidemiological study of Salmonella enterica serotype Typhimurium isolates from cattle by fluorescent amplified-fragment length polymorphism fingerprinting and Pulsed-Field Gel Electrophoresis. J. Clin. Microbiol. 39: Tamura, K., J. Dudley, M. Nei, and S. Kumar MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24: Tankouo-Sandjong, B., A. Sessitsch, E. Liebana, C. Kornschober, F. Allerberger, H. Hächler, and L. Bodrossy MLST-v, multilocus sequence typing based on virulence genes, for molecular typing of Salmonella enterica subsp. enterica serovars. J. Microbiol. Methods. 69: Tankouo-Sandjong, B., A. Sessitsch, N. Stralis-Pavese, E. Liebana, C. Kornschober, F. Allerberger, H. Hächler, and L. Bodrossy Development of an oligonucleotide microarray method for Salmonella serotyping. Microb Biotechnol. 1: Torpdahl, M., M. N. Skov, D. Sandvang, and D. L. Baggesen Genotypic characterization of Salmonella by multilocus sequence typing, pulsed-field gel electrophoresis and amplified fragment length polymorphism. J. Microbiol. Methods. 63:

95 Touchon, M., and P. C. E. Rocha The small, slow and specialized CRISPR and anti- CRISPR of Escherichia and Salmonella. PLoS One. 5:e USDA-ERS Foodborne illness cost calculator: Salmonella Vale, P. F., and T. J. Little CRISPR-mediated phage resistance and the ghost of coevolution past. Proc. R. Soc. Lond., B, Biol. Sci. 277: Voetsch, A., T. Van Gilder, F. Angulo, M. Farley, S. Shallow, R. Marcus, P. Cieslak, V. Deneen, and R. Tauxe FoodNet estimate of the burden of illness caused by nontyphoidal Salmonella infections in the United States. Clin. Infect. Dis. 38: Zheng, J., C. E. Keys, S. Zhao, J. Meng, and E. W. Brown Enhanced subtyping scheme for Salmonella Enteritidis. Emerging Infect. Dis.13:

96 86 Chapter 4 Characterization of clinical, poultry and environmental Salmonella Enteritidis isolates using multilocus sequence typing based on virulence genes and CRISPRs Fenyun Liu 1, Subhashinie Kariyawasam 2, Bhushan M. Jayarao 2, 3, Rodolphe Barrangou 4, Peter Gerner-Smidt 5, Efrain Ribot 5, Edward G. Dudley 1 and Stephen J. Knabel 1* 1 Department of Food Science, the Pennsylvania State University, University Park, Pennsylvania 16802; 2 Animal Diagnostic Laboratory, Orchard Drive, the Pennsylvania State University, University Park, Pennsylvania 16802; 3 Department of Veterinary and Biomedical Sciences, the Pennsylvania State University, University Park, Pennsylvania 16802; 4 Danisco USA Incorporation, 3329 Agriculture Drive, Madison, Wisconsin 53716; 5 Centers for Disease Control and Prevention, Atlanta, Georgia * Corresponding author. Mailing address: 437 Food Science Building, The Pennsylvania State University, University Park, PA 16802, US. Phone: sjk9@psu.edu

97 Abstract Salmonella enterica subsp. enterica serovar Enteritidis has consistently been a major cause of foodborne salmonellosis in the United States. Two major food vehicles for S. Enteritidis are contaminated eggs and chicken. Improved subtyping methods are needed to accurately track specific strains of S. Enteritidis related to human salmonellosis throughout the poultry and egg food system. A multilocus sequence typing (MLST) scheme based on virulence genes (fimh and ssel) and CRISPRs (Clustered Regularly Interspaced Short Palindromic Repeats) was developed and used to characterize 34 clinical isolates, 70 poultry isolates and 63 hen house environment isolates of S. Enteritidis. A total of 27 sequence types (STs) were identified for the 167 S. Enteritidis isolates. The MLST scheme identified four persistent and predominate STs circulating among U.S. clinical isolates, and poultry and hen house environments in Pennsylvania. It also identified a potential environment-specific sequence type. Moreover, cluster analysis based on fimh and ssel identified three epidemic clones and one outbreak clone of S. Enteritidis, as well as 11 singletons. Significant differences in virulence gene sequences between singletons and the other STs suggested that singletons might have different virulence capacity than other STs. The MLST scheme may provide information about the ecological origin of S. Enteriditis isolates, potentially identifying strains that differ in virulence capacity.

98 Introduction In the United States, Salmonella is the leading cause of bacterial foodborne disease, with approximately 1.4 million human cases each year since 1996 (41). The second most-reported serovar of Salmonella for human diseases is Salmonella enterica subspecies enterica serovar Enteritidis (S. Enteritidis), which causes nearly as many human cases as S. Typhimurium, the most prevalent serovar (8). The major food vehicle for S. Enteritidis is shell eggs and 80% of the S. Enteritidis outbreaks and approximately 50,000 to 110,000 cases are egg-associated in the U.S. each year (6, 34). The most recent S. Enteritidis outbreak was associated with eggs, in which 1,519 people got infected (9). Another common food vehicle is chicken, consumption of which is considered as another risk factor for S. Enteritidis infections in humans (22, 30). The chicken and egg food system is complex and contains a large number of niches that may be potential sources of S. Enteritidis (Fig. 4.1) (21). S. Enteritidis has been isolated from a wide variety of animals, such as rodents, wild birds and insects, which could serve as potential reservoirs (17). Additional potential reservoirs for S. Enteritidis include chicken manure, sewage and other moist and organic materials in farm environments (6). Furthermore, oral S. Enteritidis infection in poultry could occur via contaminated water and feed (6). Infection of S. Enteritidis among chickens can spread rapidly by direct contact with infected birds or with contaminated materials within densely populated poultry houses (6, 17). Additionally, eggs can become contaminated internally when the ovaries of layers are colonized by S. Enteritidis; this process is called vertical transmission (17, 31). Eggs can also be contaminated externally by feces and environments containing S. Enteritidis; this is referred to as horizontal transmission (14). In order to control S. Enteritidis in poultry, one of the interventions employed on farms is egg quality assurance programs, which involve acquisition of S. Enteritidis free chicks, control of pests

99 89 (including rodents and insects), use of S. Enteritidis-free feeds, and routine microbiologic testing for S. Enteritidis in the farm environment (6). Characterization of S. Enteritidis isolates in different niches on chicken farms and from patients can help track dissemination of specific strains related to human salmonellosis and thus identify reservoirs and routes of transmission. Before considering the epidemiology of Salmonella, it is important to first define epidemic clone (EC), and outbreak clone (OC). Epidemic clone is a strain or group of strains descended asexually from a single ancestral cell (source strain) that is involved in one epidemic, and can often include several outbreaks (11). Outbreak clone is a strain or group of strains descended asexually from a single ancestral cell (source strain) that is involved in one outbreak (11). Several molecular subtyping methods have been developed to characterize strains and study the epidemiology of S. Enteritidis, including amplified fragment length polymorphism (AFLP) (27, 15, 19, 36, 38), multiple loci variable number tandem repeat analysis (MLVA) (3, 5, 12, 35), and pulsed-field gel electrophoresis (PFGE) (18). PFGE is currently the gold standard method used by public health surveillance laboratories for tracking foodborne pathogens including Salmonella (18). The main advantage of PFGE is its high discriminatory power (i.e. ability to separate unrelated strains) for subtyping most serovars of Salmonella. However, PFGE lacks discriminatory power for highly clonal serovars like S. Enteritidis (18, 42). For example, the most recent S. Enteritidis outbreak due to shell eggs was caused by the most common PFGE pattern for S. Enteritidis in the PulseNet database and thus not all isolates may be related to this outbreak (9). The lack of adequate discriminatory power makes it difficult to track a specific strain of S. Enteritidis in the food system. Besides occasional inadequate discriminatory power, PFGE does not provide appropriate information to infer phylogenetic relationships among subtypes (33). Another subtyping method, MLVA, was reported to have higher discriminatory power than PFGE for S. Enteritidis (3, 5, 12, 35). However, in some circumstances, strains that

100 90 had the same MLVA type were separated by PFGE (5). Moreover, replicates of the same strains of Salmonella have been shown to have different number of repeat units at a specific locus (7, 13), which makes accurate interpretation of results difficult. Compared to PFGE and MLVA, Multilocus Sequence Typing (MLST), which targets nucleotide sequence differences in several DNA loci, generates discreet, highly informative, highly portable and reproducible data. Moreover, MLST is a well-accepted tool for studying the population structure, evolution and diversity of bacteria (25). Recently, a new MLST scheme based on virulence genes (fimh and ssel) and CRISPRs (Clustered Regularly Interspaced Palindromic Repeats) was shown to provide better separation of S. Enteritidis than PFGE (29). CRISPRs encode tandem sequences containing bp DRs (direct repeats) and spacers of similar size (Fig. 4.2). Spacers are short DNA sequences obtained from foreign nucleic acids such as phages or plasmids and are inserted into bacterial chromosome to protect them from infection by homologous phages and plasmids (2). Therefore, CRISPRs reflect the specific phage and plasmid pools in the environment and hence contain ecological and geographic information of the bacteria present there (24, 40). As a result, CRISPRs might be useful for tracing back a clone of S. Enteritidis to the specific niches in the farm or food processing plant where it originated. Therefore, the objective of the present study was to characterize S. Enteritidis isolates from different sources using this scheme to investigate its epidemiology. With a better understanding of the epidemiology of S. Enteritidis, more effective intervention strategies can be established to prevent future S. Enteritidis outbreaks due to eggs and poultry meat.

101 Materials and methods Bacterial isolates and DNA extraction. A summary of all Salmonella Enteritidis isolates used in this study are listed in Table 4.1. A total of 167 isolates were obtained as follows: 34 from Centers for Disease Control and Prevention (CDC), 86 from the Pennsylvania Egg Quality Assurance Program (PEQAP) and 47 from the Animal Diagnostic Lab (ADL) at the Pennsylvania State University (Table S2). All 34 clinical isolates were previously analyzed (29), among which 32 isolates were collected from 11 S. Enteritidis outbreaks and the other 2 isolates were sporadic isolates. Bacterial isolates were stored at -80 C in 20% glycerol. When needed, isolates were grown overnight in Tryptic Soy Broth (TSB) (Difco Laboratories, Becton Dickinson, Sparks, MD) at 37 C. For all isolates, DNA was extracted using the UltraClean Microbial DNA extraction kit (Mo Bio Laboratories, Solana Beach, CA) and stored at -20 C before use. PCR amplification. Primers for all four markers were designed based upon consensus alignments of the published S. Typhimurium LT2 genome (accession number AE006468) using Primer 3.0 ( (Table 4.2). PCR amplifications were performed using a Taq PCR master mix kit (Qiagen Inc., Balencia, CA) in a Mastercycler PCR thermocycler (Eppendorf Scientific, Hamburg, Germany). A 25 µl PCR reaction system contained 12.5 µl Taq PCR master mix, 9.5 µl PCR-grade water, 1.0 µl DNA template, 1.0 µl forward primer (final concentration, 0.4 µm) and 1.0 µl reverse primer (final concentration, 0.4 µm). A single PCR cycling condition was used to amplify all four markers (initial denaturation at 94 C for 10 min; 28 cycles of 94 C for 1 min, 55 C for 1 min, 72 C for 1 min; final extension at 72 C for 10 min). DNA sequencing. After PCR, products for sequencing were treated by adding 1/20 volume of shrimp alkaline phosphatase (1 U/µl, USB Corp. Cleveland, OH) and 1/20 volume of exonuclease I (10 U/µl, USB Corp). The mixture was then incubated at 37 C for 45 min to

102 92 degrade the primers and unincorporated dntps. After that, the mixture was incubated at 80 C for 15 min to inactivate the enzymes. PCR products were then sent to the Genomics Core Facility at the Pennsylvania State University for sequencing using the ABI Data 3730XL DNA Analyzer. In order to obtain complete DNA sequences of fimh and ssel, two more primers targeting the internal regions of these two genes were used together with the forward and reverse primers (Table 4.2). Both DNA strands of the amplicons were sequenced. Sequence analysis and sequence type assignment. For fimh and ssel, sequences were aligned and single nucleotide polymorphisms (SNPs) were identified using MEGA 4.0 (37). For CRISPR1 and CRISPR2, analyses of the spacer arrangements were performed using CRISPRcompar (20) and spacers were visualized as described by Deveau et al. (16). Different allelic types (ATs) (sequences with at least one-nucleotide difference or one-spacer difference in the case of CRISPRs) were assigned arbitrary numbers. The combination of 4 alleles (fimh, ssel, CRISPR1 and CRISPR2) determined its allelic profile and each unique allelic profile was designated as a sequence type (ST). Cluster analysis. Cluster analyses were performed based on allelic profile data by the unweighted pair group method with arithmetic mean (UPGMA) and results were visualized using the tree drawing tool on PubMLST ( CRISPR1 and CRISPR2 were combined into one allele for a more accurate cluster analysis, because CRISPR1 and CRISPR2 are likely to be spatially linked (39). Nucleotide sequence accession number. DNA sequences of the four genetic MLST markers were deposited in GenBank under accession numbers HQ to HQ

103 Results Results of MLST and sequence type distribution. In order to gain insights into the sources and routes of transmission of S. Enteritidis contamination, the 167 isolates were characterized using an MLST scheme based on virulence genes and CRISPRs which was previously developed in our laboratories (Dr. Stephen Knabel and Dr. Edward Dudley). fimh, ssel, CRISPR1 and CRISPR2 identified 12, 13, 14 and 20 allelic types, respectively. In total, 27 sequence types (STs) were identified for all 167 isolates (Fig. 4.6 and Table S2). There were 9, 12 and 15 STs in clinical, poultry and environmental isolates, respectively. For clinical isolates, the 9 STs were E ST1, 2, 3, 4, 5, 6, 7, 8 and 9. The number of clinical isolates in each sequence type is listed in Fig Out of the 9 STs found in clinical isolates, 5 STs (E ST2, 5, 6, 7 and 9) were not found in either poultry or environmental isolates (Fig. 4.6). Those 5 STs came from California, Georgia, Maine, Michigan and Ohio. For poultry isolates, the 12 STs included E ST1, 3, 4, 8, 12, 15, 21, 22, 23, 24, 25 and 26, and 7 of them (E ST15, 21, 22, 23, 24, 25 and 26) were only found in poultry isolates. For the 15 STs in environmental isolates (E ST1, 3, 4, 8, 10, 11, 12, 13, 14, 16, 17, 18, 19, 20 and 27), 10 STs (E ST10, 11, 13, 14, 16, 17, 18, 19, 20 and 27) were exclusively found in environment. Predominant STs. An uneven distribution of the 27 STs was observed between different sources. Overall, the 5 STs designated E ST1, 3, 4, 8 and 10 accounted for 19 %, 17%, 25%, 8% and 7% of the total isolates, respectively, accounted for 76% of all isolates. Out of the 5 predominant STs, 4 of them (E ST1, 3, 4, and 8) were found in clinical, poultry and environmental isolates. E ST1 made up 12% of clinical isolates, 33% of poultry isolates and 8% of environmental isolates, respectively. E ST3 was found in 15% of clinical, 9% of poultry and 2% of environmental isolates. E ST4 accounted for 15% of clinical isolates, 30% of poultry

104 94 isolates and 3 % environmental isolates. E ST8 comprised 21% of clinical isolates, 14% of poultry isolates and 40% of environmental isolates. E ST10 was only found in environmental isolates, where it comprised 21% of all environmental isolates. Cluster analyses. A cluster diagram based on virulence genes identified three epidemic clones (ECI, ECII and ECIII), which included STs from multiple outbreaks, were identified by fimh and ssel (Fig. 4.4). ECI contained 9 STs, E ST3, 4, 5, 8, 10, 12, 14, 18 and 27. In total, 110 (66% of total isolates) belonged to ECI, which was the largest cluster and contained 18 clinical, 41 poultry and 51 environmental isolates. ECII contained 22% of all isolates (8 clinical, 24 poultry, and 5 environmental isolates) and 3 STs (E ST1, 9 and 26). ECIII contained 3 STs (E ST2, 7 and 13) which included 6 clinical isolates and 1 environmental isolate. One outbreak clone (OC), E ST6, contained 2 clinical isolates from the same outbreak. Besides the 3 ECs and 1 OC, 11 singletons occupied a single branch on the tree. Among the 11 singletons, 5 STs (E ST11, 16, 17, 19 and 20) were found in the farm environment and 6 (E ST15, 21, 22, 23, 24 and 25) were found in poultry isolates. These 6 poultry singletons were either sampled from eggs in broiler hatcheries with hatchability problems or necropsy isolates from sick broilers. Incorporation of CRISPRs into the MLST method separated isolates within the 3 ECs (Fig. 4.5). Among the 15 STs in the 3 ECs, 4 STs (E ST1, 3, 4 and 8) were found in all sources (clinical, poultry and environmental). These STs were also the predominant STs among all isolates (Fig. 4.3). E ST12 was found in both poultry and environment. The other 10 STs were from a single source including 4 STs (E ST1, 2, 5 and 9) found only in clinical isolates, 2 STs (E ST26 and 27) found only in poultry isolates and 4 STs (E ST10, 13, 14 and 18) found only in environmental isolates. Spacer arrangements among STs. Fig. 4.6 shows the differences in spacer arrangements among STs in CRISPR1 and CRISPR2. In CRISPR1, the number of spacers ranged from 2 to 25; for CRISPR2, the number of spacers ranged from 3 to 25. Generally, there

105 95 were great similarities among STs in the 3 ECs and the OC. The singleton E ST16 also shared spacers with STs in the 3 ECs and the OC; however many other spacers were deleted. The other 10 singletons contained totally different spacers from each other, except E ST21 and E ST22, which shared most spacers within CRISPR1, and had identical CRISPR2 loci.

106 96 Table 4.1. Sources, sample types and isolation information for the 167 S. Enteritidis isolates analyzed in the present study Sources No. of isolates Clinical 34 Poultry 70 Environment 63 Samples States 1 Source 2 Year Clinical (stool; foods related to outbreaks) Poultry: 46 egg and necropsy isolates of broiler Poultry: 3 egg isolates from layer houses Poultry: 21 egg isolates from layer houses 13 states PA PA PA CDC ADL PEQAP PEQAP Environmental: 46 drag swabs PA PEQAP Environmental: 17 drag swabs PA PEQAP; ADL Clinical isolates were from 13 states, including CA, CO, CT, GA, ID, ME, MI, MN, OH, OR, PA, SC and WV. 2 Isolates were received from CDC (Centers for Disease Control and Prevention), PEQAP (Pennsylvania Egg Quality Assurance Program) and ADL (Animal Diagnostic Lab) in Pennsylvania State University.

107 97 Table 4.2. Primers used to amplify and sequence the four MLST markers Marker Primer sequence (5'-3') Note fimh CGTCGTCATAAAAGGAAAAA Forward primer for both amplification and sequencing GAACAAAACACAACCAATAGC CTCGCCAGACAATGTTTACT CATTCACTTCGCAGTTTTG Reverse primer for both amplification and sequencing Reverse primer for sequencing internal region Forward primer for sequencing internal region ssel AGGAAACAGAGCAAAATGAA Forward primer for both amplification and sequencing TAAATTCTTCGCAGAGCATC Reverse primer for both amplification and sequencing GGAGTTGAAAATCTTTGGTG Reverse primer for sequencing internal region TTTACCGAGAGAAAAGGTGA Forward primer for sequencing internal region CRISPR1 GATGTAGTGCGGATAATGCT GGTTTCTTTTCTTCCTGTTG Forward primer for both amplification and sequencing Reverse primer for both amplification and sequencing CRISPR2 ACCAGCCATTACTGGTACAC ATTGTTGCGATTATGTTGGT Forward primer for both amplification and sequencing Reverse primer for both amplification and sequencing

108 Figure 4.1. Potential routes of transmission of S. Enteritidis contamination throughout the egg food system. 98

109 99 CRISPR1 CRISPR2 Figure 4.2. Schematic view of the two CRISPR systems in Salmonella Enteritidis strain P Direct repeats and spacers are represented by black diamonds and white rectangles, respectively. The terminal direct repeats are represented by white diamonds. Numbers of direct repeats and spacers are represented by the numbers of diamonds and white rectangles, respectively. L stands for leader sequence. cas genes are in grey while other core flanking genes (ygcf, iap and ptps) are in white. Primer targeting sites are indicated by upward ponting arrows. The figure is not drawn to scale.

110 100 (a) Figure 4.3. Frequency of the five predominant sequence types (E ST1, 3, 4, 8 and 10) in clinical, poultry and environmental isolates. The five predominant sequence types accounted for 76% of all isolates analyzed in the present study. All unlabeled pie slices in Fig. (b), (c), (d) are STs unique to that given source, except the pie slice of E ST10. (b) (c) (d)