MINI!

I
I

 

23:0

.LIBRARY
Michigan State
University

 

This is to certify that the
thesis entitled

DESCRIPTIVE EPIDEMIOLOGIC AND MICROBIOLOGIC
STUDY OF SALMONELLA TENNESSEE ISOLATES
ASSOCIATED WITH PEANUT BUTTER- NATIONAL
OUTBREAK OF 2006-2007

presented by

Chan Hong Nguyen

has been accepted towards fulfillment
of the requirements for the

Master of degree in Epidemiology

 

 

WW

Major Professor’s Signature

(9 f/QV/AQO/&

Date

 

 

MSU is an Affinnative Action/Equal Opportunity Employer

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProlechreyClRC/Dateoue.indd

 

 

,,_—____.

DESCRIPTIVE EPIDEMIOLOGIC AND MICROBIOLOGIC STUDY OF
SALMONELLA TENNESSEE ISOLATES ASSOCIATED WITH PEANUT
BUTTER- NATIONAL OUTBREAK OF 2006-2007

By

Chau Hong Nguyen

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Epidemiology

2010

ABSTRACT

DESCRIPTIVE EPIDEMIOLOGIC AND MICROBIOLOGIC STUDY OF
SALMONELLA TENNESSEE ISOLATES ASSOCIATED WITH PEANUT BUTTER-
NATIONAL OUTBREAK OF 2006-2007

By

Chau Hong Nguyen

The number of outbreaks of foodbome bacterial diseases increases each year due to
newly emerging strains of pathogens or food vehicles that have not been previously
identified. The multi-state outbreak of Salmonella serotype Tennessee infections
associated with peanut butter, 2006-2007 was the first outbreak in the United States
associated with this food vehicle. Results of our investigation revealed that S. Tennessee
strains associated with the peanut butter outbreak are more likely to be highly invasive
than strains from non-outbreak sources, odds ratio (OR) = 4.025 (p-value=0.0088, 95%
confidence interval (CI): (1.419, 11.418). Descriptive epidemiologic study revealed that

S. Tennessee strains were more likely to be isolated from females than fi'om males.

Results of the epidemiologic and virulence characterization suggested that peanut butter
could have served as a conducive medium for S. Tennessee to express certain sets of
virulence genes leading to the observed high level of invasiveness. The occurrence of
this outbreak highlights the importance of hygienic practices in peanut butter

manufacturing plants for the prevention of such mass contamination.

I.Ill.l.ll'llll Iltuwllllilill rill.

ACKNOWLEDGEMENTS

I am grateful for my guidance committee, Dr. Mahdi Saeed, Dr. Seongbeom Cho, and Dr.
Hwan Chung, for assisting, guiding, directing, and supporting me through my lab work,

data analysis, and writing process of my thesis.

I am greatly indebted to my advisor and mentor, Dr. Mahdi Saeed who is a wonderful
professor, and has guided me through my undergraduate and graduate school years. Your
financial support, time, guidance and faith in me has directed and kept me on the right

path towards the completion of my project and achievement of my degree.

My special thanks to Kimberly Nemeth and Adam Sawyer (undergraduate students,
MSU), and Kristin Evon for their efforts and time in helping and assisting me in the lab

work of attachment and invasion assays.

This project was supported in part by the Michigan State University Microbiology
Research Unit (MRU), under the National Institute of Allergy and Infectious Disease
(N IAID), National Institute of Health (N ID), and US Department of Health and Human

Services, contract number N01 -AI-30058.

I would also like to thank my family, the foundation of who I am, for always being there
for me, keeping me in their prayers and my faith strong, and their encouragements for

me.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. vi
LIST OF FIGURES .......................................................................................................... vii
LIST OF ABBREVIATIONS .......................................................................................... viii
BACKGROUND ................................................................................................................... 1
Foodborne infections .......................................................................................................... 1
Salmonella .......................................................................................................................... 2
Taxonomy and nomenclature .......................................................................................... 2
Pathogenesis .................................................................................................................... 3
Clinical manifestations ................................................................................................... 4
Salmonellosis ...................................................................................................................... 5
Causes of salmonellosis .................................................................................................. 5
Transmission of Salmonella serotypes ........................................................................... 6
Prevention of salmonellosis ............................................................................................ 7
SALMONELLA SEROTYPE TENNESSEE ...................................................................... 8
Salmonella serotype Tennessee outbreak associated with peanut butter, United States,
2006—2007 ....................................................................................................................... 8
Salmonella serotype Typhimurium outbreak associated with peanut butter, United
States, 2008-2009 .......................................................................................................... 10
PFGE profiles of peanut butter derived isolates ........................................................... 11
Salmonella contamination of peanuts ........................................................................... 12
Recall of peanut butter and peanut containing products ............................................... 13
Societal impacts of the peanut butter associated outbreaks .......................................... 14

MOLECULAR, VIRULENCE TYPES, AND PHYLOGENETIC ANALYSIS OF

SALMONELLA TENNESSEE ISOLATES ..................................................................... 15

STUDY OBJECTIVES AND METHODS ....................................................................... 17
Objective 1: To describe the epidemiologic attributes of clinical cases ........................... 17
Objective 2: Virulence profiling of S. Tennessee isolates using attachment and invasion
patterns of Caco-2 cells model grown in tissue culture .................................................... 18
Objective 3: Genotypic characterization of S. Tennessee isolates by Multi-Locus
Sequence Typing (MLST) ................................................................................................ 20

iv

STATISTICAL ANALYSES ............................................................................................. 22

Descriptive analysis .......................................................................................................... 22
Univariable analysis .......................................................................................................... 22
Multivariable analysis ....................................................................................................... 23
RESULTS ............................................................................................................................ 25
Descriptive analysis of profiled S. Tennessee isolates ..................................................... 25
Descriptive epidemiologic analysis of all isolates received ............................................. 26
Human isolates .................................................................................................................. 27
Interaction of S. Tennessee with Caco-2 cells .................................................................. 28
Univariable analysis .......................................................................................................... 28
Multivariable analysis ....................................................................................................... 29
Genomic analysis of S. Tennessee isolates ....................................................................... 29
DISSCUSSION ................................................................................................................... 30
CONCLUSION ................................................................................................................... 33
TABLES .............................................................................................................................. 35
FIGURES ............................................................................................................................ 49
REFERENCES ................................................................................................................... 56

LIST OF TABLES

Table 1. Selected Salmonella foodbome outbreaks and associated food vehicle(s) ............ 35
Table 2. Salmonella Tennessee infections associated with peanut butter, United States,
2006-2007, case findings ...................................................................................................... 36
Table 3. Findings from case-control study conducted by state and local health

departments .......................................................................................................................... 37
Table 4. Selected peanut butter and other peanut containing recall products ...................... 38
Table 5. S. enterica subsp. Enterica — 7 housekeeping genes

(MLST Salmonella database) ............................................................................................... 39
Table 6. Source frequency counts and percents of profiled isolates .................................... 40

Table 7. PFGE patterns of profiled human S. Tennessee isolates compared to all profiled S.
Tennessee isolates ................................................................................................................ 41

Table 8. Frequency counts and percentages of animal sources from S. Tennessee isolates

received ................................................................................................................................ 42
Table 9. Gender and site distribution of profiled S. Tennessee from human sources .......... 43
Table 10. Gender and site distribution of all human isolates received ................................ 44
Table 11. Virulence profiling of profiled S. Tennessee isolates with sources ..................... 45

Table 12. Univariable analysis of outcome highly invasive with selected predictor variables
for profiled human S. Tennessee isolates ............................................................................. 46

Table 13. Univariable analysis of outcome highly invasive with variable outbreak for all
profiled S. Tennessee isolates ............................................................................................... 47

Table 14. Multivariable analysis of peanut butter exposure controlling for age and gender.48

vi

LIST OF FIGURES

Figure l. The pathogenic spectrum for Salmonella infections ............................................. 49
Figure 2. Reported cases of salmonellosis, United States, 1976-1999 ................................ 50
Figure 3. Reported incidence of salmonellosis, United States, 1997-2007 .......................... 51

Figure 4. Salmonella strains of other serotypes found in food and in environmental samples
during Salmonella Typhimurium outbreak investigation ..................................................... 52

Figure 5. Multi Locus Sequence Typing (MLST): 7 housekeeping genes

(Salmonella MLST database) ............................................................................................... 53
Figure 6. Sources of profiled S. Tennessee isolates ............................................................. 54
Figure 7. Neighbor-Joining tree of S. subsp. Enterica based on 7 MLST genes .................. 55

vii

OR

CI

CDC

WHO

DT104

PFGE

FDA

PCA

Caco-2

ATCC

DMEM

FBS

TSA

BHI

PBS

IGM

MLST

ST

USDA

LIST OF ABBREVIATIONS

Odds Ratio

Confidence Interval

Centers for Disease Control and Prevention
World Health Organization
Definitive Type 104

Pulse Field Gel Electrophoresis

Food and Drug Administration
Peanut Corporation of America
adenoCArcinoma of the Colon
American Type Culture Collection
Dulbecco’s Modified Eagles Medium
Fetal Bovine Serum

Tryptic Soy Agar

Brain Heart Infusion

Phosphate Buffer Saline

Intracellular Growth Medium
Multi-locus Sequence Typing
Sequence Type

United States Department of Agriculture

viii

BACKGROUND

Foodbome infections:

Foodbome bacterial infections are among the most common infections encountered by
millions of people all around the world. These infections occur due to consumption of
food or water contaminated with bacterial pathogens or their toxins, introduced into food
products due to failure of effective implementation of food safety principles during pre
and post harvesting stages of food (1). These stages include food production,
transportation, processing, distribution, preparation, and consumption (2). In 2006, more
than 3,300 foodbome illnesses were reported to the Centers for Disease Control and
Prevention's (CDC) Foodbome Disease Outbreak Reporting System (3). Foodbome
infections have emerged to become a worldwide problem (2). Common bacterial
pathogens associated with foodbome infections are Salmonella, Campylobacter, Listeria,
and Escherichia coli (4). An estimate of 76 million illnesses each year in the United
States are due to foodbome infections, of which 325,000 lead to hospitalization and 5,000
lead to death (5, 6). Surprisingly, of the 76 million foodbome illnesses, only 14 million
are caused by known agents, whereas the causative pathogens of the remaining 62 million
are unknown (6). Complete statistics of the frequency of foodbome illnesses for
calculation of incidence along with global burden and cost of the disease have continued
to be a challenge for investigators due to underreporting of the disease, or newly
emerging pathogens or agents that has not been identified previously (5, 6, 7).
Continuous efforts are required by public health investigators for improvement of food
safety and development of effective prevention strategies to decrease the occurrence of

foodbome infections (5, 6).

SALMONELLA
Salmonella is a gram-negative, rod-shaped, facultative anaerobe organism with motile
peritrichous flagella belonging to the family Enterobacteriaceae (8, 9, 10, 11, 12). Its
facultative anaerobe characteristic permits the bacteria to a wide range of growth and
activity in nutrient-poor, acidic, low water activity environments, for long periods (12).
Salmonella can exist and survive almost anywhere fi'om contaminated soil lasting over
200 days, to 10 months in dust, and over 4 years in dried whole eggs (13). Interaction

with food at temperature from 7 to 46° C can cause multiplication of the organism (13).

Salmonella was named after D.E Salmon, an American bacteriologist and veterinarian in
1886 when the first isolation of “hog cholera bacillus” was reported fi'om diarrheic swine
(12, 13, 14, 15). S. Enteritidis was isolated three years later fiom the meat and spleen of
a man who died after consuming raw meat from an ill cow, reported from the first
nontyphoidal salmonellosis foodbome outbreak (13). S. Typhimurium was first isolated

and reported in 1889 by Loeffler, fiom an ill laboratory mouse colony (13).

Taxonomy and nomenclature:

The taxonomy and nomenclature for Salmonella has been in consistent debate among
international countries over the years (12, 13, 14). In accordance with the World Health
Organization (WHO), it is now accepted that there is only one single species of
Salmonella called S. enterica with seven subspecies: I, II, IIIa, IIIb, IV, V, VI (12, 14,
16). The Kauffmann-White scheme is a detailed antigenic formulae and grouping of each

Salmonella serotype adopted for use in 1934 (14). Different strains and isolates of

Salmonella species are classified and assigned specific names based on the Kauffmann-

White scheme of serological identification (11, 12).

According to F. Kauffmann, “the species [genus Salmonella] is a group of related sero-
fermentative phage-types or a group of related sero-, bio-, phago- (or lyso-) types” (17).
Different antigens are expressed on the surface of the bacterial cells (18). Each
Salmonella serotype possess a unique antigenic formula derived fiom the combination of
antigens expressed (18). Somatic or outer membrane antigens are known as O antigens,
H is classified as the flagella antigens, and very few Salmonella bacterial cells produce
the capsular antigens, Vi (18). The name for each Salmonella serotype is related to the
disease caused to the host or after the place in which the organism was first isolated (14).

This latter approach is more commonly used for naming new Salmonella types (14).

Pathogenesis:

The pathogenic spectrum for Salmonella infections consist of various steps displayed in
figure 1. Transmission of the bacteria to a susceptible host is the first step in the disease
process (16). Upon entry into the host via ingestion, the bacteria must survive the
stomach acidity in order to initiate an infection (12). Initiation of an infection occurs
when the bacteria attaches to the small intestine and penetrates the mucosa layer of the
intestine for entry into the midlayer, and are then engulfed into the epithelial cells
achieved with the help of adhesins (12). The organism will penetrate the epithelial cells
into the small bowel and colon causing an inflammatory response leading to

inflammation and destruction of tissues (12). Furthermore, different serovars can

penetrate further into the deeper layers of the mucosa of the intestinal wall (12).
Inflammatory reactions elicited in the small intestine results in fever and diarrheal

symptoms, observed in most infected patients (12).

Clinical manifestations:

Salmonella infections can cause a wide range of diseases in humans: enteric fever,
gastroenteritis, and bacteremia (11, 19). Gastroenteritis is the most common disease
among the three (11). Symptoms associated with Salmonella infections consist of fever,
diarrhea, abdominal pains, nausea, and occasionally vomiting that develop in 12 to 72
hours after infection and can last from 4 to 7 days or prolonged (20, 21). In most
individuals, the disease is not life threatening and patients will recover without requiring
treatment (20, 21). However, in some patients, especially the young, elderly, and
immunocompromised individuals, the disease can be severe requiring hospitalization or
treatment with antimicrobials such as ampicillin, trimethoprim-sulfamethoxazole, or
ciprofloxacin (20, 21). Additionally, bacteremia and focal infections are more likely to
develop in immunocompromised individuals (19). Complete recovery of bowel habits is
usually observed in individuals who develop diarrhea (21). In a small number of
diseased individuals, a condition known as Reiter’s syndrome can develop causing
irritation in the eyes, painful urination, and pains in the joints, lasting for months or years

leading to development of chronic arthritis (20).

SALMONELLOSIS
Salmonellosis is defined as “an infection with any serotype of the Salmonella bacteria”
(20).
Among the widely distributed foodbome diseases, salmonellosis is one of the most
common, occurring worldwide (21 , 22). Small Salmonella outbreaks among the general
population are usually contained. Large outbreaks have been reported in hospitals,
schools, restaurants, and institutions (22). Over 40,000 cases of salmonellosis are
reported to the CDC each year in the United States (22). However, this represents the tip
of the iceberg since based on systemic investigations by the CDC-FoodNet, Salmonella
serotypes infect approximately 1.4 million people annually in the United States, resulting
in greater than 100,000 physician office visits, over 16,000 hospitalization, and 2.3

billion in medical care and productivity costs (23).

Causes of salmonellosis:

Most of Salmonella serotypes can cause illness in humans and animals. However about
20 serotypes are more common reported in salmonellosis than other serotypes. Of these
serotypes, S. Enteritidis and S. Typhimurium are the most frequent (22). S. Enteritidis
has become the most common cause of food poisoning in the United States within the last
twenty years (15). According to the WHO Global Salm-Surveillence, fi'om 2000-2002, S.
Enteritidis was the most common serotype among human isolates, accounting for 65% of
all isolates; consequently, making it the dominant serotype of human isolates not only in

the United States but also in other industrialized nations (13, 24).

S. Typhimurium, including the Definitive Type 104 (DT104) strains is the second most
common strain and has been found to be resistant to several antimicrobials including
ampicilin, choramphenicol, streptomycin, sulfamethoxazole, and tetracycline (25). These
classes of antimicrobials are commonly used in treatment of human and animal
infections. (22). Multi-drug resistant S. Typhimurium, (DT104), constituted 7% of
human Salmonella infections in the United States in 1998 (26). Nontyphoidal
salmonellosis became a public health event in the United States after World War II (13).
In 1942, when salmonellosis was fist nationally reported, a total of 502 clinical cases
were observed (1 3). The number of reported cases of salmonellosis in the United States
from 1976-1999 are displayed in figure 2 (27). The highest number of reported cases
observed was in 1985 with approximately more than 65,000 cases. Reported cases of
salmonellosis have remained relatively stable within the last 10 years (27). The reported
incidence of salmonellosis in the United States from 1997-2007 are shown in figure 3.
The incidence rates observed within the 10-year period has remained within a range of

14.39 to 16.03 per 100,000

Transmission of Salmonella serotypes:

Salmonella infections in human is acquired via consumption of contaminated water and
food of animal origin or produce contaminated with animal products (7, 21). Acquisition
of Salmonella can also come from handling pets (22, 28, 29) Exposure to Salmonella in
the household can home when food products, particularly of animal origin, are prepared
poorly and inadequately cooked before consumption (22, 30). Moreover, food can also

be contaminated by people who are carriers and does not follow appropriate hygienic

methods when preparing meals, followed by inadequate or no cooking, allowing the

survival of the contaminants and their multiplication to a minimum infectious dose (12).

Primarily, Salmonella are reservoired in animals and the environment. Salmonella
serotypes have been isolated from most animal species particularly, reptiles.
Additionally, Salmonella are also widespread in the environment, in areas such as water
and soil, with long survival periods without multiplication (16). Animals are infected
with Salmonella through consumption of contaminated feed and water (12). The
estimated required dosage of the bacteria for disease initiation is 105 to 1010 bacteria,
however, as low as 10 organisms can initiate an infection (13, 16). Moreover, other
factors are also considered, such as the type of strain, the amount of contaminated food

consumed, and the physiological state of the host at time of infection (16).

Prevention of salmonellosis:

Various measures can be taken to prevent from Salmonella infections, such as avoiding
consumption of raw or undercooked food products and unwashed raw produce (22).
Other measures include careful hand washing after handling pets, especially reptiles, and
washing all food surfaces and utensils used for food preparation thoroughly (22). All
food, especially meat, should be cooked thoroughly to the required cooking temperature
before consumption because food contaminated with Salmonella will look and appear

normal, therefore making it difficult to detect by the naked human eye (22).

SALMONELLA SEROTYPE TENNESSEE

Outbreaks of foodbome illness are commonly caused by consumption of food
contaminated with Salmonella (14). Salmonella outbreaks were associated with various
groups of food vehicles (listed in Table 1). There are over 2,500 identified Salmonella
serotypes of which 20 are commonly associated with human salmonellosis (21, 31). S.
Tennessee is not a common serotype, thus infections are rare, and the average numbers of
reported cases each year during 1994-2004 were approximately 52 cases (31). The
average reported cases for S. Tennessee infections represent 0.01% of all reported
Salmonella serotypes (31). There has only been one earlier outbreak of S. Tennessee
infections associated with powered milk products and infant formula in the United States

and Canada, reported to CDC in 1993 (32, 33).

In November 2006, a substantial increase in the incidence of S. Tennessee infections was
reported from most of the United States. Subsequent investigation by the states health
authorities, CDC, and the FDA demonstrated that peanut butter consumption was
incriminated in most cases, underscoring the emergence of peanut butter as a new food

vehicle for human salmonellosis in the United States (31, 34, 35).

Salmonella serotype Tennessee outbreak associated with peanut butter, United
States, 2006-2007: (3, 31, 34)

On national bases, average numbers of S. Tennessee isolates were generally from l-5
isolates per month, whereas, in October of 2006, the numbers of reported isolates

increased to 30 isolates per month. A case-control study was then conducted during

February 5- 13, 2007 leading to identification of consumption of peanut butter with Peter
Pan or Great Value brand as the cause of the illness. Both of these brands were produced
at the same manufacturing plant, Sylvester, Georgia, and S. Tennessee strains were

isolated from several unopened and opened jars of the two brands. This outbreak leads to
the emerging and identification of peanut butter as a new food source for salmonellosis in

the United States.

Based on the Salmonella Annual Summary of 2006, the numbers of S. Tennessee isolates
reported more than doubled from 2005 to 2006 due to this first reported outbreak of S.
Tennessee resulting from contaminated peanut butter exposure in the United States. The
outbreak caused more than 800 cases in 47 states, accounting for the observed increase in
reported cases within the year. PulseNet serotyping of the number of isolates reported
fiom the outbreak strain identified 3 closely related pulsed-field gel electrophoresis

patterns (PFGE): JNXX01.0001, .0011, and .0026.

CDC officials defined a case for this outbreak as “infection with Salmonella Tennessee
with a PFGE pattern matching one of the three outbreak patterns (ending in .0001 , .0011,
or .0026) in a person residing in the United States with symptoms onset on or after
August 1, 2006”. The case findings for S. Tennessee infections associated with peanut
butter as reported by CDC are listed in table 2. Of all patients, the median age was 52
years, ranging fi'om 2 months to 95 years of age. Seventy-three percent of the patients
were female compared to 27% for males. Sources of isolates came from stools (61%),

urine (35%), and other specimens (4%). Additionally many reported symptoms of

diarrhea (72%), abdominal cramps (65%), fever (43%), and dysuria (45%). The high
percentage of cases reported for dysuria suggests that S. Tennessee is more likely to

infect the urinary tract than other serotypes.

A case-control study conducted by state and local health departments to identify the food
items associated with illness obtained 64 cases and 124 controls (Table 3). The median
age for case patients was 53 years, and 58 years for controls. It was observed that cases
were more likely than controls to have consumed peanut butter (81% versus 65%) (Table
3). Additionally, patients were also more likely than controls to have eaten either Peter
Pan or Great Value peanut butter (67% versus 13%) (Table 3). Epidemiologic data
provided to the Food and Drug Administration (FDA) officials suggested Peter Pan
peanut butter as the possible source of the outbreak. On February 14, 2007, a recall was

implemented for both brands, resulting in a decline of reported cases .

Salmonella serotype Typhimurium outbreak associated with peanut butter, United
States, 2008-2009: (35, 36)

A second national outbreak associated with peanut butter contaminated with Salmonella
Typhimurium in which larger numbers of children were infected was reported between
2008—2009. Infected individuals ranged from <1 to 98 years of age, with a median age of
16 years. Twenty-one percent of illnesses occurred in individuals <5 years of age. F orty-

eight percent of patients were females and 22% were hospitalized.

10

Interviews conducted with infected patients revealed that the outbreak occurred within 3
large institutions (2 care facilities and 1 elementary school) where the patients ate their
meals. Further investigation and review of food menus revealed a common food source
eaten by infected patients: King Nut creamy peanut butter produced by Peanut
Corporation of America (PCA). Interestingly, during this outbreak investigation, CDC’s
PulseNet identified and confirmed Salmonella strains of other serotypes other than
Typhimurium, found in food and environmental samples (Figure 4). It was reported that
an S. Tennessee isolate detected during this outbreak was noted to have a PFGE pattern
indistinguishable from the 2006-2007 S. Tennessee outbreak strains, obtained from
unopened and opened jars of King Nut brand peanut butter fiom Georgia and Minnesota,
leading to a possible association between the two outbreaks. Furthermore, the two
implicated plants are located approximately 70 km from one another. However, S.
Tennessee strains from this outbreak were not associated with an increase in illness in

humans.

PFGE profiles of peanut butter derived isolates: (31, 35)

Peanut butter derived isolates have unique PFGE profiles. S. Tennessee isolates from the
2006-2007 outbreak displays three closely related PFGE patterns: JNXX01.0010,
JNXX01.0011, and JNXX01.0026. S. Typhimurium on the other hand, displayed 2
clusters of unusual PFGE patterns. The first cluster consisted of 13 S. Typhimurium
isolates with PF GE pattern JPXX01.1818. The second cluster consisted of 41 S.
Typhimurium isolate with PFGE patterns JPXX01.0459/JPXX01.1825. Isolates fi'om the

two clusters were noted to be similar in patterns and testing of the isolates confirmed that

11

the two clusters displayed the same pattern: J PXA26.0462. A different sub-typing
method, multilocus variable-number tandem-repeat analysis, showed that the two isolates
were indistinguishable and were epidemiologically similar. As a result, the two clusters

are grouped together as a single outbreak strain.

Salmonella contamination of peanuts:

Contamination of peanuts with Salmonella can occur during growth, harvest, or storage
of peanuts (31). Peanuts can easily be contaminated due to its low water activity and
high fat content characteristics in which Salmonella favors and can survive indefinitely
(31, 35). Although peanuts are initially roasted at 350 ° F to sufficiently kill existing
Salmonella, however, these organisms can be reintroduced after heat treatment in the
post-processing steps fi'om salmonellae brought into the plant on raw peanuts, humans

fiom outside environment, or animals around and in the production environment (3 5).

The S. Tennessee outbreak in peanut butter was the first outbreak in the United States
implicated for this broadly distributed food item (31). The S. Tennessee and S.
Typhimurium outbreaks outline peanut butter as a potential cause of widespread illness
(3 5). Additionally, these two outbreaks have planted upon manufacturing plants an
important message: even when a heat-treatment step is used, processed food can become
contaminated again. Food-processing plants have underscored this prevention measure,

leading to contamination of processed food as a result (48).

12

Recall of peanut butter and peanut containing products: (37, 38)

Peanut butter and peanut butter paste are widely distributed to various distributors and are
common ingredients in products such as cookies, crackers, cereal, candy, ice cream, pet
treats, and food toppings, consumed on a daily basis. The S. Tennessee and S.
Typhimurium outbreaks associated with peanut butter in the United States has lead to the
recall of various food items. A peanut butter and other peanut containing products recall
list generated by the FDA in January of 2009 entails more than 2,100 products in 17
categories of which are recalled by more than 200 companies (Table 4). The recall list
includes not only human food but also pet food related to peanut products. Over $30,000
worth of peanuts and peanut products containing PCA peanuts was seized by United
States Marshall at Westco/Westcott due to possible Salmonella contamination. The recall
of peanut butter and peanut containing food products have resulted in the PCA filing for

bankruptcy.

As a result of the S. Tennessee and S. Typhimurium peanut butter outbreaks, retailers
were advised by the FDA to stop selling recalled products. Institutions and food service
establishments were asked to check their food products to ensure that they were not
serving any food listed on the recall list. Customers were advised to consult their health
care providers if they have been ill from eating peanut products. Customers were also
informed to throw away any peanut products in the home that are listed on the recall list
or will be recalled, and avoid consumption of any peanut food products that they think
are potentially contaminated. Manufactures were asked to inform customers through the

FDA website whether or not their products contain peanuts from the PCA.

l3

Societal impacts of the peanut butter-associated outbreaks: (39)

Peanut butter and peanut containing food products became an important concern and the
outbreaks were the beginning of a food safety crisis. Due to the recall issued by the FDA
for manufactures and retailers to remove potentially contaminated products off their
shelves, this resulted in a dramatic decline in the sales of peanut butter and peanut
containing products. Stores across the country experienced approximately 25.9%
decrease in the number of peanut sales. As a result of the Salmonella outbreaks
associated with peanut butter, peanut butter has left a negative impact on consumers of
this product. Extra measures should be taken by manufactures and retailers to educate

consumers and diminish any negative impact.

14

VIRULENCE TYPES AND PHYLOGENETIC ANALYSIS OF SALMONELLA
TENNESSEE ISOLATES

As stated previously, the primary entry mode for Salmonella into its host is ingestion of
the bacteria. Once ingested, the bacteria can reach underlying sites by passing through
the intestinal lumen (40). It is noteworthy that “all Salmonella serotypes share the ability
to invade the host (41).” However, the ability of each Salmonella serotypes to attach and
induce its own uptake into cells of the intestinal lumen depends on various factors such as
its growth state (40) and the outer surface proteins expressed (42). The specific
characteristic that each Salmonella serotype possess contribute and influence its
invasiveness (42). Additionally, the ability of each serotype to establish an infection with
its host relies on its ability to attach, colonize, and invade the host’s epithelial cells upon
initial contact with the epithelial cells (42). Moreover, although Salmonella isolates are
closely related, each strain differs in the degree of attachment and invasiveness to

intestinal cell lines (43).

The attachment and invasiveness of S. Tennessee isolates associated with the national
peanut butter outbreak were studied using tissue culture assays. Tissue culture is a
technique developed in the 19408 for assessment of animal cells as in vitro systems (44).
Used in many areas of science, this technique enables scientists and researchers to induce
and grow cells outside of the host (44). Development of tissue culture techniques allow
for assessment of a specific cell line growth, differentiation, and its specific role and
function as if it was in the host cell (45). Furthermore, tissue culture systems have been

established as models for use to study the attachment, invasion, and virulence

15

mechanisms of different Salmonella serotypes entry into mammalian cells (40, 42).
Different mammalian cells such as Henle cells, Caco-2 cells, and Hep-2 cells are used in
tissue culture systems for studying and understanding the natural processes of Salmonella
infections (44). Caco-2 cells (adenoCArcinoma of the Colon) are immortalized cell line
derived from human epithelial colorectal adenocarcinoma cells (42). When grown as a
monolayer to 80% confluency, Caco-2 cells will differentiate and resemble columnar
epithelial cells (46, 47, 48), of which can be use in attachment and invasion assay of

bacterial cells.

16

STUDY OBJECTIVES AND METHODS
The emergence of the national outbreak of Salmonella serotype Tennessee associated
with peanut butter in 2006-2007 has been the focus of this descriptive epidemiologic and

microbiologic study of S. Tennessee isolates associated with the national peanut butter

outbreak.

Objective 1: To describe the epidemiologic attributes of clinical cases:
Salmonellosis is a notifiable disease; therefore, cases are required to be reported to
community health departments for maintaining an updated disease record (49, 50). In
this study, S. Tennessee isolates were collected from several participating state
Departments of Health with epidemiological data when available. Sources are listed
below:
1. Seventeen isolates from Michigan Department of Community Health,
2. Nineteen isolates fiorn Cornell University and New York Department of Health,
3. Seven isolates from Indiana Department of Health,
4. Thirty nine isolates from Tennessee Department of Health,
5. Fifty isolates from Minnesota Department of Health.
6. Thirty seven S. Tennessee isolates were received from University of
Pennsylvania, and
7. Twenty three isolates were received from the USDA National Veterinary Service
Laboratory (NV SL) Ames. Iowa. These samples included of 22 isolates fi'om

animal sources and 1 isolate was fi'om animal feed.

17

Two references S. Tennessee isolates were received from University of Calgary. A total
of 194 isolates were procured. We will use SAS systems ver 9.1.3 to perform descriptive
epidemiologic study of the association between certain demographic epidemiologic

attributes and invasive S. Tennessee strains.

Objective 2: Virulence profiling of Salmonella Tennessee isolates using attachment
and invasion patterns of Caco-2 cells model grown in tissue culture:

We conducted virulence profiling of the S. Tennessee isolates using Caco-2 cells model
grown in tissue culture to perform attachment and invasion assays. The Caco-2 cells
used for this tissue culture project was isolated from a primary colonic tumor of a 72 year
old Caucasian male (ATCC HTB-37), and expresses characteristics of enterocytic
differentiation upon reaching 80% confluency. Cells were grown in complete growth
medium comprised of Dulbecco’s Modified Eagles Medium (DMEM) supplemented with
20% (vol/vol) Fetal Bovine Serum (FBS), 1% (vol/vol) nonessential amino acids, and
antibiotics to final concentrations of 100 ug/ml penicillin and 100 ug/ml streptomycin.
Cells were maintained at 37°C in 5% C02 and 95% air in T-75 flasks (Sarstedt,
Numbrecht, Germany) containing 10 ml of complete growth media for Caco-2 cell lines.
Cells were grown, feed every 2-4 days, and subcultured when reached 80% confluency

using trypsin to detach cells from flask wall.

S. Tennessee isolates were identified and labeled in preparation for attachment and

invasion assay. Two days before attachment or invasion, selected S. Tennessee isolates

were grown on Tryptic Soy Agar (TSA) plates. Next day, an isolated colony was picked

18

and inoculated into 10 ml Brain Heart Infusion (BHI) broth tubes for bacterial growth
overnight. Prior to attachment or invasion assay, each BHI tubes were vortextcd and

each bacterial isolate was plated to determine bacterial counts by dilution.

Attachment assays were performed using isolates grown with and without 1% D-
Mannose in the medium. For attachment assays, each individual wells in 24-well plates
were seeded with 3.31 x 105 Caco-2 cells and grown as a monolayer on sterile cover slips
to 80% confluency. Prior to attachment, wells were washed twice with incomplete
DMEM (no FBS or antibiotics) (Gibco), and approximately 5 x 108 bacterial cells grown
with and without 1% D-mannose were added to cover slips in each well with 2 ml of
complete DMEM containing no antibiotics. Twenty—four well plates were then incubated
for 1 hr in a 5% C02 atmosphere at 37 °C. After 1 hour, cells in 24-well plates were
washed 3 times with sterile Phosphate Buffer Saline (PBS), fixed with methanol for 15
minutes, and stained with a 1:7 dilution Giernsa stain (Sigma) for 1 hour. Each cover slip
was then removed fi'om the wells, rinsed with PBS twice and a final rinse with distilled
water and hung in coupling jars overnight to dry. Cover slips were mounted using
Permount (Fisher) on microscope slides (Fishers) and left to dry overnight for

examination by light microscopy the following day.

Invasion assays with Caco-2 cells were performed as described above except incubation
period was 3 hours. After 3 hours of incubation, each 24-well plates were washed once
with complete DMEM and intracellular growth medium (IGM) consisting of DMEM,

20% FBS, and 1 m1 of Gentarnicin per 100 ml was added and plates incubated. Plates

l9

and inoculated into 10 ml Brain Heart Infusion (BHI) broth tubes for bacterial growth
overnight. Prior to attachment or invasion assay, each BHI tubes were vortextcd and

each bacterial isolate was plated to determine bacterial counts by dilution.

Attachment assays were performed using isolates grown with and without 1% D-
Mannose in the medium. For attachment assays, each individual wells in 24-well plates
were seeded with 3.31 x 105 Caco-2 cells and grown as a monolayer on sterile cover slips
to 80% confluency. Prior to attachment, wells were washed twice with incomplete
DMEM (no FBS or antibiotics) (Gibco), and approximately 5 x 108 bacterial cells grown
with and without 1% D-mannose were added to cover slips in each well with 2 ml of
complete DMEM containing no antibiotics. Twenty-four well plates were then incubated
for 1 hr in a 5% C02 atmosphere at 37 °C. After 1 hour, cells in 24-well plates were
washed 3 times with sterile Phosphate Buffer Saline (PBS), fixed with methanol for 15
minutes, and stained with a 1:7 dilution Giemsa stain (Sigma) for 1 hour. Each cover slip
was then removed from the wells, rinsed with PBS twice and a final rinse with distilled
water and hung in coupling jars overnight to dry. Cover slips were mounted using
Permount (Fisher) on microscope slides (Fishers) and left to dry overnight for

examination by light microscopy the following day.

Invasion assays with Caco-2 cells were performed as described above except incubation
period was 3 hours. After 3 hours of incubation, each 24-well plates were washed once
with complete DMEM and intracellular growth medium (IGM) consisting of DMEM,

20% FBS, and 1 ml of Gentarnicin per 100 ml was added and plates incubated. Plates

l9

were replaced every hour with new IGM for 3 hours. After 3 hours of incubation, plates
were performed as described above for attachment. The purpose of addition of
Gentamicin in invasion assays is to kill any extra-cellular bacteria without affecting

grth of intracellular bacteria (51).

Objective 3: Genotypic characterization of S. Tennessee isolates by Multi-Locus
Sequence Typing (MLST):

Seven housekeeping genes derived from the Salmonella multi-locus sequence typing
(MLST) database was used for genotypic characterization of 60 S. Tennessee isolates by

MLST. These housekeeping genes are listed in table 5 and displayed in figure 5.

MLST is a technique used to characterize and categorize bacteria, providing a
standardized approach to data collection (52). The technique of MLST involves PCR
amplification of fragments followed by DNA sequencing with either the forward or
reverse PCR primer (52). The sequencing step of MLST will identify the variation
(polymorphic site) at a locus and assign a specific allele number for the observed
polymorphic site of the gene (53, 54). The alleles at the chosen 7 housekeeping gene loci
provide an allelic profile of the gene which in turn defines the sequence type (ST) of each
isolate (53). It is noted that isolates of the same serotype will display the same ST (54).
The sequencing fragments of the 7 housekeeping genes ranges from 430-500 base pairs
(53). Housekeeping genes of this size are chosen because it has been demonstrated that
DNA fragments of this length can be used in most bacterial pathogens for accurate

sequencing and identification of many different alleles within the population (53).

20

Distinction of genetic variation among isolates allows for characterization of bacterial
isolates (55). The procedure of MLST enables the examination of the relationship among

isolates by comparing their allelic profiles.

21

STATISTICAL ANALYSES

Descriptive analysis:

SAS Proc fieq procedure was used to compute frequency and percentages for each
categorical variable of interest. This procedure was also used to produce cross
tabulations tables to test for association of distribution of isolation sites of isolates by age

groups and gender of cases.

Univariable analysis:
Each variable was categorized as binary, ordinal, or multinomial variables. Proc logistic
procedure was used to perform unadjusted analysis of each individual variable of interest

with the outcomes.

One of the main variables of interest is outbreak. S. Tennessee strains with PFGE pattern
provided from participating states Department of Health ending in JNXX01.0010, .0011,
or .0026 were classified as associated with cases having a positive exposure to peanut

butter and related to the outbreak. S. Tennessee strains having other PFGE patterns were

labeled as non-peanut butter associated and was used as the reference group.

Age was analyzed as a categorical variable with 4 groups: <5 years, =5 and <16 years,

=16 and <65 years, and =65 years of age. Age group of =16 and <65 years was used as a

reference group. Male was used as the reference group for the variable gender.

22

Another variable of interest was site of isolation. Three groups were formed for this
variable: stool, urine, and other. The group other consisted of blood and wound sites.
These two sites were merged together to meet the convergence criteria for analysis

because the counts observed for blood and wound individually were small.

There are two main outcome variables of interest: attachment and invasiveness. The
outcome variable attachment was coded as 0 for negative attachment results and l for
positive attachment results of bacterial cells with Caco-2 cell models grown in tissue

culture. The interaction was assessed under a light microscope.

The second outcome variable of interest is invasiveness. Invasiveness results were
assigned to 2 categories based on the percentage of Caco-2 cells containing bacterial
micro-colonies as observed under a light microscope. Outcome for invasion was coded
as either highly invasive (greater than or equal to 75% of Caco-2 cells containing
bacterial micro-colonies as observed under a light microscope) or invasive (between 25
and 75% of Caco-2 cells containing bacterial micro-colonies as observed under a light

microscope).

Multivariable analysis:

In addition to the univariable analysis, a multivariable analysis was performed to test for
possible interactions. Possible interactions were: age and outbreak, gender and outbreak,
and site of isolation and gender. Variables that were tested for confounding was age and

gender. Confounding variables were tested by observing the percent change in the

23

unadjusted and adjusted OR estimates of the variables of interest. If the OR estimates for
the tested variable changed by more than 10%, this variable was considered as a

confounder.

24

RESULTS
We have performed attachment and invasion profiling on 96 randomly chosen isolates of
S. Tennessee from a variety of sources including human, food, animal, animal feed, and

environmental sources.

Descriptive analysis of profiled S. Tennessee isolates:

Descriptive epidemiologic data on cases of S.Tennessee infections received from
participating states department of health were not complete for all variables. Of the 55
profiled human isolates, 11 isolates had missing information for sex. Additionally, 9
isolate samples were missing age information of the patient. Not all isolates had a PFGE
pattern listed; therefore, it could not be determined if these isolates were associated with
the peanut butter outbreak or not. For each analysis, missing observations for the
explanatory variables were automatically deleted by SAS but missing observations are

included in the tables for comparison.

Proc freq analysis for the variable source revealed that 55 of the isolates tested were from
human, 23 from animal sources, 14 fiom food items (peanut butter, dry powered eggs), 2
were isolated from environmental sources, 1 from animal feed, and 1 isolate source was
unknown. Table 6 displays the frequency counts and percentages of the sources of the
profiled S. Tennessee isolates. Figure 6 displays a flowchart of the number of isolates
profiled and their corresponding sources. Of the 96 isolates profiled, 29 (43.28%) were
related to the peanut butter outbreak compared to 38 (56.72%) non-related isolates and 29

isolates were unknown.

25

0f the 55 profiled human isolates, information for PFGE patterns was listed for 27
(49.91%) isolates. Table 7 displays the PF GE patterns of profiled human isolates in
comparison to all profiled isolates. Approximately half of the profiled human isolates
had missing information for the variable PFGE. Of the 27 isolates with known PFGE
patterns, 16 of the isolates displayed one of the 3 closely related PFGE pattern reported
from the outbreak strain by CDC. Of all profiled isolates, 24 of the isolates displayed the

3 closely related PFGE pattern associated with the outbreak.

A total of 23 USDA samples were received of which 22 were from animal sources and l
was from animal feed. Table 8 is a listing of the distribution of animal sources for
profiled S. Tennessee animal isolates. Thirty-six percent of the animal samples came
from chicks, followed by 27% fiom cattle. S. Tennessee strains from animal sources
were non-peanut butter related to the outbreak. All 22 S. Tennessee animal isolates
displayed Mannose resistant local attachment under microscopic observation. An equal
number of profiled isolates (50%, 1 1/22) were observed to be invasive and highly

invasive.

Descriptive epidemiologic analysis of all isolates received:

A total of 194 isolates were received from 5 participating department of community
health centers (Michigan, New York, Minnesota, Tennessee, and Indiana), the USDA,
and University of Calgary and Penn State. For isolates received fi'om New York, only
information such as date of isolation and source of isolate were available, therefore they

were removed from the analysis of human isolates. Proc means analysis of 95 human

26

isolates with available information revealed that the mean age of patients were 52 years
of age with a range of 1 year to 94 years, similar to the profiled isolates. Additionally,
81.32% of the human isolates were isolated from patients between 16 years of age and
above. Eighty-nine human isolates had information listed for gender, of which 67%

were isolated fi'om females compared to 33% for males.

Human isolates:

A total of 55 human isolates were randomly chosen for attachment and invasion of Caco-
2 cell models in tissue culture. Age information was available for 46 patients. The mean
age of patients was 43 years of age, ranging fi'om 1 year to 94 years. Approximately
36.36% of human isolates were isolated from patients between 16 and 64 years of age
and 23.64% were from patients 65 years and over. The most common site of isolation for
human isolates was from stool (49.09%). Of the profiled human isolates, cross tabulation
table of gender by site of isolation revealed that 49.09% (27/55) percent of human
samples tested were fiom females of which 100% (27/27) were recovered from stool and
urine versus 82.35% (14/17) recovered from stool and urine for males (p.value= 0.0152).
Table 9 and 10 compares the gender and site distribution of profiled isolates and non-
profiled S. Tennessee isolates. S. Tennessee was more commonly isolated in stool and

urine of females than males.

Interaction of S. Tennessee with Caco-2 cells:

Ninety-three (97%) of the isolates tested displayed a positive attachment to the Caco-2

cell models maintained in tissue culture medium as recommended by ATCC instruction

27

for maintaining these cells. There were no differences in attachment patterns of isolates
observed between different sources under a light microscope. Additionally, isolates were
also tested whether they were sensitive or resistant to D-Mannose used at 1% in the tissue
culture medium. Of the 96 isolates tested, 3 were observed to be Mannose sensitive and
93 isolates expressed local attachment in the presence of 1% Mannose (Mannose
resistant). Table 11 displays the interaction profiles of tested S. Tennessee isolates with
corresponding sources. Fifty- eight percent (55/96) of the isolates tested were invasive,
compared to 42% (40/96) for highly invasive (Table 11). Comparison of invasion
patterns of all profiled S. Tennessee strains associated with the outbreak revealed that
75.86% (22/29) were observed to be highly invasive under microscopic examination.
Because 97% of the isolates were positive for attachment, therefore we could not assess
this variable with exposure variables of interest. Univariable analyses of outcome

variable attach with exposure variable of interests revealed insignificant results.

Univariable analysis:

The univariable analysis of outcome invasion with selected predictor variables of interest
for profiled S. Tennessee human isolates and all profiled S. Tennessee isolates are
displayed in table 12 and 13. For profiled human isolates, outbreak, age, and gender
were found to be insignificant at the 5% alpha value (Table 12). However, for all
profiled S. Tennessee isolates, it was observed that S. Tennessee strains associated with
the peanut butter outbreak were more likely to be highly invasive than strains from non-
outbreak sources OR= 4.025 (p-value=0.0088, 95% confidence interval (CI): 1.419,

11.418) (Table 13).

28

Multivariable analysis:

A multivariable analysis was performed to test for possible interactions. All interactions
included in the model were found to be insignificant. Age and gender were found to be
insignificant in the univariable analysis. Calculation of the percent change in the OR
estimates of the covariates age and gender revealed that age is a confounding variable.
The percent change in the OR estimate for age was 50.048 %. The percent change in the
OR estimate for gender was 0.490%; therefore gender is not a confounding variable.

However, gender was also controlled for in the multivariable model (Table 14).

Genomic analysis of S. Tennessee isolates:

Sixty S. Tennessee isolates were screened by MLST and Multi-locus VNTR Analysis
(MLVA). Two genotypes were identified by MLST and 3 were identified by MLVA.
Two sequence types (ST) were identified: a major type and minor type. Two non-
outbreak strains were identified for the minor type. Sequence variation was observed in
dnaN, hisD, and thrA. No sequence variation was observed in the 7 genes in isolates
during the outbreak period (August 2006 to June 2007). The neighbor- joining tree of S.

subspecies enterica based on the 7 MLST genes are displayed in figure 7.

29

DISCUSSION

The emergence of S. Tennessee in peanut butter has caused over 800 cases, perhaps many
more. The emergence of S. Typhimurium again in 2008-2009 lead to a higher number of
cases compared to S. Tennessee. The occurrence of the two outbreaks underscores the
ability of S. Tennessee as a newly emerging foodbome pathogen. In the United States,
newly recognized emerging foodbome bacterial pathogens and well recognized
pathogens are associated with new food vehicles. Examples of such recognized
foodbome pathogens known for causing most severe illness are: Camplylobacterjejuni,
Salmonella, Escherichia coli 01572H7, Listeria monocytogenes, and Toxoplasma (56,
57). These foodbome pathogens are foodbome zoonoses, having an animal reservoir
from which they spread to humans (58). Additionally, the spread of these foodbome
bacterial pathogens are global. Pathogens such as Salmonella has spread around the
world since the 19808, whereas S. Typhimurium DT 104 is appearing in North America
and Europe (58). New foodbome bacterial pathogens are being identified at an

increasing rate, suggesting that many more remain to be discovered (58).

The information available on the interaction of Salmonella spp. and host cells is limited.
There is little knowledge on the mechanism of bacterial attachment, colonization, and
invasion (59). The specific determinants involved in these processes have not been fully
identified or understood (59). The process of bacterial attachment to human epithelial
cells is an essential step for initiation of bacterial infections and bacterial invasion leads
to observations of disrupted Caco-2 cell membranes. Furthermore, the adherence of the

Salmonella spp. to the intestinal cell surface of the host is essential for initiation of

30

bacterial pathogenicity (59). Attachment and invasion profiling of S. Tennessee strains
were conducted to as a method to the virulecne of S. Tennessee strains associated with
the peanut butter outbreak, 2006-2007. Caco-2 cells, when grown to 80% confluency
differentiate and resemble columnar epithelial cells (46, 47 , 48). The use of Caco-2 cell
attachment assays for S. Tennessee bacterial strains revealed a high percentage of the
tested isolates to express attachment to this type of cell model. Furthermore, in the
presence of Mannose, S. Tennessee strains were Mannose-resistant as they displayed
attachment to Caco-2 cells in the presence of this sugar. The high percentage of the
tested isolates that attached to Caco-2 cells suggests that these strains of Salmonella
enterica serovar Tennessee expresses certain surface structures that plays a role in
adhering to Caco-2 cells (59), which may be essential for initiation of an infection. It was
also observed that S. Tennessee strains associated with the peanut butter outbreak were
highly invasive in comparison to isolates from other sources. This observation suggests
that S. Tennessee strains associated with the outbreak are more likely to be invasive than
strains from non-outbreak sources. The high level of attachment and invasiveness
observed within the S. Tennessee strains associated with the peanut butter outbreak could
be due to peanut butter as the source of food vehicle. Peanut butter could have served as
a conducive medium enabling the S. Tennessee strains to express certain virulence genes

involved in these processes.

Descriptive epidemiologic study of clinical cases revealed that a higher number of S.

Tennessee strains profiled and those not profiled were isolated fiom females than fiom

males. Similarly, CDC reported a higher percentage of female patients (73%) compared

31

to males. S. Tennessee infections are rare, however, they are more likely than other
serotypes to infect the urinary tract (31). Urinary tract infections are common among
females. A high proportion of isolates were from the urine, therefore possibly explaining

the high number of cases among females in comparison to males.

Salmonella infections are more prevalent among the young, immunocompromised, and
elderly. We observed a high number of isolate samples obtained from patients 65 years
of age and above. For the young, the number of isolate samples obtained was lower than
other age groups. However, information for age was unknown for 9 clinical samples out
of the 55 profiled isolate samples. It could be possible that these samples were isolated

from those less than 5 years of age, but this remains unknown.

32

CONCLUSION

All profiled S. Tennessee isolates procured fi'om participating health departments all
displayed attachment to the Caco-2 cells model grown in tissue culture, an essential
initial step leading to invasiveness of the strains. All profiled S. Tennessee strains were
observed to be invasive; however, S. Tennessee strains associated with the peanut butter
outbreak were highly invasive compared to invasive isolates from other sources or
isolates that were not associated with the outbreak. Food vehicles such as peanut butter,
having a high fat and low water content, could play a role serving as a conducive medium
for optimal growth and expression of certain surface structures that enhances the strain’s

attachment and invasiveness as observed in the S. Tennessee outbreak strains.

The emergence of known or newly recognized serotypes appear to be a continuous
challenging process for health communities as seen in the S. Tennessee outbreak and
again in S. Typhimurium, in which both occurred in a high fat low water content food
vehicle. The occurrence of these two outbreaks suggests that our perceptions need to be
reviewed from a perception that is based on the fact that food sources such as peanut
butter previously thought safe can be considered hazardous. Therefore, consumption of
peanut butter can be a risk factor in the etiology of sporadic non-typhoidal Salmonella
infections among adults and children. It is not sufficient enough to only educate food
producers, handlers, and consumers in basic food safety. More in house prevention
programs and testing should be sought to mitigate public exposure to emerging
contaminated food items and protecting consumers from severe illnesses resulting fiom

foodbome bacterial pathogens.

33

Food industries should use the occurrence of this outbreak to identify lessons to be
learned and develop applicable procedures for use among their industries. An increase in
more regular and sensitive testing policies for peanut butter and other food items for
Salmonella, prior to the release will improve the safety of these food products, prevent
distribution of contaminated peanut butter jars and future damaging outbreaks caused by

Salmonella.

34

Table 1. Selected Salmonella foodbome outbreaks and associated food vehicle(s)
[Sourcez compiled from various CDC-MMWR reports]

 

 

 

Year Serotype Outbreak source Ref
2009 Saintpaul Raw alfalfa sprouts (60,
61
2008-2009 Typhimurium Peanut Butter (35)
2008 Saintpaul Jalapeno peppers/ raw produce (62)
2008 Litchfield Cantaloupe (63)
2006-2007 Tennessee Peanut butter (31)
2007 l 4,5,12:i:- Frozen pot pics (64)
2006 Typhimurium Raw tomatoes (65)
2004 Braenderup & Roma tomatoes (66)
J aviana
2003 Enteritidis Raw almonds (67)
2004 Typhimurium Ground beef (61)
2002 Newport Beef (68)
2001 Kottbus Alfalfa sprouts (69)
2000-2002 Poona Cantaloupe (70)
2000-2001 Listeriosis Homemade Mexican-style cheese (71)
1999-2001 Enteritidis Raw shell eggs (72)
1999 Muenchen Orange juice (73)
1998 Agona Toasted oats cereal (74)
1996-1998 Enteritidis Raw shell eggs (75)
1994-1995 Enteritidis Raw shell eggs (76)
1994 Typhimurium Raw ground beef (77)
1994 Enteritidis Ice cream (78)
1994 Montevideo Beef jerky (79)
1990-1993 Enteritidis Ziti (baked) eggs (80)
1993 Tennessee Powered milk and infant formula (32)
1991 Enteritidis Raw shell egg (81)
1990 Enteritidis Bread pudding (82)
1989 Enteritidis Grade A shell eggs (83)
1981-1982 Dublin Raw milk (84)

 

35

 

Table 2. Salmonella Tennessee infections associated with peanut butter, United States,
2006-2007, case findings [Sourcez Morb Mortal Wkly Rep. 2007. 56; 521-4]

 

 

 

 

 

 

Variable %

Gender

Female 73

Male 27
Symptoms *

Diarrhea 72

Abdominal cramps 65

Fever 43

Dysuria 45
Source of isolates

Stool 61

Urine 35

Other specimens 4

 

 

*Symptoms onset dates were known for 481 of 628 patients and ranged from August 1, 2006-
April 23, 2007

36

Table 3. Findings from case-control study conducted by state and local health
departments [Source: Morb Mortal Wkly Rep. 2007. 56; 521-4]

 

 

 

 

 

 

 

 

 

 

 

 

Variable Controls Cases Matched 95% Confidence
(N=124) (N=65) Odds Interval (CI)
Ratio
No. % No. % (mOR)
Ate peanut butter 82 65 53 81 1.9 (0.8-5.2)
Ate peanut butter more than 50 40 43 66 3.5 (1.4-9.9)
once a week
Ate either Peter Pan or Great 16 13 44 67 10.9 (3.8-43.0)
Value peanut butter

 

37

 

Table 4. Selected peanut butter and other peanut containing recall products [Source:
http://www.accessdata.gfiQv/scriptsl/peanutbutterrecall/index.cfin]

 

Product recall category

Distributor (brand)

 

Brownie product recalls

Boston Cookies

 

Family Fresh Markets

 

Food Bonanza

 

Cake and pie product recalls

Awrey Bakeries

 

Food 4 Less

 

Kroger

 

Candy product recalls

Buffalo Chips

 

Brach’s

 

Fannie May

 

Cereal product recalls

Nature’s Path

 

Bear Naked

 

Michaelene’s

 

Cracker product recalls

Meijer

 

Keebler

 

Little Debbie

 

Dressing and seasoning product recalls

Kariba Farms

 

WOW

 

Fruit and vegetable product recalls

Eating Eight

 

Nutty Nanners

 

Trader’s Joe

 

Ice cream product recalls

#216 Schwan’s

 

Shop ‘11 Save

 

Sysco

 

Peanut product recalls

Marker Pantry

 

Piggly Wiggly
Peanut Corporation of America

 

Peanut butter product recalls

King Nut

 

Town & Country

 

Peanut Corporation of America

 

Peanut paste product recalls

Peanut Corporation of America or
Parnell’s Pride

 

Pet food product recalls

Breadfarrm

 

American Nutrition, Inc.

 

Next Gen Pet Products

 

Pre-packaged meals product recalls

Dinners ready

 

Reser’s

 

Ethnic Gourmet

 

Snack bar product recalls

Special K Protein

 

Nutrisystem

 

Kashi TLC

 

 

Topping product recalls

Barefoot Contessa

 

 

Best Choice

 

38

 

Table 5. S. enterica subsp. Enterica — 7 housekeeping genes (MLST Salmonella database)

 

 

 

 

 

 

 

 

 

 

Gene Protein Size (bp)
hemD Chorismate synthase 432
dnaN DNA polymerase III, beta-subunit 501
thrA Uroporphyrinogen III synthase 501
Pure Histidinal dehydrogenase 399
sucA Phosphoribosylarninoimodazole carboxylase 501
aroC 2-oxoglutarate dehydrogenase decarboxylase 501
component
hisD Aspartokinase I 501

 

39

Table 6. Source frequency counts and percents of profiled isolates

 

 

 

 

 

 

 

 

Source No. %
Human 55 57.30
Animal 23 23.96
Food 14 14.58
Feed 1 1 .04
Environmental 2 2.08
Missing 1 * 1 .04
Total 96 100

 

 

 

 

 

*1 S. Tennessee isolate source was unknown

40

Table 7. PFGE patterns of profiled human S. Tennessee isolates compared to all profiled
S. Tennessee isolates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PFGE patterns associated Profiled human isolates All profiled isolates
with S. Tennessee outbreak No. % No. %
JNXX01.001O 3 5.45 4 4.17
JNXX01.0011 9 16.36 16 16.67
JNXX01.0026 4 7.27 4 4.17
Other PFGE pgtterns
JNXX01.0001 3 5.45 4 4.17
JNXX01.0002 1 1.82 1 1.04
JNXX01.0012 2 3.64 2 2.08
JNXX01.0014 1 1.82 3 3.13
JNXX01.003O 1 1.82 1 1.04
JNXX01.0039 1 1.82 1 1.04
JNXX01.0049 2 3.64 2 2.08
Missing PFGE 28 50.91 58 60.42
Total 55 100 96 100

 

 

 

 

 

 

41

 

Table 8. Frequency counts and percentages of animal sources from S. Tennessee isolates
received

 

 

 

 

 

 

 

 

 

 

 

 

 

Animal type No. %
Alphaca l 4.55
Avian 2 9.09
Cattle 6 27.27
Chick 8 36.36
Goat 1 4.55
Swine 3 1 3 .64
Turkey 1 4.55
Total 22 1 00

 

42

Table 9. Gender and site distribution of profiled S. Tennessee fiom human source

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gender Profiled Site of isolation for profiled isolates from human sources
human
isolates N=55
No. % Stool Urine Other Missing
No. % No. % No. % No. %
Female 27 49.09 15 53.8 12 46.15 0 0 O 0
5
Male 17 30.91 12 70.5 2 11.76 3 17.64 0 O
9
Missing 11 20.00
Total 55 100

 

 

 

 

 

 

43

Table 10. Gender and site distribution of all human isolates received

 

 

 

Gender Profiled Site of isolation for profiled isolates
human
isolates N=95
No. % Stool Urine Other Missing
No. % No. % No. % No. %

 

Female 60 63.16 31 55.36 25 44.64 0 0 4 6.7

 

Male 29 30.53 20 76.92 4 15.38 2 7.69 3 10.34

 

 

 

 

 

 

 

 

Missing 6 6.31

 

 

 

 

 

Total 95 100

 

 

Table 1 1. Virulence profiling of profiled ' S. Tennessee isolates with sources

 

 

 

 

 

 

 

 

Interaction Sources

Profiles Human Animal Food Feed Environmental Total*
Attachment 53 23 1 3 1 2 92
positive 55.79% 24.21% 13.68% 1.05% 2.1 1% 96.84
Attachment 2 0 1 0 O 3
negative 2. l 1% 0% 1 .05% 0% 0% 3.16%
Highly 34 1 1 9 0 1 55
invasive 35.79% 1 1.58% 9.47% 0% 1.05% 57.89%
Invasive 21 12 5 1 1 40

22.11% 12.63% 5.26% 1.05% 1.05% 42.11%

 

 

 

 

 

 

 

*A total of 96 S. Tennessee isolates were profiled, l isolate source was unknown

 

 

' Profiled isolates are S. Tennessee isolates chosen randomly for virulence profiling of attachment and
invasion pattern using Caco-2 cells model grown in tissue culture. S. Tennessee isolates where virulence
profiling was not performed were classified as non-profiled.

45

Table 12. Univariable analyses of outcome highly invasive with selected predictor
variables for profiled human S. Tennessee isolates

 

 

 

 

 

 

 

 

 

 

Variables Odds Ratio 95% CI P-value Missing
Outbreak 4.444 (0.803, 24.609) 0.0876 11
Gender 0.521 (0.143, 1.893) 0.3218 25
Age: 9

<5 years 1.333 (0.196, 9.083) 0.7689
=5 and <16 years 1.666 (0.257, 10.789 0.5921
=65 year 0.778 (0.190, 3.187) 0.7269
Site: 6
urine 0.643 (0.185, 2.234) 0.4870
other 0.225 (0.018, 2.814 0.2472

 

* Male was used as a reference group for gender
Age group of =1 6 and <65 was used as a reference group for age
Stool was used as a reference group for site (other consists of blood and wound)

46

 

Table 13. Univariable analysis of outcome highly invasive with variable outbreak for all
profiled S. Tennessee isolates

 

Variable Odds Ratio 95% CI P-value Missing

 

Outbreak 4.025 (1.419, 1 1.418) 0.0088 29

 

 

 

 

 

 

 

47

Table 14. Multivariable analysis of peanut butter exposure controlling for age and gender

 

 

 

 

Variables Odds Ratio 95% CI P-value Missing
Outbreak 4.668 (0.491 , 44.364) 0.180 11
Gender 0.980 (0.084, 11.384) 0.987 25
Age: 9

<5 years 0.212 (0.085, 52.412) 0.648
=5 and <16 years 0.339 (0.027, 4.178) 0.399
=65 year 0.913 (0.083, 10.014) 0.940

 

 

 

 

 

 

 

* Male was used as a reference group for gender
Age group of =16 and <65 was used as a reference group for age

48

Figure 1. The pathogenic spectrum for Salmonella infections (12, 13)

Transmission

 

Transmission of the bacteria to susceptible
1 host via ingestion of contaminated products

8

Attachment, penetration, invasion

 

 

 

 

Attachment of bacteria to small intestine
Penetration into mucosal layers
2 Invasion of epithelial cells

0

Inflammation and destruction

 

 

 

 

3 Destruction of tissues and inflammatory
reactions resulting in fever, diarrhea, and
colitis

 

 

 

49

Figure 2. Reported cases of salmonellosis, United States, 1976-1999
[Source: Morb Mortal Wkly Rep. 2009. 56; 79-85]

 

70,000
60,000
3 50,000 ~ -
§ 40,000 -
'8

1% 30,000 ~
a: 20,000
10,000

 

 

 

 

 

 

 

 

 

50

.—a
5‘
U.

Reponedhnfidmux
:3 _.
Ur Ur

y.—
A

135

Figure 3. Reported incidence of salmonellosis, United States, 1997-2007
[Source: Morb Mortal Wkly Rep. 2009. 56; 79-85]

 

 

 

 

 

 

 

 

 

if

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Ykzr

51

Figure 4. Salmonella strains of other serotypes found in food and in environmental

samples during Salmonella Typhimurium outbreak investigation"
[Source: http://www.cdcw/flmonella/flphimurium/sfiains table.html|

 

PFGE image

 

 

 

 

 

 

 

Location Other DNA
of sample Salmonella fingerprint
origination serotypes ID
(PFGE
pattern)
Georgia Mbandaka TDRX01.001 1
Georgia Senfienberg JMPXOl .0049
Georgia Tennessee JNXX01.001 1
Minnesota Tennessee JNXXOI .0026

 

Source in
which
strain was
found

 

 

PCA plant
in Blakely,
Georgia-

floor crack

 

PCA plant
in Blakely,
Georgia-

floor crack

 

 

Unopened
container of
King Nut
brand

peanut
butter

 

 

 

Unopened
container of
King Nut
brand

peanut
butter

 

*These strains are not associated with an increase in human illness

52

 

Figure 5. Multi Locus Sequence Typing (MLST): 7 housekeeping genes
(Salmonella MLST database)

thrA

S. enterica

4.6 - 5.1 Mb

 

53

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ea as ea Se as: age age
as ea.“ ea.“ ens sea é: .Ee
E Nuz E i 2.2 i 2.2
Boh— 8958 BEES—Em 888“ Sam 82:8 E 82g :25: 8958 So...— Bousm gap:
_ _ _ _ _ _ _
5% 53
fig gee
$2 i i
gas 38% Baez Baggage
_ _ _
_
i
Ea as: ease .m

 

 

 

moan—ca 088989. .m. BEE“— .«o 805cm .0 unswE

54

egos com

mmaammmSmmmmaammSm

 

manages.“
«cod
.Illl
N3. 3% IT 2:

3.5 a e.
3.: 8: < Enema; f 2:

man“ 30?. < Esgm .
5 mm m E m 3.
53m t Z _

33w canon—«Baum . 2: am
32:50 eagafim . E 2
8:32.22 .3 sass A T
55.3. aaogo o ,
3.2.6 5583 e
«2.5 95:39...— 2:
3

859,50 $254— L
3.3m razoz 2 a
33m 88320: a ,
33m «885:
25:2 39:3 8 ‘
09$ .55 003030 .
are... $2283.38.“ 2 L
33m 28¢. om
aim Essay. ,
£3950 835M. 2:

:2 one 8380.. 2

«£28 .8 £35 em

a: 5:85. 2:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

V

 

 

 

 

 

macaw Ema—E e no woman «cream inane. .m we cob 9:50 7.22ch Z .h 0.5me

55

References

. Balkin K, Szumski B, Barbour S, et a1. Food-borne illnesses. New York: Greenhaven
Press Thomson Learning Inc., 2004.

. Van Beneden C, Keene W, Strang R, et a1. Multinational outbreak of Salmonella
enterica serotype newport infections due to contaminated alfalfa sprouts. Journal
American Medical Association 1999. 281; 158-62.

. CDC, Centers for Disease Control and Prevention. Salmonella annual summary, 2006.
Availabe at:
httn://www.cdc.gov/ncidod/dbmd/phlisdata/salmtab/2006/SglmonellaAnnualSumm
m2006.fif [Accessed on July 29, 2009].

. Helms M, Simonsen J, Molbak K. . Foodbome bacterial infection and hospitalization: a
registry-based study. Clinical Infectious Diseases 2006. 42; 498-506.

. Scallan, E. Activities, achievements, and lessons learned during the first 10 years of the
foodbom diseases active surveillance network: 1996-2005. Clinical Infectious
Disease 2007. 44; 718-25.

. Mead P, Slutsker L, Dietz V, et al. . Food-related illness and death in the United States.
Emerging Infectious Diseases 1999. Availabe at:
http://www.cdc.ggv/ncidod/eid/vo15no5/mead.htm [Accessed on August 13, 2009].

. Voetsch A, Van Gilder T, Angulo F, et a1. FoodNet estimate of the burden of illness
caused by nontyphoidal Salmonella infections in the United States. Clinical
Infectious Diseases 2004. 38; S1 127-34.

. Tietjen M, Fung D. Salrnonellae and food safety. Reviews in Microbiology 1995. 21;
53-83.

. Rao P, Riccardi SW, Birrer D. History of Salmonella. Availabe at:
http://www.columbigedu/cu/biologv/courses/g41 58/presentations/2004/Salmonella
,pp_t [Accessed on June 23, 2009].

 

10. Todar, K. Salmonella and salmonellosis. Availabe at:
httn://www/textbool_(ofb§cteriologv.net/salmonelléahtml [Accessed on June 23,
2009]

1 l. Bennasar A, Luna G, Cabrer B, et al. . Rapid identification of Salmonella typhimurium,
S. enteritidis, and S. virchow isolates by polymerase chain reaction based

fingerprinting methods. International Microbiology 2000. 3; 31-8.

12. Guthrie, R. Salmonella. CRC Press, 1992.

56

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Evans A, Brachman P. Bacterial infections of humans: epidemiology and control. New
York: Plenum Press, Plenum Publishing Corporation, 1998.

Bell C, Kyriakides A. Salmonella: a practical approach to the organism and its control
in food. United Kingdom: Blackwell Science, 2002.

Salmonella information. Availabe at: httn://www.salmonella.org[info.html [Accessed
on June 23, 2009].

Darwin, Miller. Interaction of Salmonella with the intestinal mucosa. Clin Microbiol
Rev 1999.

Kauffrnann, F. Serological diagnosis of Salmonella-species Kauffmann-White-Schema.
Denmark: Williams & Wilkins Company, 1972.

Kauffrnann, F. Serological diagnosis of Salmonella-species Kauffinann-White-Schema.
Denmark: Williams & Wilkins Company, 1972.

Hohmann, E. Nontyphoidal salmonellosis. Clinical Infectious Disease 2001. 32; 263-9.

CDC, Centers for Disease Control and Prevention. Salmonellosis. Availabe at:
http://wwwcdegov/salmonellg [Accessed on June 23, 2009].

WHO, World Health Organization. Drug-resistant Salmonella. Availabe at:
http://www.who.inUmediggmfidfaasheets/fs139/en/print.html [Accessed on June
23,2009]

Salmonellosis. Availabe at: http://www3.niaid.nih.gov/topics/salmonellosis/defult.htm
[Accessed on January 11, 2010].

Vugia D, Samuel M, Farley M, et a1. Invasive Salmonella infections in the United
States, F 00de, 1996—1999: incidence, serotype distribution, and outcome.
Clinical Infectious Diseases 2004. 38; S149-56.

Galanis E, Wong D, Patrick M, et a1. .Web-based surveillance and global Salmonella
distribution, 2002-2002. Emerging Infectious Diseases 2006. 12; 381-8.

Fiskk T, Lundberg B, Guest J, et a1. . Invasion infection with multidrug-resistant
Salmonella enterica serotype Typhimurium definitive type 104 among HIV-
infected adults. Clinical Infectious Disease 2005. 40; 1016-21.

Rabatsky-Ehr T, Whichard J, Rossiter S, et a1. . Multidrug resistant strains of

Salmonella enterica Typhimurium, United States, 1997-1998. Emerg Infect Dis
2004. 10; 795-801.

57

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

CDC, Centers for Disease Control and Prevention. Summary of notifiable disease,
United States, 2007. Morb Mortal Wkly Rep. 2009. 56; 79-85.

Younus M, Wilkins M. J, Davies H. D, et a1. Case-control study of disease

determinants for non-typhoidal Salmonella infections among Michigan children.
BCM Research Notes 2010. 3; 1-23.

Younus M, Wilkins M. J, Davies H. D, et al. The role of exposure to anumals and other
risk factors in sporadic, non-typhoidal Salmonella infections in Michigan children.
Zoonoses and Public Health 2010. 1-7.

CDC, Centers for Disease Control and Prevention. Preliminary FoodNet data on the
incidence of infection with pathogens transmitted commonly through food- 10
states, 2007. Morb Mortal Wkly Rep. 2008. 57; 366-70.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella
serotype Tennessee infections associated with peanut butter- United States, 2006-
2007. Morb Mortal Wkly Rep. 2007. 56; 521-4.

CDC, Centers for Disease Control and Prevention. Salmonella serotype Tennessee in
powdered milk products and infant formula- Canada and United States, 1993. Morb
Mortal Wkly Rep. 1993. 42; 516-7.

Salmonella serotype Tennessee in powdered milk products and infant formula, Canada
and United States, 1993. Morb Mortal Wkly Rep 1993. 42; 516-7.

Multistatc outbreak of Salmonella serotype tennessee infections associated with peanut
butter United Staste, 2006-2007. Journal American Medical Association 2007. 298;
33-5.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella
associated with peanut butter and peanut butter containing products- United States,
2008-2009. Morb Mortal Wkly Rep. 2009. 58; 1-6.

CDC, Centers for Disease Control and Prevention. Salmonella strains tables for
outbreak related to peanut butter and peanut butter containing products. Availabe

at: hgpzflwwwcdc.gov/salmonella/typhimurium/strains tablehtml [Accessed on
January 19, 2010].

Peanut butter and other peanut butter containing products recall list. Availabe at:

http://www.gccessdata.gov/scripts/peanutbutterrecall/index.cfrn [Accessed on
January 11, 2010].

FDA, US. Food and Drug Administration. Safety: peanut product recall. Availabe at:
http://wwww.fdaggov/Safetv/Recalls/ngorProductRecgls/Peanut/default.htrn
[Accessed on January 1 1, 2010].

58

39.

40.

41.

42.

43.

45.

46.

47.

48.

49.

50.

51.

Salmonella outbreak taints peanut butter sales. Availabe at:
http://blgg.nielsen.com/nielsenwire/consfumer/sgalmonella-outbreaLk-taints-peanut-
butter-sales/ [Accessed on January 11, 2010].

Lee C, Falkow S. The ability of Salmonella to enter mammalian cells is affected by
bacterial growth state. Proc. Natl. Acad. Sci 1990. 4304-08.

Ohl M, Miller S. Salmonella: a model for bacterial pathogenesis. Annu. Rev. Med.
2001. 52; 259-74.

Fadl A, Venkitanarayanan K, and Khan M. Identification of Salmonella Enteritidis
outer membrane proteins expressed during attachment to human intestinal epithelial
cells. Journal of Applied Microbiology 2002. 92; 180-6.

Mellor G, Goulter R, Chia T, et a1. Comparative analysis of attachment of shiga-
toxigenic Escherichia coli and Salmonella strains to cultured HT-29 and Caco-2
cell lines. Applied and Environmental Microbiology 2009. 75; 1796-9.

. Kirsop B, Doyle A. Maintenance of microorganisms and cultured cells: a manual of

laboratory methods. London: Academic Press Limited, Academic Press Inc., 1991.

Unchern, S. Basic techniques in animal cell cultures. Availabe at:
lip://www.phgrm.chula_.ac.th/surachai/aggdemic/stu/cell cglture handbookpgf
[Accessed on June 23, 2009].

Lee 1, Horn K, Bai G, et a1. NMR metabolic analysis of Caco-2 cell differentiation.
Journal of Proteome Research 2009. 30; A-E.

Finlay B, Falkow S. Salmonella intercations with polarized human intestinal Caco-2
epithelial cells. Journal of Infectious Diseases 1990. 162; 1096-106.

Engle M, Goetz G, and Alpers D. Caco-2 cells express a combination of colonocyte
and enterocyte phenotypes. Journal of Cellular Physiology 1998. 174; 362-9.

MDCH- communicable disease reporting in Michigan. Availabe at:
http://wwwmichgian.gov/mdch/O.1607,7-132-12538-,00.htm1 [Accessed on June
23,2009]

Massachusetts Department of Public Health, Bureau of Communicable Disease
Control. Guide to surveillance, reporting and control. Salmonellosis (non-
typhoidal) 2006. Availabe at:
http://www.ma_ss.gov/Eeohh_52/docs/dph/disease rgportingzguide/salmonellosisrtf
[Accessed on June 23, 2009].

Gahring L, Heffron, Finlay B, et a1. Invasion and replication of Salmonella
typhimurium in animal cells. Infection and Immunity 1990. 58; 443-8.

59

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64

Maiden, M. Miltilocus sequence typing of bacteria. Annu. Rev. Microbiol 2006. 60;
561 -88.

Enright M, and Spratt B. Multilocus sequence typing. Trends in Microbiology 1999. 7;
482-7.

Torpdahl M, Skov M, Sandvang D, et al. Genotypic characterization of Salmonella by
multiloc sequence typing, pulsed-field get electrophoresis and amplified fragment
length polymorphism. Journal of Microbiological Methods 2005. 63; 173-84.

Urwin R, and Maiden M. Multi-locus sequence typing: a tool for global epidemiology.
Trends in Microbiology 2003. 11; 479-87.

Tauxe, R. Food safety and irradiation: protecting the public from foodbom infections.
Emerging Infectious Diseases 2001. 7; 516-21.

Altekruse S, Cohen M, and Swerdlow D. Emerging foodbome disease. Emerging
Infectious Diseases 1997. 3; 285-93.

Tauxe, R. Emerging foodbome diseases: an evolving public health challenge.
Emerging Infectious Diseases 1997. 3; 425-34.

Mikcha J, Ferreira A, Ferreira C, et al. Hemagglutinating properties of Salmonella
Enteric serovar Enteritidis isolated from different sources. Brazillian Journal of
Microbiology 2004. 35; 54-8.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype
Saintpaul infections associated with eating alfalfa sprouts, United States, 2009.
Morb Mortal Wkly Rep 2009. 58; 1-3.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella

serotype Typhimurium infections associated with eating raw ground beef, United
States, 2004. Morb Mortal Wkly Rep 2006. 55; 180-2.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype

saintpaul infections associated with multiple raw produce iterns- United States,
2008. Morb Mortal Wkly Rep. 2008. 57; 929-34.

CDC, Centers for Disease Control and Prevention. Investigation update: outbreak of
Salmonella Litchfield infections, 2008. Availabe at:
http://www.cdc.gov/salmonella/litchfield/ [Accessed on June 23, 2009].

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella

infections associated with frozen pot pies, United States, 2007. Morb Mortal Wkly
Rep. 2007. 57; 1277-8.

60

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella

infections associated with raw tomatoes eaten in restaurants, United States, 2005-
2006. Morb Mortal Wkly Rep. 2007. 56; 909-11.

CDC, Centers for Disease Control and Prevention. Outbreaks of Salmonella infections
associated with eating Roma tomatoes, United States and Canada, 2004. Morb
Mortal Wkly Rep. 2005. 54; 325-8.

CDC, Centers for Disease Control and Prevention. Outbreaks of Salmonella serotype
Enteritidis associated with raw almonds, United States and Canada, 2003-2004.
Morb Mortal Wkly Rep. 2004. 53; 484-7.

CDC, Centers for Disease Control and Prevention. Outbreak of multidrug-resistant
Salmonella Newport, United States, January- April 2002. Morb Mortal Wkly Rep.
2002. 51; 545-8.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype
Kottbus infections associated with eating alfalfa sprouts, Arizona, California,
Colorado, and New Mexico, February-April 2001. Morb Mortal Wkly Rep. 2002.
5 1 ; 7-9.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella
serotype Poona infections associated with eating cantaloupe from Mexico, United
States and Canada, 2000-2002. Morb Mortal Wkly Rep. 2002. 51; 1044-7.

CDC, Centers for Disease Control and Prevention. Outbreak of Listeriosis associated
with homemade Mexican-style cheese, North Carolina, October 2000- January
2001. Morb Mortal Wkly Rep. 2001. 50; 560-2.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype
Enteritidis infection associated with eating shell eggs, United States, June 1999-
2001. Morb Mortal Wkly Rep. 2003. 51; 1149-52.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype
Muenchen infections associated with unpasteurized orange juice, United States and
Canada, June 1999. Morb Mortal Wkly Rep. 1999. 48; 582-5.

CDC, Centers for Disease Control and Prevention. Multistatc outbreak of Salmonella
serotype Agona infections linked to toasted oats cereal, United States, April-May,
1998. Morb Mortal Wkly Rep. 1998. 47; 462-4.

CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype

Enteritidis infection associated with eating raw or undercooked shell eggs, United
States, 1996-1998. Morb Mortal Wkly Rep. 2000. 49; 73-9.

61

76. CDC, Centers for Disease Control and Prevention. Outbreaks of Salmonella serotype
Enteritidis infection associated with consumption of raw shell eggs, United States,
1994-1995. Morb Mortal Wkly Rep. 1996. 45; 737-42.

77. CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella serotype
Typhimurium infection associated with eating raw ground beef, Wisconsin, 1994.
Morb Mortal Wkly Rep. 1995. 44; 905-9.

78. CDC, Centers for Disease Control and Prevention. Emerging infectious diseases
outbreak of Salmonella enteritidis associated with nationally distributed ice cream
products- Minnesota, South Dakota, and Wisconsin, 1994. Morb Mortal Wkly Rep.
1994.43; 740-1.

79. CDC, Centers for Disease Control and Prevention. Outbreak of Salmonellosis
associated with beef jerky, New Mexico, 1995. Morb Mortal Wkly Rep. 1995. 44;
7 85-7 .

80. Mishu B, Koehler J, Lee L, et al. Outbreaks of Salmonella Enteritidis infections in the
United States, 1985-1991. Journal of Infectious Diseases 1994. 169; 547-52.

81. CDC, Centers for Disease Control and Prevention. Outbreak of Salmonella Enteritidis
infection associated with consumption of raw shell eggs, 1991. Morb Mortal Wkly
Rep. 1992. 41; 369-72.

82. CDC, Centers for Disease Control and Prevention. Epidemiologic notes and reports
update: Salmonella Enteritidis infections and shell eggs, United States, 1990. Morb
Mortal Wkly Rep. 1990. 39; 909-12.

83. CDC, Centers for Disease Control and Prevention. Epidemiologic notes and reports
update: Salmonella Enteritidis infections and grade A shell eggs, United States,
1989. Morb Mortal Wkly Rep. 1990. 38; 877-80.

84. CDC, Centers for Disease Control and Prevention. Salmonella Dublin and raw milk
consumption, California. Morb Mortal Wkly Rep. 1984. 33; 196-8.

62