.4'
H. 3.519
. a.. y

. .

Wank.“ mm
M 9

(2|:- 7 a

. an. .
. i

63M» w:

 

. \ ‘ iluflk
I 3.01.!!!»
.1 .‘shaa‘vx
A.)

1‘4“!
. Agatha;
.a. u2§~. :1

1....

 

This is to certify that the
thesis entitled

EVALUATION OF A SURFACE PLASMON RESONANCE BIOSENSOR
FOR THE IDENTIFICATION OF SALMONELLA
TYPHIMURIUM AND ESCHERICHIA COLI 0157:H7

IN THE PORK PRODUCTION CHAIN
presented by

Cynthia Anne Meeusen

has been accepted towards fulfillment
‘ of the requirements for

Masters degree in Agricultural Engineering

Major p%fessor

Date 12/8/00

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY

Michigan State
University

 

 

 

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

 

lo é3§L110332002 out: a. it) @304

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuopBS-sz

EVALUATION OF A SURFACE PLASMON RESONANCE BIOSENSOR FOR
THE IDENTIFICATION OF SALMONELLA TYPHIMURIUM AND ESCHERICHIA
COLI O157:H7 IN THE PORK PRODUCTION CHAIN
BY

Cynthia Anne Meeusen

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

MASTER OF SCIENCE

Department of Agricultural Engineering

2000

ABSTRACT

EVALUATION OF A SURFACE PLASMON RESONANCE BIOSENSOR FOR THE
IDENTIFICATION OF SALMONELLA TYPHIMURIUM AND ESCHERICHIA COLI O157:H7 IN
THE PORK PRODUCTION CHAIN

By
Cynthia Anne Meeusen

An off-the—shelf surface plasmon resonance (SPR) sensor was converted
to a biosensor to detect Salmonella enterica spp. Typhimurium and Escherichia
coli O157:H7 in contaminated samples. The selectivity and specificity Of the SPR
biosensor were assayed using pure and mixed cultures Of S. Typhimurium and E.
coli 01572H7. Field samples were Obtained from swine farms and representative
small and medium-sized pork production facilities and analyzed using the SPR
biosensor. In addition to the standard plate count, the VlDASTM immunosensor
was used to validate the results of the SPR biosensor.

The Optimum antibody concentration for use on the SPR biosensor was
300 ug/ml. The detection limit of the SPR biosensor for S. Typhimurium and E.
00” O157:H7 in pure culture was 107 CFU/ml. Inoculated samples reached this
limit Of detection after 5 1/2 hours Of enrichment. Concentrations of non-target
bacteria beyond 107 CFU/ml caused a decrease in the magnitude of the sensor
response to the presence of the target pathogen.

The SPR biosensor has demonstrated potential for rapid, versatile. and
accurate pathogen detection. With further studies and refinements, the SPR
biosensor shows promise as a complementary pathogen detection system for

food safety.“

Copyright by
Cynthia Anne Meeusen
2000

This thesis is dedicated to my family,
Alma, Wayne, Josephine, Sara, Andrew, and Allie Meeusen
for their love and belief in me,
and to
Dr. Evangelyn Alocilja,
for her unswerving enthusiasm and support.
Thank you!

ACKNOWLEDGMENTS

Many Thanks to

Dr. Evangelyn Alocilja, for serving not only as my major advisor, but also
my mentor, my councilor, my teacher, my advocate, and my friend.

Or. Wes Osburn for kindly allowing us the use Of his lab-space, and for his
support while on my committee.

Dr. Elliot Ryser for serving on my committee, as well as the generous use
of his laboratory and resources.

Dr. Rajesh Sharma, for his helpful advice and cooperation.
Dr. Jamie Prater, for her assistance with all my microbiology questions.

Steve Marquie, for the constant encouragement and belief that I could
make it.

TABLE OF CONTENTS

List of Tables ................................................................................... viii
List of Figures......... x

introduction 1

Chapter 1

Review of the Literature 5
The Pork industry .................................................................... 5
Escherichia coli 0157: H7 ........................................................ 7
Salmonella Typhimurium ......................................................... 16
Food Safety and Current Control Methods ............................. 23
Current Detection Methods of Microorganisms ...................... 25
Biosensors .............................................................................. 32
Surface Plasmon Resonance ................................................. 35
References ............................................................................. 45

Chapter 2

Methods and Materials 52

Chapter 3

Sensitivity and Specificity Assays for Salmonella Typhimurium and
Escherichia coli 01572H7 Utilizing a Surface Plasmon Resonance
BIosensor 69

Abstract .................................................................................. 69

Introduction ............................................................................ 71

Methods and Materials ........................................................... 73

Results and Discussion .......................................................... 80

Conclusions ............................................................................ 100

References ............................................................................. 1 03
Chapter 4

Detection Of Salmonella Typhimurium and Escherichia coli O157:H7
in the Pork Production Chain using a Surface Plasmon Resonance
Biosensor and a VIDAST'" lmmunosensor... 105

Abstract .................................................................................. 105
introduction ............................................................................ 1 06
Methods and Materials ........................................................... 109
Results and Discussion .......................................................... 116
Conclusions ............................................................................ 124
References ............................................................................. 1 26

Recommendations for Further Research... ......128

vi

Appendices

Appendix A: Research Procedures and Protocols... 131
Appendix B: Data for Chapter 3 139
Appendix C: Data for Chapter4 151

Bibliography......... .....158

vii

LIST OF TABLES

Table 1.1. Leading causes of foodbome illness in the United States ......... 23
Table 1.2. Benefits and features Of biosensors for use in the food industry.34
Table 2.1. Typical preparation Of the SPR biosensor .................................. 53

Table 2.2. A typical dilution series experiment. The dilution series consists
Of SPR analysis Of 10 varying bacteria concentrations (CF U/ml). 60

Table 2.3. Environmental samples from slaughterhouse ......... _ ................... 62
Table 2.4. Farm samples ............................................................................ 62
Table 2.5. SPR biosensor procedure for Salmonella spp. assays... 65
Table 2.6. SPR biosensor E. coli O157:H7 assay procedure... 66

Table 3.1. A typical dilution series experiment. The dilution series consists
Of SPR analysis of 10 varying bacteria concentrations (CFUlml)... 78

Table 3 2. ANOVA results for S. Typhimurium specificity In mixed cultures
P-values... 99

Table 3.3. ANOVA results for E. coli 0157: H7 specificity in mixed cultures.

P-values... 99
Table4.1. Farmsamples 111
Table 4.2. Environmental samples from slaughterhouse... 111
Table 4.3. SPR biosensor procedure for Salmonella spp. assays... 113
Table 4.4. SPR biosensor E. coli O157:H7 assay procedure... . 1 14
Table4.5. SamplesinoculatedwithE.coliO157:H7........................... 115

Table 4.6. Incidence of Salmonella spp. and E. Coli O157:H7 in the pork
production chain - Results for the VIDASTM immunosensor... 117

Table 4.7. P-Values Of positive Salmonella spp. samples in VIDAS-
enriched method, after application Of first and second antibody,
compared to negative samples... 119

Table 4.8. P-Values Of positive Salmonella spp. samples in T88,
after application of first and second antibody, compared to
negativesamples............................................................... 119

viii

Table 4.9. P-Values of known-positive E. coli O157:H7 samples in
VIDAS—enriched method, after application of first and second
antibody, compared to negative samples... 121

Table 4.10. P-Values of known-positive E. coli O157:H7 samples in
T88, after application of first and second antibody, compared .
to negative samples... 121

Table 4.11. Average SPR biosensor response for positive and negative
samples, comparing the difference Of the two with the average
standarddeviation............................................................... 122

LIST OF FIGURES

Figure1.1. PorkProcessing Flowchart 7

Figure 1 ..2 (A) Direct method and (B) indirect method of fluorescent
labeling... .. 29

Figure1.3: Total internal reflectance................................................ 37

Figure 1.4: Effect Of the critical angle (Go) on reflection and refraction.
(a) incident light angle less than 9c, (b) Incident light angle greater
than 9c, causing total intemai reflectance... ...39

Figure 1.5. The Kretschmann geometry, with angle Of incidence 9. When
6 = 93p, virtually no light is reflected to the photodetector... ...43

Figure 2.1. Side view of the Spreetatm SPR biosensor 53

Figure 2.2. A typical SPR Biosensor Preparation. ( 1) Sensor initially
in NaOH/Triton solution, (2) Phosphate Buffered Saline (PBS)
solution, (3) Neutravidin solution, (4) PBS solution, (5) Biotinylated
antibody, (6) PBS solution, (7) NaOH/Triton in PBS, (8) BSA,
(9)NaOH/Tritonin PBS... 55

Figure 2.3. (A) Antibody-antigen interaction on the SPR biosensor
gold—film surface using one antibody layer a. (B) Antibody-antigen
interaction on the SPR biosensor gold-film surface in a
sandwich-type interaction, with two layers of antibody, a and b ..... 65

Figure 3.1. Side view Of the Spreetam‘ SPR biosensor 73

Figure 3.2. Average response of the SPR biosensor to anti-Salmonella spp.
antibody concentration of 3 pg, 30 I19, and 300 ug/ml... 81

Figure 3.3. Average response of the SPR biosensor to anti-E. coli O157:H7
antibody concentration Of 3 pg, 30 ug, and 300 ug/ml 83

Figure 3. 4. Average background responses of SPR biosensor to serial
dilutions Of S Typhimurium, E. coli 0157: H7, and sterile TSB
without antibody. .. .. .. 84

Figure 3.5. Average negative control of SPR biosensor - Response to
serial dilutions of sterile TSB using anti-Salmonella spp. and

anti-E. coli O157:H7 antibodies... ...86
Figure 3.6. Average SPR biosensor response for S. Typhimurium ......... 87
Figure 3.7. Average SPR biosensor response for E. coli O157:H7 ......... 88

Figure 3.8. Linear region Of S. Typhimurium dilution series... .. 89

Figure 3.9. Linear region of E. coli O157:H7 dilution series... 90
Figure 3.10. Timed growth of S. Typhimurium and E. coli O157:H7 ......... 93
Figure 3.11. Average specificity for S. Typhimurium... 94
Figure 3.12. Average specificity for E. coIiO157zH7... ...95

Figure 3 13. Specificity for anti- Salmonella spp.- prepared SPR biosensor
in mixed culture... . . ..97

Figure 3. 14. Specificity for anti-E. coli 0157: H7- prepared SPR biosensor
in mixed culture... .. . . . .............98

Figure 4.1. (A) Antibody-antigen interaction on the SPR biosensor gold-film
surface using one antibody layer a. (B) Antibody-antigen interaction '
on the SPR biosensor gold-film surface in a sandwich-type interaction,
with two layers Of antibody, a and b... 113

xi

INTRODUCTION

Contamination Of meat products by foodbome pathogens is a major food
safety concern. . Foodborne bacterial pathogens are believed to be the most
frequently occurring hazard of the nation’s food supply. It has been estimated ‘
that one in ten persons in the United States experiences bacteria related food
poisoning each year, with billions of dollars lost due to medical costs, lost
productivity, and product recalls associated with outbreaks Of food-borne illness
(Hui et al., 1994). in the United States, food-borne diseases cause an estimated
76 million illnesses, 325. thousand hospitalizations, and 5 thousand deaths
annually (Buzby et al., 1996). Researchers at the Economic Research Service
(ERS) of the United States Department of Agriculture (USDA) estimate the
annual cost of human illnesses for just seven food-borne pathogens from all food

sources to be $56-$94 billion (Buzby ef al., 1996). Meat and poultry sources
I account for $45-75 billion of this total cost (Buzby et al., 1996).

Pork products are growing in popularity among consumers Of meat, and
cOntinue to expand their share of local and global markets. In 1996, the US.
pork industry exported 413 thousand metric tons of pork valued at $1.1 billion
(US. Meat Export Federation, 1997). Every time a pathogenic outbreak occurs
in a pork product, consumer health is put at risk, and the viability of the pork

industry is threatened.

Escherichia coli O157:H7 has emerged as an important enteric pathogen
Of considerable public health significance. illnesses caused by E. coli O157:H7
can range from mild, watery diarrhea to life-threatening conditions such as
hemoiytic uremic syndrome and thrombotic thrombocytopenic purpura. The
combination of the severe consequences of infection, its low infectious dose,
and its association with many common foods makes E. coli O157:H7 a bacterial
pathogen of particular concern. E. coli 0157:H7 has been‘ identified in many
commonly consumed foods, including pork products (Jay, 2000).

Salmonella enterica has been identified as the most prevalent and costly
Of known foodbome pathogens (Davies, 1997). it is the main cause of
documented foodbome illness in most developed countries (Buzby et al., 1996).
Adding to the concern about Salmonella infection is the existence Of several
antibiotic resistant strains, especially the multi-drug resistant strain 8.
Typhimurium DT 104. With the emergence of this strain, the incidence and
severity of Salmonella-related human illness is increasing. Salmonella is of
particular concern in the pork industry. Currently 11% Of all outbreaks of
salmoneliosis in humans have been associated with pork (Isaacson et al., 2000).
Adding to the problem of Salmonella infection in pigs is their ability to become
long-term carriers Of the organism.

In an effort to reduce the occurrence and numbers Of pathogens on meat
and poultry products, recent food safety initiatives have emphasized the
monitoring and control of foodborne pathogens. Contributing to the prevalence

of foodborne disease is the fact that these foodbome pathogens do not

necessarily make the food look, smell, or taste bad. There is no definitive way to
tell simply by looking at a food whether or not it is contaminated (Moutvilie,
1987). Because Of this, tests must be performed that are accurate and sensitive
to very low levels Of microbial contamination. Classical approaches to quality
control have relied heavily on microbiological determinations of both raw
materials and end products, but the time required for results is too long for many
products (Jay, 2000). Fot these reasons, there is a growing need in the food
industry for pathogen detection systems that are sensitive to low levels of
bacteria, specific to the target organism, inexpensive, and capable of yielding
results at or near real-time.

Surface Plasmon Resonance (SPR) is an Optical phenomenon that occurs
as a result of total internal reflection of light at a metal film-liquid interface.
Although the light is totally reflected, a component Of the incident light
momentum, the evanescent wave, penetrates into the less dense liquid medium.
In a SPR biosensor, the evanescent wave interacts with'surface plasmons (free
oscillating electrons) in the thin metal film surface. When SPR occurs, energy
from the incident light is lost to the metal film, resulting in a decrease in light
intensity. The resonance phenomenon occurs only at a precisely defined angle
Of incidence, which is dependent on the refractive index Of the medium adjacent
to the metal surface. The refractive index changes in direct proportion to the
mass and the make-up of the medium present. When antibodies are affixed to
the metal surface, the angle of incidence that causes SPR depends on the

amount of antibody-antigen substrates present. By using antibodies specific to

pathogens Of interest, it is possible to utilize the SPR phenomenon to measure
the amount of pathogenic bacteria preSent in a sample by measuring the change
in refractive index.

The goal of this research is to assess the feasibility Of using a SPR
biosensor to detect Salmonella Typhimurium and E. coli O157:H7 in
environmental and meat samples from representative pork packing facilities.
The objectives Of this investigation include determining the sensitivity and
specificity of the SPR biosensor assay, and comparing the use of the SPR
biosensor as a diagnostic tool to the VIDAST“ immunoassay for detecting
Salmonella spp. and E. coli O157:H7 in samples obtained during pork

production.

CHAPTER 1

REVIEW OF THE LITERATURE
l. The Pork Industry

Foodborne bacterial pathogens are believed to be the most frequently
occurring hazard in the nation’s food supply. It has been estimated that one in
ten persons in the United States experiences a bacterial foodbome illness each
year (Hui et al., 1994). More than 200 known diseases are transmitted through
food, causing an estimated 76 million illnesses, 325,000 hospitalizations, and
5,000 deaths annually (Mead et al., 1999). Of the confirmed cases Of human
foodborne illness and deaths reported to the Centers for Disease Control and
Prevention (CDC), over 90% are attributed tO bacteria (Buzby et al., 1996).
Worldwide, diarrheal illnesses are the leading cause of childhood death and the.
second leading cause of death in general, behind cardiovascular disease (Morse
et al., 1994). One estimate Of the annual U.S. cost for human illness from all
food sources is $56-$94 billion. Meat and poultry sources account for $45-75
billion of this total cost (Buzby et al., 1996).

Pork products are growing in popularity among consumers of meat, and
continue to expand their share of local and global markets. In 1996,. the US.
pork industry exported 413 thousand metric tons Of pork valued at $1.1 billion
(US. Meat Export Federation, 1997). Every time a an outbreak involves a pork
product, consumer health is put at risk, and the viability of the pork industry is

threatened.

It is generally agreed that the intemai tissues Of healthy slaughter animals
are free of bacteria at the time of slaughter (Jay, 2000). However, husbandry
conditions during transport to slaughter may favor the growth or shedding of a
particular pathogen, leading toithe pathogen becoming an important component
in the feces of pigs (Letellier et al., 1999). During evisceration, fecal bacteria .
may contaminate the meat, and eventually cause diseases in humans. Other
primary sources and routes of microorganisms tO fresh meats include the stick
knife, animal skin or hide, the gastrointestinal tract, the hands of handlers,
containers, the handling and storage environment, and/or lymph nodes (Jay,

2000).

When many animals are slaughtered in a single time period, there is a
tendency for any bacteria present in the animals to be spread among the
carcasses (Jay, 2000). The pork processing chain is a multi-staged process,
(Figure 1.1), with the potential for contamination by pathogenic bacteria at each
stage.

Pork products have been implicated in a number of outbreaks of
foOdborne illness. Fecal shedding and growth Of pathogens in the farm
environment can lead to subsequent contamination Of the slaughterhouse
environment and the finished food-product. Most outbreaks are associated with

intensively reared weaned pigs (Ryser, 2000).

 

Hogs from Farm

I

Trucking
I
Pen Area
Stunning and Sticking

I

Scalding water

I

Dehairing / Singeing

I

Dehairing / Cleaning

I

Evisceration

J.

Head Removal

I

Final Wash

l

Chill Room
(18-24 h chilling)

I

Meat
processing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.1. Pork processing flowchart (Ryser, 2000).

Recent outbreaks of foodborne illness caused by Salmonella have
caused pork to be more closely analyzed as a source of bacterial pathogens
(lsaacson et al., 1999). Currently 11% of all outbreaks of salmonellosis in
humans have been associated with pork (lsaacson et al., 2000). in one study of
pork and beef carcasses, salmonellae were recovered from 27% of 49 pork
samples (Jay, 2000). One hundred twenty-one Salmonella strains were isolated
from 225 pork carcasses in northeast Georgia (Epling and Carpenter, 1990).

Adding to the problem of Salmonella infection in pigs is their ability to
become long-term carriers of the organism. Shedding of Salmonella by these
animals may contaminate the environment, instruments, equipment, and meat at
the slaughterhouse (Letellier et al., 1999). Carrier pigs do not have any clinical
signs of infection, which makes them difficult to identify (lsaacson, 2000).

Escherichia coli 0157:H7 is a slightly less well-known pathogen in pork
products, but because of its potential virulence, it is still important. Enteric
bacteria are common meat contaminants, due to their presence in the
gastrointestinal tract. Of 442 meat samples examined, 86% yielded enteric
bacteria. The mean number for 94 pork sausage samples tested was 7.9
CFU/gram (Jay, 2000). E. coli was one Of the most commonly found, with 29%
Of meat samples testing positive (Jay, 2000).

Because Of growing consumer demand for pork and pork products, more
research is focussing on the pork industry chain as a potential source of human
foodborne illness. As a whole, foodbome pathogens are a major food safety

concern. Any pathogenic outbreak associated with a pork product puts

consumer health at risk, and threatens the vitality Of the pork industry.
improvements in monitoring the pork industry chain for pathogenic bacteria
would save time and money, and most importantly, lower public health risks for
the consumer.

ll. Escherichia coli O157:H7

E. coli O157:H7 is a facultative, non-sporing rod-shaped facultative
member of the Enterobacten’aceae family, which has emerged as an important
enteric pathogen of considerable public health significance in the United States,
Canada, and Europe. it has caused many outbreaks and numerous spOradic
cases of hemorrhagic colitis, hemolytiC-uremic syndrome, and diarrheal illness
occurring in day-care centers, schools, nursing homes, and the community
(Ratnam et al., 1988). Raw or under-cooked hamburger and beef, raw milk,
unpasteurized apple cider, contaminated water, mayonnaise, and unwashed
vegetables have all caused outbreaks (Monk et al., 1995; Jay, 2000). The
Center for Disease Control and Prevention (CDC) estimates that there may be
20,000 illnesses a year due to E. coli O157:H7 infection (Jay, 2000).

E. coli in its non-pathogenic forms is common in warm-blooded animals
(Buzby et al., 1996). A German pediatrician, Dr. Theodore Escherich, first
described whatwas then called Bacterium coli commune in 1885 (Neill et al,
1994). Use of the suffix commune recognized that this organism is distributed
widely in the intestinal flora of many animals, and is the predominant facultative

anaerobe in the human bowel, helping to maintain normal physiologic function Of

the intestine (Neill et al., 1994; Varnam and Evans, 1991). As a human
pathogen, the organism, now called Escherichia coli, was recognized as a cause
of infant diarrhea as early as the 17005 (Jay, 2000). E. coli was established as a
foodbome pathogen in 1971 when imported cheeses contaminated with an
enteroinvasive strain turned up in 14 American states, causing illness in nearly
400 individuals (Jay, 2000).

Five virulence groups of E. coli are recognized: enteroaggregative
(EAggEC), enterotoxigenic (ETEC), enterOinvasive (EIEC), enteropathogenic
(EPEC), and enterohemorrhagic (EHEC) (Jay, 2000; Neill et al., 1994; Varnam
and Evans, 1991). The five main strains each have a distinct clinical pattern,
and differences in epidemiology, pathogenicity, and O:H antigens.
Enteroaggregative (EAggEC) E. coli

Enteroaggregative E. coli, also designated enteroadherant, is composed
of strains that have been epidemiologically linked to diarrhea disease, but lack
properties typical of the other four groups (Neill et al., 1994). EAggEC strains
are related to EPEC but display unique aggregative adherence (Jay, 2000).
Some EAggEC strains produce heat-stable enterotoxin (ST), designated EAST1.
The distinguishing clinical feature Of EAggEC strains is a persistent diarrhea that
lasts more than 14 days, especially in children. It is unclear whether members of
this group are foodbome pathogens (Jay, 2000).

Enterotoxigenic (E TEC) E. coli
Enterotoxigenic E. coli strains produce ST enterotoxin and/or heat-labile

(LT) enterotoxin (Neill et al., 1994). ETEC strains colonize the small intestine by

10

means Of fimbrial colonization factor antigens (CFAs). Once attached, the ETEC
strains produce enterotoxins. Unlike EPEC strains, which produce diarrhea
primarily in the very young, ETEC strains cause diarrhea in both children and
adults (Jay, 2000). ETEC infeCtion resembles cholera with a severe dehydrating
diarrhea and high mortality, most commonly in children (Neill et al., 1994). ‘
Approximately 108-101o colony forming units (CFU) are necessary for diarrhea by
an ETEC strain in an adult human (Jay, 2000). In areas Of poor sanitation,
transmission is Often caused by fecal shedding of ETEC strains from
asymtomatic adults (Neill et al., 1994).
Enteroinvasive (EIEC) E. coli

Enteroinvasive E. coli strains derive their virulence from their ability to
invade epithelial cells (Neill et al., 1994). They do not generally produce
enterotoxins as do ETEC strains, but they enter and multiply in colonic epithelial
cells and then spread to adjacent cells in a manner similar to the shigellae (Jay,
2000). Members Of this group are most Often found in the colon, and cause
voluminous bloody or non-bloody diarrhea. Although foods are a proven source
Of. this strain, person-tO-person transmission is known (Jay, 2000; Neill et al,
1994).
Enteropathogenic (EPEC) E. coli

Enteropathogenic E. coli strains have been epidemiologically linked to
diarrheal illness, produce neither the LT or ST enterotoxins, and are non-
invasive (Neill et al., 1994; Jay, 2000). These strains possess adherence factor

plasmids that enable adherence to the intestinal mucosa, and the development

11

Of attachment-effacement (AlE) lesions in which surrounding microvilli are
destroyed (Neill et al., 1994). The AIE lesions in the intestine appear to be the
most important virulence factor Of EPEC strains, which do not produce
detectable quantities of Shiga toxin (Jay, 2000). EPEC strains cause diarrhea in
children generally under one year old. Severe disease resulting in prolonged
diarrhea or death is rare (Neill et al., 1994).
Enterohemorrhagic (EHEC) E. coli

Enterohemorrhagic E. coli are similar to EPEC strains in that they both
possess the chromosomal gene eae (E. coli attaching and effacing gene),
resulting in the production Of attachment-effacement lesions that disrupt
microvilli on the intestinal wall Of the host (Kudva et al., 1997). in contrast to
EPEC, however, EHEC strains affect only the large intestine. All EHEC strains
produce Shiga toxin 1 (Stx1) and/or Shiga toxin 2 (Stx2), also referred to as
verotoxin 1 (VT1) and verotoxin 2 (VT2), because of their toxicity to African
green monkey kidney tissue cells (Vero cells) (Jay, 2000; Buchanan and Doyle,
1997). EHEC strains produce a clinical illness of severe bloody diarrhea
(‘hemorrhagic colitis’) and possess certain virulence determinants such as SLT
production, possession Of a 60-Mda plasmid, and eliciting attaching —and-
effacing histopathologic lesions in an animal mode. Serotype O157:H7 is
considered the most well characterized, ‘prototypic’ EHEC strain (Neill et al.,

1994)

12

Escherichia coli O157:H7

The H7 antigen type was initially isolated in 1944 from a human diarrheal
specimen, whereas the 0157 type was first isolated and named from diarrheal
swine feces (Orskov et al., 1977). The first 0157:H7 strain was recovered in
1975 from a woman with bloody diarrhea (Jay, 2000; Neill et al., 1994). E. coli
0157:H7 was next recovered in 1978 from diarrheal stools in Canada (Jay,
2000). E. coli O157:H7’s recognition as a human pathogen was finalized in
1982, following two outbreaks of severe bloody diarrhea in Oregon and
Michigan, transmitted through consumption Of fast-food hamburgers (Neill et al,
1994; Ratnam eta/., 1988; Buchanan and Doyle, 1997).

Following a Pacific Northwest outbreak in 1993, the United States
Department Of Agriculture Food Safety and inspection Service (USDA-FSIS)
undertook a multistate study Of the prevalence and incidence Of E. coli 0157:H7
in both beef and dairy herds. The largest number per gram found was 15, and
the average was approximately 4 CFUIg Of fresh beef. in a nationwide survey
conducted between 1994 and 1998, the USDA found E. coli O157:H7 in 23 of
23,900 ground beef samples (Jay, 2000). 00er and Schoeni (1987) found E.
coli O157:H7 in 3.7% of 164 beef, 1.5% Of 264 pork, 1.5% of 263 poultry, and
2.0% Of 205 lamb samples.

One distinguishing feature Of serotype 0157:H7 is its lack of rapid (less
than 48 hrs) fermentation Of sorbitol. The most common screening procedure for
this organism is based on this feature, because 93% Of E. coli isolated from

humans do ferment sorbitol (Neill et al., 1994).

13

Toxins

The toxins produced by EHEC strains of E. coli (verotoxin, verocytotoxin)
have been referred to as Shiga-like toxins. Shiga toxin is a potent toxin that is
produced by Shigel/a dysenteriae. The terms ‘Shiga-like toxin’ and ‘Verotoxin’
are considered synonymous, although specific toxins (e.g. SLT-ll and VT-2) may
not be identical (Neill et al., 1994).

Epidemiology

An estimated 2.8 cases of E. coli O157:H7'related illness per 100,000
people were reported in 1998 (USDA-FSIS, 1998). Human infections with E. coli
O157:H7 are usually linked to consumption Of contaminated and improperly
cooked beef, unpasteurized milk, or fresh fruits and vegetables, water, or apple
cider (Besser et al., 1993; Griffin and Tauxe, 1991; Hofmann, 1993).

People Of all ages, from 1 to 80 years Old, have developed E. coli
O157:H7 infection (Neill et al., 1994). The mean incubation period for this
serotype is generally 12 days after eating contaminated foods (Buchanan and
Doyle, 1997), although longer periods of 3.1 to 8 days have been reported (Neill
et al., 1994). illness typically begins with a short prodrome Of mild non-bloody
diarrhea that progresses within a day to grossly bloody diarrhea, Often
associated with severe abdominal pain and cramping (Neill et al., 1994). The
period Of overtly bloody diarrhea accompanied by severe abdominal pain and
moderate dehydration typically lasts 4-10 days (Buchanan and Doyle, 1997).
Because Of these symptoms, cases frequently have been misdiagnosed as

having a noninfectious cause such as mesenteric infarction, inflammatory bowel

14

disease, Meckel’s diverticulum, intussusception, and anatomic gastrointestinal
bleeding (Neill et al. 1994).

E. coli O157:H7 can be carried asymptotically, like many other enteric
pathogens such as Salmonella, Shigella, and Campy/Obacter (Neill et al., 1994).
The most common method of transmission is the fecal-oral route, and illness can
result from very low levels Of bacteria (Buchanan and Doyle, 1997).

Several complications may arise in patients with hemolytic colitis, Of which
hemolytic uremic syndrome is the most common. Other potential complications
include thrombotic thrombocytopenic purpura, and acute renal failure (Buchanan
and Doyle, 1997).

Hemolytic Uremic Syndrome

E. coli O157:H7 has been blamed for 85-95% of all hemolytic uremic
syndrome (HUS) cases (Griffin, 1995). First described in 1955 (Neill et al.,
1994), HUS consists Of an acquired Coombs’ negative hemolytic anemia,
thrombocytopenia, and acute renal failure. The onset of HUS is typically 7 days
after the onset of gastrointestinal symptoms (Buchanan and Doyle, 1997).
Approximately 10% of patients with hemorrhagic colitis develop HUS (Simmons,
1997). This proportion varies among reported series and outbreak
investigations, most likely due to the presence of various risk factors, such as
extremes in age, antimicrobial use, antimotility agents, and toxin type of the
infective strain (Neill et al., 1994).

Symptoms of HUS include pallor, intravascuiar destruction of red blood

cells (microangiopathic hemolytic anemia), thrombocytopenia, lack of urine

15

formation (OligO-anuria), edema, and acute renal failure (Buchanan and Doyle,
1997). HUS is a highly morbid disease, with a fatality rate of 3-5%, and a
considerable risk of long-term morbidity such as hypertension, chronic renal
failure, and disability (Neill et al., 1994; Buchanan and Doyle, 1997).
Thrombotic ThrOmbocytopenic purpura

A second complication associated with E. coli O157:H7 infection is
thrombotic thrombocytopenic purpura (TTP). TTP consists of the cardinal
features of HUS (hemolytic anemia, thrombocytopenia, and renal failure) along
with fever and altered mental status (Neill et al., 1994). TTP, however, generally
causes less renal damage, and is restricted primarily to adults (Buchanan and

Doyle, 1997).

iii. Salmonella enterica serotype Typhimurium

The salmonellae are gram-negative, non-sporing, motile facultative
members Of the Enterobacteriaceae family (Jay, 2000). The USDA has identified
Salmonella enterica as the most prevalent and costly of the known foodbome
pathogens (Davies, 1997). Salmonella is a main cause of documented
fOOdborne human illnesses in most developed countries (Buzby at al., 1996).
Each year'in the United States an estimated 8-18 thousand hospitalizations.
2,400 cases of septicemia, and 500 deaths are associated with Salmonella
infections (USDA, 1999). There are more than 2 thousand serotypes (or
serovars) of the genus, Of which 50-150 have been linked to disease outbreaks

(Ziprin, 1994). The term ‘serotype’ refers to a group of related microorganisms

16

distinguished by its composition of antigens (Buzby et al., 1996). Of the
Salmonella serotypes, salmonellae have been placed in two species, S. enterica
and S. Bongori (Jay, 2000). The serotypes are divided into five sub-species,
most of which are classified under the S. enterica type species (Jay, 2000;
Buzby et al., 1996). The strains are classified by their serotype designation. For
example, the serotype typhimun‘um is found among the S. enterica species, but
is referred to as Salmonella enterica serotype Typhimurium,'or S. Typhimurium
(Jay, 2000; Buzby et al, 1996).

Salmonella was first cultured by Gaffky in 1884 (Jay, 2000). The same
year, T. Smith and D. E. Salmon first described it in the environment, and the
genus was subsequently named in Salmon’s honor (Ziprin, 1994; WVA, 1997).
Until 1949, S. Typhi was the predominant strain found to cause typhoid fever in
humans (Tauxe, 1991 ). As typhoid fever was nearly eliminated due to advances
in medicine and food safety, the non-typhoid strains, including S. Typhimurium,
became leading sources of Salmonella infection in humans (Tauxe, 1991). The
major Salmonella serotypes causing salmonellosis and human gastrointestinal
disease are S. enterica spp. Typhimurium, Enteritis, and Typhi (Lin and Tsen,
1999, WVA, 1997). A survey of Salmonella isolates obtained from patients in
Taiwan revealed that S. enterica spp. Typhimurium is the most common serotype
causing foodbome disease in humans (Wang et al., 1994). Due to the
resistance of S. Typhimurium strain DT 104 to a range of antibiotics, it is a

pathogen of increasing concern (WHO, 1997).

17

The salmonellae can be placed into three epidemiological groups: those
that infect humans only, the host-adapted serotypes, and unadapted (no host
preference) serotypes (Jay, 2000; WHO, 1997).

Salmonella infecting humans only

Salmonellae that cause enteric fever only in humans and higher primates
include S. Typhi, S. Paratyphi A, and S. Paratyphi 0 (Jay, 2000; WHO, 1997).
This group includes the agents Of the most severe of all diseases caused by
salmonellae; typhoid and paratyphoid fevers (Jay, 2000).

Host-adapted Salmonellae serotypes

The host-adapted salmonellae serotypes include S. Gallinarum (poultry),
8. Dublin (cattle), S. Abortus-equi (horses), 8. Abortus-ovis (sheep), and S.
Choleraesuis (swine). As the name suggests, these are serotypes of Salmonella
that have been isolated from only one host-species. These serotypes are
infrequently human pathogens (Jay, 2000; WHO, 1997). However, when these
strains do cause disease in humans, it is Often invasive, and can be life-
threatening (WHO, 1997).

Unadapted Salmonellae serotypes

Unadapted Salmonella serotypes can be pathogenic for humans and
other animals, and include most foodbome serotypes, including S. Typhimurium
(Jay, 2000). These strains typically cause only gastroenteritis that is mild and
self-limiting, but can be severe in the young, the elderly, of those with weakened

immune systems (WHO, 1997).

18

Antibiotic-resistant S. Typhimurium strain DT 104

Adding to the virulence Of salmOneilae infection is the recent discovery of
several antibiotic-resistant strains, especially the multi-drug resistant S.
Typhimurium strain Definitive Type (DT) 104 (Foster, 1997; WHO, 1997; WVA,
1997). With the emergence of this strain, the incidence and severity of
Salmonella-related human illness is increasing, with some countries seeing a
20-fold increase in incidence in the past 10-15 years (WHO, 1997). Today 8.
Typhimurium strain DT 104 is a leading cause Of ‘Salmonellosis in the United
States (Tauxe, 1991; lFST, 1997).

Strain DT 104 is a subpopulation of S. Typhimurium that reacts in a
specific way when tested against a battery Of bacteriophages. An additional
characteristic of S. Typhimurium strain DT 104 is that isolates, known as
resistance type (R-type) ACSSuT, are commonly resistant to several antibiotics,
including ampicillin, chloramphenicol, streptomycin, sulfonamides, and
tetracycline (Foster, 1997; WVA, 1997; Karpiskoval et al., 1999).

S. Typhimurium DT 104 was first identified in 1970, and was sensitive to
antimicrobial agents. Muiti-drug resistant strains of S. Typhimurium DT 104
emerged in cattle in 1988 in England and Wales (WHO, 1997; Foster, 1997). it
has since been isolated from sheep, pigs, horses, goats, emus, cats, dogs, elk,
mice, coyotes, ground squirrels, raccoons, chipmunks, and birds (Jay, 2000). S.
Typhimurium DT 104 infection in humans has been associated with the
consumption of beef, chicken, lamb, pork, sausage, raw milk, and meat paste as

well as with the handling of sick. animals (Foster, 1997; lFST, 1997).

19

This drug resistant strain is Of particular concern, because antimicrobial
therapy is used extensively to control S. Typhimurium infection in animals
(WHO, 1997). The emergence of an antibiotic-resistant strain has made
infections with S. Typhimurium in food animals difficult to control. Over the past
three decades, many countries have reported sharp rises in salmonellosis
(WVA, 1997). Between 1990 and 1995, the number of S. Typhimurium DT 104
isolates from humans in Britain increased from 259 to 3,837 per year—a 15-fold
increase. The percentage of drug-resistant isolates increased from 39% in 1990
tO 97% in 1995 (Foster, 1997).

Serotypes that are resistant to antimicrobial agents seem to depend more
heavily on characteristics of the host (e.g. extremes in age, and/or strength'of
the immune system) than do those serotypes that are sensitive to antimicrobial
agents (Buzby et al., 1996). Infection with multi-drug resistant S. Typhimurium
DT 104 has been associated with hospitalization rates twice that Of other
zoonotic foodbome Salmonella infections, and with ten-fold higher case-fatality
rates (WVA, 1997). More than one-third Of patients have required
hospitalization, and 3% have died (Foster, 1997).

Epidemiology

Each year in the United States, 840 thousand to 4 million people become
sick, and up to 4 thousand people die as a result of infection with S. enterica
(Buzby and Roberts, 1996; Tauxe, 1991). S. enterica is one of the most costly of
foodborne pathogens, with estimated medical costs associated with treatment of

salmonellosis ranging from $0.69 to $3.8 billion per year (lsaacson et al., 1999).

20

Salmonellae are widespread in animals, especially in poultry and swine, and the
animal frequently shows no symptoms Of the disease. Eggs, poultry, meat, and
meat products are the most common vehicles for foodbome salmonellosis (Jay,
2000), however environmental sources, including water, soil, insects, factory
surfaces, kitchen surfaces, animal feces, fresh produce, raw meats, raw poultry, .
and raw sea foods have also been implicatediin human illness (00er et al,
1997; Jay, 2000). Additionally, thousands of cases Of human salmonellosis in
the United States and other industrialized countries have been transmitted by ice
cream, chocolate, potato salad, cheddar cheese, raw milk, black pepper, paté,
aspic, ham, pasteurized milk, and drinking water (Foster, 1997).
Characteristics of Disease

Normally, S. enterica-caused human illness is limited an acute
gastroenteritis (Buzby et al., 1996). Humans vary in their susceptibility to
Salmonella, depending on the virulence Of the serotype, the individual’s immune
system, and the quantity of Salmonella ingested (Buzby et al., 1996). Typically.
107-109 CFU/g are necessary for salmonellosis (Jay, 2000). However, the
infectious dose may be as low as one cell for some Salmonella serotypes
(CAST, 1994; Buzby et al., 1996). individuals most susceptible to Salmonella
infection include the very young, the elderly, and those with weakened immune
systems (Buzby at al., 1996).

Symptoms of salmonellosis typically develop in 12-36 hours after
ingestion (Jay, 2000; Buzby et al., 1996), and include dehydration, nausea.

fever, vomiting, abdominal pain, headache, chills, and diarrhea, usually

21

accompanied with prostration, muscular weakness, moderate fever,
restlessness, and drowsiness (Jay, 2000; WHO, 1997; Buzby et al., 1996). The
incidence is particularly high in children and the elderly; accounting for up to
60% of all laboratory confirmed cases (WHO, 1997). Symptoms persist for an
average of 2-3 days. The average mortality rate is 4.1%, higher in the very
young (under 1 year Old) and up to 15% in people over the age Of 50 (Jay,
2000). Up to 5% of patients may become carriers of the organism.

Salmonella infections can cause secondary—disease syndromes, and
sometimes chronic illnesses (Buzby et al., 1996). Many Salmonella serOtypes
can penetrate the intestinal lining in humans. infrequently, the organism may
invade the bloodstream, causing septicemia (Buzby et al., 1996). Complications
of septicemia include endocarditis, meningitis, and pneumonia. Although an
enterotoxin and a cytotoxin have been identified in pathogenic salmonellae, they
seem to play a minimal role in infection (Jay, 2000).

Control of salmonellae requires reducing infection in food animals and
lowering the risk of contamination at all stages in the food production chain
(WHO, 1997). It is very unlikely that the eradication of salmonellae in domestic
animals is possible in the foreseeable future (WHO, 1997). For this and other
reasons, a method of detection that is rapid, accurate, and reliable is needed in

the food industry to control and monitor the incidence Of these bacteria.

22

IV. Food Safety and Current Control Methods

Bacterial pathogens are a major growing concern in both the food industry
and the public eye. Microorganisms can enter the human food chain on the
farm, during transportation, processing, at retail food outlets, and during food
preparation both in homes and public venues (Table 1.1). Bacterial foodborne
illness can result when a significant amount of living bacterial pathogens are
ingested.

Table 1.1. Leading Causes of Foodborne Illness in the United States
(Bryan, 1990)

 

 

Factors 1961 -1 982
Improper Cooling 44%
Lapse Of 12 or more hours between preparation/eating 23%
Contamination by handlers 18%
Addition of uncooked raw ingredient 16%
Inadequate cooking/canning/heating 16%

 

 

Although it may not be possible, or even desirable, to achieve a zero
tolerance for all such organisms, the production of foods with the lowest possible
numbers is the desired goal. in an effort to reduce the occurrence and numbers
of pathogens on meat and poultry products, reduce the incidence of foodborne
illness associated with consuming these products, and provide a framework for
modernization of the meat and poultry inspection system, the USDA-FSIS
mandated new requirements on July 25, 1996. The new regulations required
that all slaughter facilities and plants producing raw ground products conduct
regular microbial testing to verify the adequacy of a plant’s process controls for

the prevention and removal Of fecal contamination and associated bacteria, and

23

that they develop and implement HACCP (Hazard Analysis and Critical Control
Points) programs (lDEXX, 1998).

HACCP is a system Of science-based process controls designed to
identify and prevent biological, chemical, and physical hazards. The system is
intended to lead to the production Of safe foods by minimizing the hazards in raw
materials. It is designed to prevent problems before they occur, and to correct
deviation as soon as they are detected (Anderson, 1994). The HACCP
approach is used extensively in the meat, poultry, and seafood industries to
produce products in compliance with health and safety requirements. - By using
a HACCP system, control is transferred from end product testing into the
processing and manufacturing of foods — that is, from testing for failure to
preventing it.

The HACCP system has seven principles:

1. Conduct a hazard analysis, identifying where significant hazards can

occur, and describe preventative measures.

2. Identify the critical control points (CCPs) in the process - points at

which controls can be put into place to reduce food safety hazards.

3. Establish critical limits for preventative measures associated with each

identified CCP.

4. Establish CCP monitoring. requirements. Establish procedures for

using the results of monitoring to adjust the process and maintain

control.

24

5. Establish corrective actions to be taken when monitoring indicates that
there is a deviation from an established critical limit.
6. Establish effective record-keeping procedures that document the
HACCP system.
7. Establish procedures for verification that the HACCP system is
working correctly.
The HACCP system is a proactive, systematic approach to controlling
foodbome hazards. Rather than the traditional approach Of end-product testing,
HACCP places emphasis on the safety Of all ingredients and all process steps,

on the premise that safe products will result if these are controlled.

V. Current Detection Methods Of Microorganisms

in addition to the new federal regulations for process control, consumers
are increasingly aware Of the risk to human health due to foodbome pathogens
and as such demand a safe, high-quality, and nutritious food supply. This
means that having a reliable system for monitoring the quality and safety Of
foods is of increasing importance, and fast, reliably sensitive screening methods
are needed by the food industry.

Although some methods of analysis are better than others, every method
has its associated inherent limitations. None permit the determination Of exact
numbers of microorganisms in foods (Jay, 2000). it is not feasible to test an
entire batch of food, as the sampling methods Often destroy or contaminate the

product, and take time during which the food is losing its freshness and

25

commercial value. Because Of the small sample size used for testing, food
analysis methOds must be sensitive to low numbers Of pathogens. They must be
specific, as a product recall based on faulty identification presents enormous
losses Of time and money. This factor is quite important, as there are Often
many similarities between pathogenic organisms and non-pathogenic strains. A
pathogen detection procedure must be able to differentiate between these
microbes. Additionally, it is very advantageous for the detection method to be
quick. In the world of food processing, the product is by nature unstable, and
must move through the food processing plant quickly in order to maintain
product quality. if a detection method takes too long, the product could have
already reached shelves. This can result in large-scale product recalls that are
enormously expensive, not only in monetary terms but also in the public name
recognition Of the company. Finally, it is advantageous for pathogen detection
methods to be relatively easy to perform. Many smaller food processing plants
cannot afford to have an full-time laboratory technician on-site. Therefore, if a
detection method could be easily taught to a non-microbiologist, it could
increase the quantity Of product tested while saving the company money.

Common current types of pathogen detection in food products include
conventional plating, molecular, and immunological methods. Each of these are
described briefly with their respective advantages and disadvantages.
Conventional Plating Methods

Conventional plating methods are by far the most widely used procedures

for determining the numbers of viable cells in a food product (Jay, 2000). These

26

.‘5

traditional approaches to microbial assays require enrichment and cultivation,
isolation, morphological examination, and biochemical testing to identify human
pathogens. These assays often require 4-6 days (Deshpande, 1994).

In the conventional Standard Plate Count (SPC) method, a portion Of the
food is homogenized, serially diluted, plated onto suitable agar media, and
incubated for a given time. Once colony growth has been established, visible
colonies are counted by hand or by use of an electronic counter. Variation of
the SPC method includes membrane filtration, microscope colony counts, use of
dry films such as Petrifilm, dye reduction, and contact plating (Jay, 2000).

An advantage of conventional plating methods for identification and
enumeration of pathogenic bacteria in food products is the accuracy and
reproducibility Of the results. Such methods have been refined and perfected
over many years, and microbiologists that perform them have typically had years
of laboratory experience performing them.

Drawbacks to conventional plating methods, however, are many: the cost
of equipment and consumables, slow sample throughput, the laborious nature Of
the techniques, and the level of experience and skills required for the analysis,
and probably most significantly the length Of time necessary to achieve results.
it is not always feasible to hold an entire ‘batch’ of processed food until the
microbial plating has been completed. Rather, random samples are taken and
tested, and the rest Of the batch allowed to continue through. If the samples

reveal the presence of pathogenic microorganisms, the product have most likely

27

already been packaged and shipped. This can lead to large product-recalls,

which are expensive both in terms Of money and the reputation Of the company.

Molecular Methods

Most methods of detecting and characterizing pathogenic bacteria based
on molecular or immunological methods were developed since 1960 (Jay, 2000). '
Molecular methods for detection of pathogenic bacteria are based on some
combination of the metabolic activity of microorganisms on specific substrates,
genetic typing, measurements of growth response, and/or measurements of
certain parts of cells (Jay, 2000). Molecular methods for detecting and
Characterizing foodborneipathogens include nucleic acid probes, polymerase
chain reaction (PCR), restriction enzyme analysis, random amplification of
polymorphic DNA (RAPD), pulsed field gel electrophoresis, restriction fragment
length polymorphism (RFLP), and ribotyping (Jay, 2000). These methods are all
relatively new teChnologies, and will most likely grow in importance as more
foodbome organisms are reclassified by nucleic acid analyses.

Molecular methods based on genetic typing of the pathogen Of interest
Often use a DNA probe to detect homologous DNA or RNA sequences, and/or
some combination Of restriction endonucleases. Advantages Of these methods
are their high specificity and selectivity — using the DNA fingerprint Of an
organism as a target is a good assurance of finding an exact match to your

target of interest.

28

Disadvantages Of the molecular methods are the time and expertise
required to perform the assays. Most of these methods include several steps,
many with extended wait-periods, and although DNA characterization techniques
have rapidly advanced in the past few years, they still requires a laboratory
technician with extensive training and expertise (Jay, 2000).

Immunological Methods

Immunological methods rely on antibody-antigen interactions to detect
pathogens in foods. Methods include the use of fluorescent antibodies .for
photodetection or fluorescence microscopy of the antibody-antigen complex.
The direct fluorescent antibody (FA) technique employs a specific antibody to

which is coupled the fluorescent compound (Figure 1.2).

. ,_ ., , A , Fluorescent
‘WWMbLabel
.. Fluorescent ’ ‘v * ’ -.— . . V . .
WM’Label . Anthody2.

‘ Antibody 1

. Antibody1 __ __
' " : Antigen

Antigen

     

(A) ' (8)

Figure 1.2. (A) Direct method and (B) indirect method Of fluorescent labeling

The indirect method uses a second antibody (Antibody 2) that is specific
to Antibody 1. The fluorescent label is coupled to Antibody 2, and when this
couples to Antibody 1, the antigen is indirectly fluorescently labeled. Although
more complicated, the use of the indirect method eliminates the need to pr'epare

FA for each organism of interest (Jay, 2000).

29

Radio-immunoassay is another technique in which a radioactive label is
added to an antigen, allowing it to react to its specific antibody, then measuring
the level of antigen-antibody complexes by use of a radioactivity counter (Jay,
2000) 1

Enzyme linked immunosorbent assay (ELISA) is another popular method
Of detection. it consists of an enzyme coupled to either an antigen or an
antibody. The enzyme remaining in the sample after rinsing is assayed tO
determine the amount of antibody-antigen coupling, and thus the amount of
antigen present in the sample. A cOmmonly used enzyme is horseradish
peroxidase, and its presence is measured by the addition of peroxidase
substrate. Colorimetric determination of enzyme substrate is then used to
measure the presence Of the antigen. A ‘sandwich’ ELISA is a variation of this
technique in which the antigen has at least two binding sites. The antigen reacts
first with excess solid phase antibody, and then with labeled antibody.

A disadvantage Of ELISA is that in most cases, sample enrichment takes
about 48 hours, and detection sensitivity is in the range Of 105 or 106 CFUlml
(Alocilja, 2000).

The Vitek Immunodiagnostic Assay System“

The Vitek lmmunodiagnostic Assay System (VIDASW‘) is an automated
enzyme immunoassay for the detection Of antigens using the Enzyme Linked
Fluorescent Assay (ELFA) method. .

A Solid Phase Receptacle serves as the solid phase as well as the

pipetting device for the assay, The Solid Phase Receptacle is coated with

30

antibodies. Reagents for the assay are ready-tO-use and pre-dispensed in the
sealed reagent strips (bioMerieux, 1998).

The assay is completely automated by the VIDAS instrument. An aliquot
of the heated enrichment broth is placed into the reagent strip and the sample is
cycled in and out of the Solid Phase Receptacle for a specific length of time.
Antigens present in the sample will bind to the monoclonal antibodies coating
the interior of the Solid Phase Receptacle. Unbound components are eliminated
during washing steps. Alkaline phosphatase labeled antibodies are cycled in
and out of the Solid Phase Receptacle and bind to any antigen captured on the
Solid Phase Receptacle wall in a sandwich reaction (Verozny-Rozand et al.,
1997). A final wash step removes unbound conjugates.

The“ final substrate (4-Methyl-umbelliferul phosphate) is cycled in and out
Of the Solid Phase Receptacle. The conjugated enzyme catalyzes the hydrolysis
of this substrate into a fluorescent product (4-Methyiumbelliferone).
Fluorescence is measured at 450 nm and expressed as a relative fluorescent
value (RFV) (Verozny-Rozand et al, 1997; biOMerieux, 1997).

When the VIDAS Assay is completed, the results are analyzed
automatically by the computer, a test value is generated, and a report is printed
for each sample. Test values are compared to a set of thresholds and each

sample is interpr‘eted as positive or negative (bioMerieux, 1997).

31

Vi. Biosensors

Biosensors are analytical instruments that allow direct measurement Of
biological materials, including foodbome pathogens (Kress-Rogers, 1997).
Many of these technologies Offer potential in assuring the safety of food products
without interrupting the manufacturing process.

A biosensor is made from a biological sensing element attached to a
signal transducer. The total effect Of a biosensor is to transform a biological
event into an electrical signal (Canh, 1993). The sensing element can be
enzymes, antibodies, deoxyribose nucleic acid (DNA), or microorganisms, either
integrated with, or in intimate contact with a physicochemical transducer (Turner
and Newman, 1998; Scott, 1998; Jonsson, 1998). The transducer may be
electrochemical, optical, thermometric, or piezoelectric. Electrochemical
transducers measure changes in current or voltage; optical transducers measure
changes in fluorescence, absorbance or reflectance; thermometric transducers
measure change in temperature due to the heat resulting from biological
reactions; and piezoelectric transducers measure changes in the resonant
frequency of a piezoelectric crystal due to small changes of mass or density at
the crystal surface (Rogers and Gerlach, 1996). The transducer signal
transforms the specific recognition of the analyte into a signal that can be readily
quantified, usually in real or near-real time.

Biosensors were born out Of the combination of existing sensors and
biological systems. They were first reported by Clark and Lyons in 1962, when

. biological molecules were coupled to transducers that converted the biological

32

signal into an electrical signal (Turner and Newman, 1998). These first sensors
were amperometric, comprised of glucose oxidase immobilized to oxygen
electrodes (Rogers and Gerlach, 1996; Canh, 1993). In 1967, Updike and Hicks
prepared an enzyme electrode by polymerizing a gel containing glucose oxidase
onto an oxygen electrode (Canh, 1993).

The biological component of the biosensor is used to confer specificity on
the device (Turner and Newman, 1998). In a biocatalysis-based sensor, the
biological component produces or consumes a component that can be detected
by the transducer. In an affinity-based sensor, the biological component binds a
molecule that the transducer then measures. The most widely used
biomolecules in biosensors are enzymes and antibodies, although other
biocatalytic components include whole cells, tissue slices, organelles, lectins,
and DNA (Jonsson, 1998).

Medicine has been the strongest driving force in biosensor research to
date, but the appeal of biosensor technology to the food industry is becoming
more Obvious. The potential advantage Of using biosensors in food analysis is a
rapid, specific quantification without the need for extensive sample preparation

(Table 1.2).

33

Table.1.2: Benefits and features of biosensors for use in the food industry
(Turner and Newman, 1998).

 

Feature Benefit
Target specificity Ability to definitively identify single compounds,
or a broad range Of compounds
Electronic integration Compact instrument design makes them easy
' to use

Operation in complex matrices Lack of need of sample preparation

Fast response time High throughput in automatic analyzers, and
decreased time for spot checks

Continuous signal Accurate monitoring of fluctuations, and ability
to take corrective action

Small size Portable, inexpensive, many simultaneous
assays

Mass producability inexpensive, disposable (hygienic), can be

widely dispersed

Biosensors are characterized by their specificity, and their sensitivity.
Specificity (or selectivity) is the ability to recognize a single compound among
other substances in the same sample. Specificity is achieved through ‘the
reaction Of an analyte with a specific biological component (Scott, 1998). The
specificity of biosensors is determined by both the bioreceptor and the method of
transduction. I

The sensitivity of a sensor is given by the change in its response as a
function Of the Change in input signal monitored (Canh, 1993). When the
variation in the phenomenon ceases to yield an appreciable variation in signal,
the detection limit has been reached, which usually corresponds to the limit of
the linear range at low concentrations. The linear range of a biosensor is

Obtained from a calibration curve of its response to different analyte

34

concentrations (Canh, 1993). For the measurement to be carried out properly, it
is necessary to know the response time of the biosensor, or the time taken to
reach a steady state from the instant of the variation in the concentration under
invesfigafion.

Today, biosensors are powerful tools that allow scientists to monitor
biOSpecific interactions in real time and to derive information about binding
kinetics and equilibrium, structure, and function. A key technology to biosensor
application for safety monitoring in the food industry is the construction of

immunosensors, which have application in microbial identification.

Vli. Surface Plasmon Resonance

Surface plasmon resonance (SPR) is a type of optical biosensor that uses
the properties of evanescent waves (Canh, 1993). It is widely used in the
biosensor, pharmaceutical, and analytical chemistry communities (Salamon et
al., 1999). The SPR sensor can be made into a highly specific biosensor to
detect biospecific interactions between proteins and biomolecules by forming a
functionalized sensor surface that is specific for a particular analyte. For
example, coupling an antibody onto the sensor surface converts the sensor into
an immunosensor. The SPR biosensor used in this study works by immobilizing
antibodies directly on to the metal surface and introducing a quantity of antigen.
The sensor is equipped with a thin gold film to which biotinylated antibodies
specific to the organism of interest are bound via avidin-biotin interaction. As

the antigens bind to the antibodies, the refractive index at the sensor surface

35

changes and affects the SPR coupling conditions. The change in refractive
index can be monitored and displayed as a function Of time. Thus, SPR has the
ability to measure, in real time, interactions of biomolecules due to the interfacial
refractive index changes caused by the antibody-antigen interactions.

initial applications of SPR involved investigating the optical properties
inherent to thin metal films (Earp and Dessy, 1996). Since then, SPR has been
used for assessment of antigen-antibody interactions, studies Of bio-recognition
interactions at surfaces, and screening foodsthfs or other materials for
pesticide, antibiotic or drug residues (Fratamico et al., 1998; Cranfield
Biotechnology Centre, 2000; Pfaff et al., 1994). Medical applications, such as
the human enzyme creatine kinase (CK), the .anticonvulsant drug phenytoin,
human chorionic gonadotrophin (hCG), and other biological contaminants in
matrices, such as blood, serum, plasma, saliva, and urine, have been
successfully measured with SPR (Wortberg et al., 1997). Additionally, SPR has
been used for characterization of anisotropic biological membranes (Salamon et
al., 1999), analysis of protein binding reaction kinetic parameters (Natsume et
al., 1994; Malmborg et al., 1992; Karlsson et al., 1991; Faegerstam et al., 1992;
O’Shannessy et al., 1993; O’Shannessy et al., 1994; Masson et al., 1994;
Shinohara et al., 1994), antibody detection and characterization (Medina, 1997),
pesticide analysis (Harris et al., 1996), detection Of chemical residues in milk
(Sternesjoe et al., 1995; Stemesjoe et al., 1996), chemical analysis Of ginseng
roots (Kajiwara, 1998), detection of Clostridium botulinum toxin (Ogert et al.,

1992), detection of Staphylococcal enterotoxin (Tempelman et al., 1996),

36

analysis Of Bacillus thuringiensis toxin binding (Masson et al., 1995), protein
interactions (He at al., 1997), lectin binding assays (Ozeki et al., 1998),
monitoring chemotaxis control in E. coli (Schuster et al., 1993), and biosensing
techniques (Liedberg et al., 1983). Fratamico et al. (1998) used an SPR
biosensor to detect E. coli O157:H7.

SPR relies on the excitation of the surface plasmon of a thin metal layer
covering the surface of the waveguide. The angle at which the incident light
best couples to the surface plasmon is sensitive to the refractive index in the
vicinity Of the metallized surface. Binding of large molecules such as antibodies
can thus be monitored as the system goes out of resonance (Turner and
Newman, 1998).

Waveguides and Modes

A waveguide is a physical medium through which light can be guided. A
common example Of this is the use offiber optics. In an SPR biosensor using
the Kretschmann prism arrangement, the Optical waveguide is a planar surface
coated with a thin metal film. The propagation of light occurs through total
internal reflection in the waveguide. When this occurs, the light ray is confined

within the waveguide, with very little leakage into the surroundings (Figure 1.3).

/\}\/eu~‘1y

02

 

 

Figure 1.3: Total internal reflectance. n1>n2, with angle of incidence 9.

37

The guiding medium must have a higher index of refraction (n) than the
surroundings in order for light to propagate by total intemai refraction. As long
as the angle of incidence (9) is larger than the critical angle, 9., (Figure _1.4),
suchthat .

ec = sin" (n./n2) (1.1)
there will be total intemai reflection (Sutherland and Daehne, 1987).

As long as a critical angle is surpassed, total intemai reflection can occur
at a boundary interface between any two refractive indices. The critical angle
defines a minimum angle of incidence for a particular interface, and depends on
the ratio _Of the refractive indices of the media involved (Earp and Dessey, 1996).
Evanescent waves I

An evanescent wave is an electromagnetic field that propagates along a
surface, but decays exponentially perpendicular to. it (Purvis et al., 1998). When
light is reflected at an optical interface where refractive index is changing, such
as in a waveguide, an evanescent wave develops as energy decays away from
the point Of reflection into the surrounding medium (Turner and Newman, 1998).
At a certain angle of incidence, there is total internal reflectance in the
waveguide. Anytime light undergoes total internal reflection, an evanescent field
is created. This energy field extends beyond the waveguide boundary into the
medium for a distance similar to the wavelength Of the light. There is a net flow

of energy across the reflecting surface to maintain the evanescent field.

38

Normal
Incident angle 9

    

 

 

Critical angle.....,__\\ a, Reflected
n1
02
Refracted
In > n; _ (a)
Normal

Critical angle

Total internal reflection

   
 

Incident angle

 

02

 

”1 > n2 (b)
Figure 1.4: Effect of the critical angle (BC) on reflection and refraction. (a)

Incident light angle less than 90, (b) Incident light angle greater than 9c, causing
total internal reflectance.

39

This transfer of energy results in attenuation in reflectance, and can be detected
with an optical photodetector.

The penetration depth Of the evanescent field will depend On the
wavelength of the light, the refractive index ratio Of the waveguide to the
surroundings, and the photon intensity in the plasmon mode. The penetration tip
of the field energy can be estimated as:

(1,: it (1.2)
41r(n12sin26-n22)"2

Where i. = the wavelength of light, n, = refractive index of the waveguide
material, n2 = refractive index Of the surroundings, and 9 = the angle of
incidence. The evanescent field is only present when there is total internal
reflectance in the Optical waveguide.

Surface Plasmons

A surface plasmon (SP) is an oscillation of electrons on the surface of a
solid, typically a conductor. Under certain conditions, the photon’s energy is
transferred to the surface of the metal as packets of electrons called plasmons.
This energy transfer occurs at a specific wavelength Of light, when the quantum
energy carried by the photon exactly matches the quantum energy level of the
plasmons. Plasmons are electron clouds that behave as if they were single
charged particles. Part of their energy is expressed as evanescent waves,
which extend about 100 nm above and below the metal surface, decaying

exponentially as a function Of distance. The interaction between the plasmon’s

40

evanescent wave and the matter within the evanescent field determines the
resonance wavelength or angle of incident light that resonates with the plasmon.
The magnitude of the change in resonance wavelength or the angle of incidence
is directly and linearly proportional to the change in composition at the surface
(Purvis et al., 1998).

The main criterion for a material to support SP waves is that the real part
of the dielectric permittivity be negative. The dielectric permittivity (e) is a
dimensionless quantity that is proportional to the square Of the refractive index
(n) of the material within the region of Optical wavelengths. The SP is affected by
changes in dielectric permittivity Of materials in contact with the thin metal film.
As these values change, they alter the coupling efficiency of the light into the
plasmon mode. TO find the coupling efficiency, the angle of incidence of the
light beam is scanned through a range of values. A distinct minimum in

reflectivity will be observed at a discrete angle, which is labeled the SP coupling

angle (95p). At this particular angle of incidence, light is most efficiently coupled . '

into the plasmon mode, and the reflection from the metal film is most attenuated.
Sensing is done by relating 95p to changes in the dielectric permittivity, and thus
the refractive index of the sample (Earp and Dessy, 1996).

In an SPR biosensor, thin gold or silver films are most Often used due to
their optical qualities and the ease and accuracy with which they can be
deposited onto a substrate. The metal film is deposited onto a glass substrate
that will be optically coupled to a waveguide, and the other side of the thin film is

exposed to the analyte sample. Surface plasmon resonance occurs as the

41

evanescent wave propagates through the metal and excites surface plasmons.
The resonance wavelength can be determined by measuring the light reflected
by (the metal surface. At most wavelengths, the metal acts as a mirror, reflecting
virtually all incident light. At the SP coupling angle (esp), surface plasmons are
created and the incident light is almost completely absorbed as SPR occurs
(Quantech, 1998).

SPR can be achieved by varying the frequency of the light, or by varying
the angle of incidence, as changing the wavelength at a fixed angle is equivalent
to Changing the angle at a fixed wavelength. In either case, at some-point
resonance occurs, and the reflected intensity of the light drops significantly. The
position of the SPR is extremely sensitive to the refractive index of the sample
(Texas Instruments, 1999).

The Kretschmann Geometry

The Kretschmann prism arrangement is the most frequently used
geometry in SPR sensor design (Earp and Dessy, 1996). Only transverse
magnetic polarized light (the electric field polarized in the plane of incidence)
may couple to surface plasmons (Texas instruments, 1999). The Kretschmann
prism arrangement facilitates this coupling of a light wave onto a surface
plasmon (SP) (Figure 1.5). In the Kretschmann geometry, the angle of incidence
Of the light-beam with respect to the metal surface, and the reflected light

intensity are measured.

42

Glass Prism

 

From light To
source photodetector
9
Glass Slide

 

 

Metal Film

AVA

 

Liquid Sample

Figure 1.5. The Kretschmann geometry, with angle Of incidence 9. When
9 = esp, virtually no light is reflected to the photodetector.

Incident light passes through a prism and onto the thin metal film, to
which antibodies for the analyte Of interest have been bound. The SP coupling
angle is sensitive to changes in the refractive index of the biomolecule layer.
Light is emitted at varying angles Of incidence, and at the SP coupling angle,
SPR occurs and the optical photodetector receives almost no light. Both the
angle of incidence Of the light-beam and the reflected light intensity are
measured.

Increasing the concentration of proteins in a given area will create a
refractive index change that is directly proportional to mass loading. The
sensitivity of the Kretschmann prism device will depend on how accurately the
resonance angle can be measured. The response Of an SPR instrument due to
bimolecular binding is essentially the same for various proteins and

biomolecules at similar concentrations. This is because the refractive indices for

43

many different macromolecules are basically the same, regardless of
composition (Stenberg et al., 1991).

In the SPR biosensor used in this research, antibodies are immobilized
directly to the metal surface via avidin-biotin interaction, and as the antigen
binds to the antibodies, the refractive index at the sensor surface changes.
Changes in the analyte layer on the sensor surface affect the effective refractive
index of the metal film-antibody interface, which affects the 'SP coupling angle.
Varying the incident angle of light locates the SP coupling angle as the angle at
which virtually no light is reflected. This Change in the amount of light striking
the photodetector, and the corresponding angle of incidence are the sensor

outputs.

44

REFERENCES

Alocilja, E.C., Ryser, E., Osburn, W., 2000. Distribution and Prevalence of
Foodborne Pathogens in the Pork Industry Chain. Project proposal to
USDA-National Research initiative Competitive Grants Program ‘
(NRICGP)

Anderson, A. 1994. What is the HACCP? Journal of Human Nutrition and
Dietetics 7(1):53-60

Besser, R.E., Lett, S.M., Weber, J.T., Doyle, M, P., Barrett, T.J., Wells, J.G.,
Griffin, PM. 1993. An Outbreak of Diarrhea and HemO/ytic Uremic
Syndrome from Escherichia coli O157:H7 in Fresh-Pressed Apple Cider.
Journal Of the American Medical Association 269222172220

BiOMerieux, 1998. VIDAS E. coli O157:H7 (ECO) Manual. biOMerieux, lnc.,
595 Anglum Drive, Hazelwood, MO 63042-2320

Bryan, F. L. 1990. Application of HACCP to ready-to—eat chilled foods. Food
Technology 44(7):70-77

Buchanan, R.L., and Doyle, MP. 1997. Foodborne Disease Significance of
Escherichia coli O157:H7 and other Enterohemorrhagic E. coli.
FOOdTechnology 51:69-76

Buzby, J., Roberts, T. 1996. Microbial foodborne illness. Choices; First
Quarterzi4-17 '

Buzby, J., Roberts, T., Lin, C.-T., MacDonald, J., 1996. Bacterial Foodborne
Disease: Medical Costs and Productivity Losses. USDA Agricultrual
Economic Report NO. 741

Canh, TM, 1993. Biosensors. Chapman and Hall, London, UK

CAST- Council for Agricultural Science and Technology, 1994. Foodborne
Pathogens: Risks and Consequences. Task Force Report NO. 122,
Washington DC.

Cranfield Biotechnology Centre, 2000. Available at URL: mtpzllwww.cranfiel¢
ac.uklbiotech/sm htm

Davies, P. 1997. Food safety and its impact on domestic and export markets.
Swine Health Product 5:13-20

Deshpande, SS, 1994. Immunodiagnostics in agricultural, food, and
environmental quality control. Food Technology. June 1994;136:141

45

Doyle, MP, and Schoeni, J.L. 1987. Isolation of Escherichia coli O157:H7 from
retail fresh meats and poultry. Applied Environmental Microbiology
53:2394-2396

Doyle, M.P., Beuchat, L.R., Mountville, T.J., 1997. Food Microbiology:
Fundamentals and Frontiers. ASM Press. Washington DC.

Earp, R.L., Dessy, RE, 1998. Surface Plasmon Resonance. In Ramsey, G.,
( ed.) Commercial Biosensors, Applications to Clinical, Bioprocess, and
Environmental Samples. Wiley-lnterscience Pub., pp 99-164

Earp, R.L., Dessy, RE, 1998. Surface Plasmon Resonance. In Ramsey, 6.,
(ed) Commercial Biosensors, Applications to Clinical, Bioprocess, and
Environmental Samples. Wiley-lnterscience Pub., pp 99-164

Epling, L.K., and Carpenter, J.A., 1990. Antibiotic Resistance of Salmonella
isolated from pork carcasses in Northeast Georgia. Journal of Food
Protection, 532253-254

Faegerstam, L.G., FrostelI-Karlsson, A., Karlsson, R., Persson, B., Roennberg,
l., 1992. Biospecific interaction analysis using surface plasmon resonance
detection applied to kinetic, binding site and concentration analysis.
Journal Of Chromatography 597:397-410

Foster, EM, 1997. Historical Overview of Key Issues in Food Safety. Emerging
Infectious Diseases, Special Issue. 3(4). Available from URL:
http://wwwcdcgov/nciggg/EiD/vol3no4/foster.htm

Fratamico, P.M., Strobaugh, T.P., Medina, M.B., Gehring, AG, 1998. Detection
Of Escherichia coli O157:H7 using a surface plasmon resonance
biosensor. Biotechnology Techniques, Vol 12, NO. 7, pp 571-576

Griffin, PM. 1995. Escherichia coli O157:H7 and other enterohemorrhagic
Escherichia coli. In Infections of the Gastrointestinal Tract. Blaser, M.J.,
Smith, P.D., Ravdin, J.l., Greenberg, H.B., and Guerrant, R.L. (eds).
Raven Press Ltd, New York, NY

Griffin, PM, and Tauxe, R.V. 1991. The Epidemiology of Infections Caused by
Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the
associated hemolytic uremic syndrome. Epidemiology Review. 13:60-98

Harris, R.D., Luff, B.J., Wilkinson, J.S., Wilson, R., Schiffrin, D.J., Piehler, J.,
Brecht, A., Abuknesha, R.A., Mouvet, C., 1996. Integrated optical surface
plasmon resonance biosensor for pesticide analysis 1995/6 Research
Journal, Optical Fibres Group, University of Southampton, contact
riourna@ecs.soton.ac.%

46

He, X., Shen, L., Malmborg, A.C., Smith, K.J., Dahlbaeck, B., Linse, S., 1997.
Binding site for C4b-binding protein in vitamin k-dependent protein 5 fully
contained in carboxy-terminal laminin-g-type repeats. A study using
recombinant factor lX-protein S chimeras and surface plasmon
resonance. Biochemistry 36:3745-3754

Hoffmann, SI. 1993. Southwestern Internal Medicine Conference: Shiga-Like
Toxins in Hemolytic—Uremic Syndrome and Thrombotic Thrombocytopenic
Purpura. American Journal of Medical Science 306:398—406

Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, eds. 1994. Foodborne
Disease Handbook, Volume 1: Diseases Caused by Bacteria. Marcel
Dekker, Inc. New York, NY

IDEXX Food Safety Net, 1998. The Comprehensive Hazard Analysis Critical
Control Point (HACCP) Course Participant’s Guide. IDEXX Food Safety
Net Services, Westbrook, ME '

IFST (Institute of Food Science and Technology), 1997. Salmonella
Typhimurium DT104 Position Statement 11 Nov. 1997. Available at URL:
httpzllwww.ifst.orglhottop20.htm

isaacson, R.E., Firkins, L.D., Weigel, R.M., Zuckermann, F.A., DiPietrO, J.A.,
1999. Effect of transportation and feed withdrawal on shedding of
Salmonella Typhimurium among experimentally infected pigs. AJVR,
60:1155—1158

Jay, J.M, 2000. Modern Food Microbiology. (Sixth ed.) Aspen Publishers, inc.
Gaithersburg, MD

Jonsson, U., 1998. Optical Affinity Biosensors. In in AD. Scott (ed) Biosensors
for Food Analysis, Athenaeum Press Ltd, Gateshead, Tyne and Wear,
UK. PP. 37-46

Kajiwara, H., 1998. Motif 2 in adenosine kinase homologous hinseng polypeptide
showed affinity to D-n'bose by capillary zone electrophoresis and surface
plasmon resonance Journal of Chromatography 817:173-179

Karlsson, R., Michaelsson, A., Mattsson, L., 1991. Kinetic analysis Of mOnocIonal
antibody-antigen interactions with a new biosensor based analytical
system Journal of Immunological Methods, 145(1991)229—240

Karpiskoval, R., Benes, 0., Dedicova, D., 1999. Emergence Ofa multidrug
resistant Salmonella Typhimurium DT 104 in the Czech Republic.
Eurosurveillance 4:56-58

47

Kress-Rogers, E. (ed.) 1997. Handbook of Biosensors and Electronic Noses.
CRC, Boca Raton, Florida

Kudva, l.T., Hatfield, P.G., Hovde, C.J. 1997. Characterization of Escherichia coli
0 157:H 7 and other Shiga Toxin-Producing E. coli Serotypes Isolated from
Sheep. Journal of Clinical Microbiology, 35(4)892-899

Letellier, A., Messier, S., Quessy, S., 1999. Prevalence of Salmonella spp. and
Yersinia enterocolitica in finishing swine at Canadian abattoirs. Journal of
Food Protection, 62:22-25

Liedberg, B., Nylander, C., Lundstroem, l. 1983. Surface plasmon resonance for
gas detection and biosensing Sens. Actuat. 4:299-304

Lin, J.S., Tsen, H.Y., 1999. Development and Use Of Polymerase Chain
Reaction for the Specific Detection of Salmonella Typhimurium in Stool
and Food Samples. Journal of Food Protection, Vol. 62, No. 10:1103-
1110

Malmborg, A.C., Michaelsson, A., Ohlin, M., Jansson, B, Borrebaeck, CK, 1992.
Real time analysis Of antibody-antigen reaction kinetics Scandinavian
Journal of immunology 351643-650 '

Masson, L., Mazza, A., Brousseau, R., 1994. Stable immobilization Of lipid
vesicles for kinetic studies using surface plasmon resonance Analytical
Biochemistry 218: 405-412

Masson, L., Mazza, A., Brousseau, R., Tabashinik, B., 1995., Kinetics of Bacillus-
thuringiensis toxin binding with brush border membrane vesicles from
susceptible and resistant larvae of Plutella xylostella. The Journal of
Biological Chemistry 270(20):11887-11896

Mead, P.S., Slutsker, L., Dietz, V., McCaig, L., Bresee, J., Shapiro, Griffin, P.,
and Tauxe, R., 1999. Food-related illness and death in the United States
Emerging Infectious Disease, Vol. 5

Medina, MB, 1997. Hygromycin B Antibody Production and Characterization by
a Surface Plasmon Resonance Biosensor Journal of Agricultural Food
Chemistry 452389-394

Monk, J.D., Beuchat, L.R., Doyle, MP, 1995. Irradiation inactivation of
foodborne microorganisms. Journal of Food Protection 58(2)197-208

Morse, D.L., Birkhead, G.S., Guzewich, J.J., 1994. Investigating foodbome
disease, in Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, (eds)
Foodborne Disease Handbook, Volume 1: Diseases Caused by Bacteria.
Marcel Dekker, Inc. New York, NY, pp. 547-603

48

Moutville, T.J., 1987. Food Microbiology. CRC Press, NY p. 643

Natsume, T., Koide, T., Yokota, S., Hirayoshi, K., Nagata, K., 1994. Interactions
between collagen-binding stress protein HSP47 and collagen The Journal
of Biological Chemistry 269(49)31224-31228

Neill, M. A., Tarr, P.l., Taylor, D.N., TrOfa, AF, 1994. Escherichia coli in Hui,
Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, (eds) Food Disease
Handbook, Volume 1: Diseases Caused by Bacteria. pp. 169-213

O’Shannessy, D.J., Brigham-Burke, M., Soneson, K.K, Hensley, P., Brooks, l.,
1993. Determination of rate and equilibrium binding constants for
macromolecular interactions using surface plasmon resonance: Use of a
nonlinear least squares analysis methods Analytical Biochemistry
212:457-468

O’Shannessy, D.J., Brigham-Burke, M., Soneson, K.K., Hensley, P., Brooks, l.,
1994. Determination of rate and equilibrium binding constants for
macromolecular interactions by surface plasmon resonance Methods in
Enzymology, 240:323-349

Ogert, R.A., Brown, J.E., Singh, B.R., Shriver-Lake, L.C., Ligier, PS, 1992.
Detection of Clostridium botulinum toxin a using a fiber optic-based
biosensor Analytical Biochemistry 205:306-312

Orskov, l., Orskov, F., Jann, B. 1977. Serology. Chemistry, and Genetics of O
and K antigens of Escherichia coli. Bacteriology Review, 41 :667-710

Ozeki, M., Furuichi, Y., Umekawa, H., Takahashi, T., 1998. A surface plasmon
resonance assay for the binding at Amaranthus Hypochondriacus var.
Mexico Lectin to glycoprotein. Biochemistry and Molecular Biology
international. 44(1):211-216

Pfaff, M., Goehring, W., Brown, JD, Timple, R. 1994. Eur. Journal of
Biochemistry 225:975-984

Quantech, 1998. Available at URL: http2/lwww.biosensor.com/quantechlplasintr

Ratnam, 8., March, 88., Ahmed, R., Bezanson, S., Kasatiya, S. 1988
Characterization of Escherichia coli Serotype 0157:H7. Journal Of Clinical
Microbiology. Oct 1988:2006-2012

Rogers, K.R., Gerlach, CL, 1996. Environmental Biosensors, A status report.
Environmental Science and Technology, 30(11)486:491

Ryser, ET, 2000. Strain specific typing bacterial pathogens in the pork industry
chain. Final report for NPPC project #99-026

49

Salamon, 2., Brown, M.F., TOIlin, G., 1999. Plasmon resonance spectroscopy:
probing molecular interactions within membranes. TIBS 24 — June
1999:213-219

Schuster, S.C., Swanson, R.V., Alex, L.A., Bourret, R.B., Simon, MI, 1993.
Assembly and function Of a quaternary signal transduction complex ,
monitored by surface plasmon resonance. Nature 365:343-347

Scott, A.O. (ed), 1998. Biosensors for Food Analysis, Athenaeum Press Ltd,
Gateshead, Tyne and Wear, UK

Shinohara, Y., Kim, F., Shimizu, M., Goto, M., Tosu, M., Hasegawa, Y., 1994.
Kinetic measurement of the interaction between an oligosaccharide and
lectins by a biosensor based on surface plasmon resonance European
Journal of Biochemistry 223: 1 89-1 94

Simmons, NA 1997. Global Perspective on Escherichia coli O157:H7 and Other
Verocytotoxic E. coli spp.: UK Views. Journal of Food Protection, Vol. 60,
No. 11:1463-1465

Stenberg, E., Persson, 8., R005, H., Urbaniczky, C., 1991. Quantitative
determination Of surface concentration of protein with surface plasmon
resonace using radiolabe/ed proteins. Colloid Interface Science
143(2):5‘3-527

Stemesjoe, A., Mellgren, C., Bjoerck, L., 1995. Determination of sulfamethazine
residues in milk by a surface plasmon resonance-based biosensor assay.
Analytical Biochemistry 226: 1 75-1 81

Stemesjoe, A., Mellgren, C., Bjoerck, L., 1996. Analysis Of sulfamethazine in milk
by an immunosensor assay based on surface plasmon resonance. in
lmmunoassays for residue analysis, American Chemical Society, 463-470

Sutherland, R., Daehne, C. 1987. IRS devices for Optical immunoassay, in
A.P.F. Turner, Karube, l., Wilson, 66. (eds) Biosensors: Fundamentals
and Applications. Oxford Science Publications, Oxford, pp. 655-678

Tauxe, R.V. 1991. Salmonella: A postmodern pathogen. Journal of Food
Protection 542563-568

Tempeiman, L.A., King, K.D., Anderson, G.P., Ligier, PS, 1996. Quantitating
Staphylococcal enterotoxin b in diverse media using a portable fiber-Optic
biosensor Analytical Biochemistry 233250-57

Texas instruments, Inc. Dallas, Texas, USA. 1999. Operation Manual, Spreeta
Experimenter’s Kit

50

Turner, A.P., Newman, JD, 1998. An Introduction to Biosensors, in AC. Scott
(ed) Biosensors for Food Analysis, Athenaeum Press Ltd, Gateshead,
Tyne and Wear, UK, pp. 13-27

U.S. Meat Export Federation, 1997. Total US. Pork Exports. Available from
URL: httpzllwww.usmef.org/expstathORK8796.html

USDA, 1999. Food Safety: The Agricultural Use Of Antibiotics and its
Implications for Human Health. Letter Report, 4/28/99, GAO/RCED-99-
74. Available at URL: http:/lwww.texas-naturaI-beef.com/food_safety.htm

USDA-FSIS, 1998. Report to Congress, FOOdNet: An Active Surveillance
System for Bacterial Foodborne Diseases in the United States.
Washington, DC. '

Varnam, AH, and Evans, MG. 1991. Foodborne Pathogens. Mosby-Year Book,
Inc., St. Louis, MO

Verozny-Rozand, C., Mazuy., C., Ray-Gueniot, S., Boutrand-Loei, S., Meyrand,
A., Richard, Y., 1997. Detection of Escherichia coli O157:H7 in French
food samples using an immunomagnetic separation method and the
VIDAST'“ E. coli 0157. Letters in Applied Microbiology, 252442-446

Wang, T.K., Tsen, T.C., Lee, J.H. 1994. Analysis of Salmonella serovars in
Taiwan by the phase induction method. Chinese Journal Of
Microbiological Immunology. 27:13-24

WHO (World Heath Organization), 1997. Multi-drug resistant Salmonella
Typhimurium. WHO Fact Sheet NO. 139, Available at URL: httgzll
www.who. int/inf-fs/en/fact1 39. html

Wortberg, M., Orban, M., Renneberg, R., Cammann, K. 1997. Fluorimetric
lmmunosensors. in Kress-Rogers, Ed, Handbook Of Biosensors and
Electronic Noses. CRC, Boca Raton, Florida, p. 369-405

WVA (World Veterinarian Association), 1997. Multi-Drug Resistant Salmonella
Typhimurium. WVA Bulletin 14(2). Available at URL: http:/lwww.who.ch

Ziprin, R.L., 1994. Salmonella. In Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver,
D.O (eds) Foodborne Disease Handbook, VOIume 1: Diseases Caused
by Bacteria. Marcel Dekker, Inc. New York, NY, pp. 253-318

51

CHAPTER 2

MATERIALS AND METHODS

Spreeta SPR Biosensor

The Spreeta Miniature Integrated Surface Plasmon Resonance Liquid
Sensing System (Texas Instruments, lnc., Dallas, Texas) was used in this study
(Figure 2.1).

Reflecting

. , . W. Mirror
' ”~33; ‘ SUbStI‘ate

Optical Plastic

  
  
     

Surface Plasmon

 

 
    

    

 

. .- -'-t " 3,14,: .fj.7::._:~:,,:~: ,5:-. :~' 15:: ‘ , .~ '- “:“-:~'.;
: ':- -.j-:,j:.:-":-.‘.1:j:,:::,:>:;'.. ' ' ‘ ‘ " ' ‘ ‘
' ' . - - L: 4 .-'. - \ - ,
r : . r : \ t
I, : \ . _ : ". I
' . \ r ’ . .‘ -.
¢\ .1 .’ = t 'j.
0 f '.
‘-. :- r g g. S

\

Temperature Photodiode
Sensor Array

Light Emitting
Diode Polarizer

Figure 2.1. Side view Of the Spreeta Sensor (courtesy of Texas Instruments,
1999, URL:

Near—infrared light (840nm) from a light emitting diode (LED) is polarized
to enhance SPR. The light beam reflects Off of the gold sensing film and is
directed by the gold mirror onto a linear array of silicon photo-diodes. The active
sensing region is the area on the thin gold film that is actively used in liquid
sensing. it is a strip approximately 4.5 mm long by 0.1 mm wide on the face of
the sensor. Except for the sensing surface, the sensor is coated with an Opaque

material to block out external light (Texas instruments, 1999).

52

The angle at which a ray of light is incident upon the sensing film is
directly mapped into a specific point on the photodiode array. Depending on the
refractive index of the liquid next to the gold film, at some angle the reflected light
intensity will experience a minimum corresponding to where SPR occurs. This
minimum is then‘processed by the Spreeta computational software, and plotted
as the change in index of refraction over time (Texas Instruments, 1999).

The SPR biosensor was assembled according to the Spreeta Operation
Manual. Neutravidin binding and attachment of biotinylated antibodies to the
surface was done following the method Of Spreeta’s Application Brief 004, with
modifications (Texas Instruments, 1999). Table 2.1 enumerates the steps and
time to prepare the SPR biosensor.

Table 2.1. Typical preparation of the SPR biosensor.

 

 

Step Action Time (minL
1 . Clean gold surface by immersing in NaOH/Triton X-100 3
Solution
2 Baseline established by immersing in PBS 3
3 Avidinate gold surface by immersing in neutravidin 3
4 Rinse off excess neutravidin by immersing in PBS 3
5 Attach antibody to gold surface via avidin-biotin interaction, 10
by immersing gold surface in biotinylated antibody
6 Rinse off unbound antibody by immersing in PBS 2
7 Rinse with PBS-NaOH/Triton X-100 solution 2
8 immerse in BSA 2
9 Rinse with PBS-NaOH/Triton X-100 solution 2
Total time: 30

 

The total time to prepare the SPR biosensor for an assay was 30 min.
The Spreeta computational software recorded the index of refraction Of each
sample approximately every 4.8 seconds. To initialize each sensor, the gold

surface was cleaned by immersing it in 10 ml of 0.1 N NaOH in 1% Triton X-100

53

Solution (.NaOHfTriton) (Sigma Chemical CO., St. Louis, M0) for 3 minutes
(Figure 2.2-1). The gold surface of the sensor was then immersed in 10 ml
phosphate buffered saline (PBS; pH 7.2) (Sigma Chemical CO., St. Louis, M0)
for 3 minutes, until a steady PBS baseline was established (Figure 2.2-2). Then
the sensor was immersed in 10 ml 100 ullml neutravidin in PBS (Figure 2.2-3)
(Pierce Chemicals, Rockford, II), then back into 10 mi of PBS (Figure 2.2-4).
Once a new baseline was achieved, the sensor was immersed in 10 ml Of 300
ug/ml anti-Salmonella spp. or anti-E. coli 0157:H7 antibody (Kirkegaard & Perry
Laboratories Inc., Gaithersburg, MD) for 10 minutes, to bind the antibody to the
avidinated surface via avidin-biotin interaction (Figure 2.2-5). Unbound antibody
was rinsed Off by immersing the sensor in 10 ml PBS (Figure 2.26). The sensor
was then immersed in NaOH/Triton in PBS for two minutes (Figure 2.2-7), then
Bovine Serum Albumin (BSA)( Pierce Chemicals, Rockford, II) for two minutes to
block any non-specific binding sites (Figure 2.28), and finally back into
NaOH/Triton in PBS for two minutes, to rinse off any remaining BSA (Figure 2.2-
9). The procedures for the assays performed after preparation Of the SPR
sensor differed depending on the objective of the experiment. '
Statistical Analysis

Statistical analysis was conducted using a single factor ANOVA (analysis
of variance) function on experiment replications. For all studies, the SPR
biosensor average responses were considered statistically different when the P-

value was less than 0.05 (95% confidence level).

54

.me E cogEIOaz av {mm as .mma E caeEanz E 525.8 was 6v $39.6
622585 E 525.8 mma E 55.8 £222.32 5 525.8 am“: 8:3 85:8 2232a
E 525.8 cagEIOaz 5 bass 328 E 5:932“. .8585 «am .835 < .3 26E

EE. 2.:
mm cm 9 2 m o

 

 

I _ . . _ . 33
- an?
. 83
4 B?
L 83

In? .

 

.vme

 

 

 

Sume

55

uonoeueu jo xepU|

Standard Plate Counts

Viable plate counts were performed by serially diluting the sample into
0.1% buffered peptone-water (BPW; Becton Dickinson and CO., Sparks, MD).
One ml of each serial dilution was plated on Petrifilm aerobic plates (3M, St.
Paul, MN), and incubated for 18-24 hours at 37°C.
Preparation and Biotinylation of Antibodies

Lyophilized affinity purified antibody to Salmonella common structural
antigens (GSA-1) and to Escherichia coli O157:H7 were purchased (Kirkegaard
& Perry Laboratories Inc., Gaithersburg, MD). All antibodies were stored at 4°C
until rehydrated. Antibodies were rehydrated according to the manufacturer’s
instructions for conjugations in carbonate buffer by adding 0.1 ml Of 0.01 M acetic
acid (Spectrum Products, lnc., Gardena, CA) to 1 mg Of the antibody. Once
totally dissolved, 0.1 ml Of 0.177 M carbonate-bicarbonate solution was added to
the antibody, and mixed thoroughly. The antibody was then heated in a 37°C
water bath for 30 minutes, then allowed to cool slowly to room temperature.
Biotin labeling was done according to the method of DeMarco et al. (1999). Six
hundred microliters of PBS was added to the rehydrated antibodies. Two
milligrams of succinimidyl-6-(biotinamido) hexanoate (EZ—Link NHS-LC-Biotin;
Pierce Chemicals, Rockford, ll) was added to 1 ml of N,N-dimethylformamide
(DMF; Aldrich Chemicals, Milwaukee, WI), and 75 pl of this solution was added
' to the rehydrated antibody. The antibody solution was then placed on ice for 2

hours to achieve an insertion of approximately 2 biotins per molecule

56

lmmunoglobulin-G (lgG) (DeMarco et al., 1999). The biotinylated antibody was
then serially diluted with sterile PBS, toiachieve a concentration Of 300 uglml.
Bacteria

Characterized strains of S. Typhimurium and E. coli O157:H7 were
obtained from Michigan State University collections. E. coli O157:H7 was
verified at the Bacteriology Laboratory at the Veterinary Diagnostic Center of the
University of Nebraska, Lincoln (Younts, 1999). For verification as E. coli
0157:H7, the isolates were subjected to PCR for the presence of the eae gene,
Shiga toxin (Stx) structural gene, and the 0 antigen biosynthesis (rfb) loci
(Younts, 1999). All culturing was done in a certified Biological Safety Level 2
environment at the MSU Meat Microbiology Laboratory. Cultures were
maintained in tryptic soy broth (TSB; Becton Dickinson, Sparks, MD) at-4°C.
Cultures for viable cell counts and SPR assays were grown in T88 at 37°C for
18-20 hours. Cultures were serially diluted in sterile BPW for use in the
experiments.
SPR Biosensor Signal Normalization

The output of the biosensor was the index of refraction at which surface
plasmon resonance occurred. TO normalize the output, the index of refraction for
the zero controls was subtracted from the responses. The resulting biosensor
output was thenthe magnitude of the change in index of refraction (the increase

in index of refraction above the index of refraction of the zero control).

57

Comparisons of Antibody Concentration

Anti-Salmonella and anti-E. coli 0157:H7 biotinylated antibodies were
prepared at 3 concentrations in sterile PBS; 300 uglml, 30 ug/ml, and 3 uglml.
To analyze the difference in response of the SPR biosensor to differing antibody
concentrations, 3 ten-fold dilution series of S. Typhimurium and E. coli O157:H7
were made in sterile BPW, and run on each of the 3 antibody concentrations.
Viable plate counts Of the serial dilutions were performed to verify the
concentration of the organisms. Results were compared to see if there was any
difference in response Of the SPR biosensor to increasing concentratiOns Of
anfibody.
Background Response

The response of the SPR sensor to assays performed without the use of
an antibody was monitored to establish the background response of the system.
Two separate experimental trials were performed on the sensor using serial
dilutions of S. Typhimurium, E. coli O157:H7, and uninocuiated TSB. Viable
plate counts of the serial dilutions were performed to verify the concentration of
the organisms. The SPR sensor was prepared in the same way as in Table 2.1,
with sterile PBS introduced in place of the antibody. For each bacterium, ten-fold
dilutions of a 10° CFU/ml culture grown in TSB were made in sterile BPW. For
the TSB, serial dilutions were made in sterile BPW, and the series run as if they
were inoculated samples. Sterile BPW was used as the zero (uninocuiated)
control. The results Of the assays were then compared to find any difference in

response of the SPR sensor to inoculated versus uninocuiated samples in TSB.

58

Negative Controls

The negative control used for the serial dilutions of S. Typhimurium and E.
coli 0157:H7 was TSB diluted in BPW. From a trial experiment conducted
earlier, the SPR biosensor baCkground signal showed that the biosensor was
sensitive to the presence of sterile TSB. Since the organism was grown in TSB,
this sensitivity could potentially be misleading, so assays Of sterile TSB on
prepared biosensors were performed to find the contribution of TSB in the
sample to the SPR response. Six 10-foid serial dilutions of TSB in sterile BPW
were performed, 3 assays using anti-Salmonella spp. antibody, and 3 with anti-E.
coli O157:H7 antibody. The results were analyzed to find the amount of SPR
biosensor response that was due to TSB. Results were also compared to the
background signal data (assays performed without antibody), to see if there was
a statistical difference between the background response and an antibody-
prepared SPR biosensor response to sterile TSB. The SPR biosensor response
to sterile TSB was? then used as the negative control, in order to account for the
TSB effect on the biosensor response.
SPR Biosensor Sensitivity Assays

SPR biosensor assays were performed on a series of dilutions Of S.
Typhimurium and E. coli O157:H7 in sterile BPW, from starter cultures inoculated
to approximately 108 CFU/ml (Table 2.2). Concentration was verified by

conducting viable plate counts.

59

Table 2.2. A typical dilution series experiment. The dilution series consists of
SPR analysis Of 10 vaiying bacteria concentrations (CFU/ml).

 

 

Step - Action Time (min)

1 SPR sensor preparation (Table 2.1) 30
2 Immerse sensor in sterile BPW (zero control) 2
3 Immerse sensor in sample 2
a. repeat for 9 additional bacterial concentrations 18

4 Rinse gold surface with PBS , 1
Total time: 53
Time for one biosensor assay: 35

 

After the SPR biosensor had been prepared according to Table 2.1, the
biosensor was immersed in the zero control (10 ml of sterile BPW) for 2 min.
Once a steady reading-was established, the sensor was then immersed in 10 ml
Of the sample for 2 min. This process was repeated for each Of the dilutions
prepared. Assays were prepared on each sensor with the lowest concentration
(0 CFU/ml) first, and the highest concentration (10° CFU/ml) last. The index of
refraction for each dilution was computed as the average of the readings taken
over each 2 min period.

The total time for a dilution series experiment, with 10 samples of
increasing concentration (10° — 10° CFU/ml) was 53 min. The time for one SPR
biosensor sample assay (procedure without step 3a) was 35 min.

Growth Curves

Growth activity of S. Typhimurium and E. coli O157:H7 was monitored by
performing a series Of three growth curves for each bacterium. In each growth
curve, 10 III Of10° CFU/ml stock solution of s. Typhimurium or E. coli O157:H7
(about 10 CFU) were introduced to 100 ml of sterile TSB. The inoculated

solutions virere grown at 37°C in a stationary incubator. At half-hour intervals for

60

ten hours, aliquots of 1 ml were removed from the vials, serially diluted, and plate
counts performed. The results from the plate counts were plotted over time to
determine the time required by the bacteria to grow to the level of detection of the
SPR biosensor. I
Specificity Assays

Specificity assays were done to establish the ability Of the SPR biosensor
to select for the target antigen. The biosensor was prepared as in Table 2.1,
then exposed to increasing concentrations Of three sample dilutants:

1. Sterile TSB (negative control).

2. Bacteria that the antibody was not specific for:

2.1. E. coli O157:H7 dilutions for anti-Salmonella spp. prepped SPR
biosensor.

2.2 S. Typhimurium dilutions for anti-E. coli O157:H7 prepped SPR
biosensor.

3. The bacteria of interest:

3.1. E. coli O157:H7 dilutions for anti-E. coli'O157zH7 prepped SPR
biosensor.

3.2 S. Typhimurium dilutions for anti-Salmonella spp. prepped SPR
biosensor.

Plate counts of the serial dilutions were performed to verify the
concentration of the organisms. Three replications were done for each dilution
series for both anti-Salmonella spp. - prepped SPR biosensor and anti-E. coli

O157:H7-prepped SPR biosensor, making a set Of 18 assays.

61

Specificity in Mixed Cultures

Sample cultures composed of mixtures Of S. Typhimurium and E. coli
0157:H7 were analyzed on the SPR sensor to establish the ability Of the SPR
biosensor to select for the target antigen in the presence of other bacteria. The
sample cultures were prepared from stock solutions of known concentration, on
which viable plate counts were performed to verify the concentration Of the
organisms. I
Field Samples

Samples were obtained from a previous study, in which site visits were
made to small-sized (~200 heads/week) and medium-sized (>1000 heads/week)
commercial meat packing facilities, as well as to the farms that supplied them
with hogs (Tables 2.3-2.4) (Ryser, 2000).

Table 2.3. Environmental samples from slagqhterhouse (Ryser, 2000)

 

 

 

 

 

Number
Method and of
Site Salee Quantity samples
Pen Alleyway Fecal material, floor Scoop, 509 1
Prechill Swab from chillroom wall Swab 7
Dehairing Hairs scooped from dehairing 259, hair scoop 6
machine machine
Drain Drain near evisceration Swab 3
Table 2.4. Farm samples (Ryser, 2000)
Number
Of
Site 4 Sample Method and Quantity samples
Pen Alleyway Fecal material, Scoop, 509 1
floor
Hogs back Composite Swab with neutralizing buffer 2
solution
Fecal matter Composite Scoop, 509 1
Feed Composite Scoop, 509 2
Water Nozzle Composite . Swab from nozzle tip 1

 

62

Twenty-four samples were analyzed by mini-VIDAS and SPR biosensor for
Salmonella spp. and E. coli O157:H7.

Field Sample Analysis

Salmonella spp. and E. coli O157:H7 were assayed using both the mini-

ViDAST" system and the SPR biosensor:

VIDAST" Salmonella

Recovery of Salmonella spp. using the VIDAS” system was done

according to the BAM/AOAC method (bioMerieux, 1998). The samples were pre-
enriched in lactose broth, then 1 ml Of the pre-enriched sample was incubated in
9 ml of buffered peptone water (BPW) for 18 hours at 37°C. Following this, 1 ml
of the BPW enrichment was transferred into 10 ml of selenite cystine broth and
10 mi tetrathionate broth. The selenite cystine and tetrathionate broths were
incubated for 8 hours at 37°C and 42°C, respectively. After enrichment, 1 ml Of
selenite cystine broth and 1 ml of tetrathionate broth were transferred to separate
tubes containing 10 ml Of M-broth and incUbated for 18 hours at 42°C. ‘ After
incubation, the M-broth enrichments were mixed and 1 ml Of each was
transferred into a hermetically sealed tube. The tubes were heated for 15
minutes in a water bath at 100°C. After cooling to room temperature, the
samples were screened on the mini-VIDAS system (bioMerieux, St Louis, MO).
Positive samples were biochemically confirmed by the Food Microbiology Lab of

Michigan State University (Ryser, 2000).

63

VIDASTM E. coli O157:H7

Recovery of E. coli 0157:H7 was done using the VIDASTM ECO system
(bioMerieux, 1998). One ml of each sample was added to 9 ml m-TSB with
novo-biocin, then incubated at 41°C for 6 hours. One ml of the enriched culture
was then transferred into 9 ml MacConkey broth with cefixime and potassium
tellurite (CT-Mac) (Mast Diagnostics, Merseyside, UK), and incubated for 18
hours at 37°C. After incubation, the CT-Mac solution was mixed, and 1 ml of
each suspension transferred into a hermetically sealed tube. The tube was
heated for 15 min in a 100°C water bath, and allowed to cool. The samples were
then screened for E. coliO157zH7 with the VIDASTM ECO assay (bioMerieux,
1998)
SPR Biosensor Salmonella Assays

The same enriched samples from the VIDAS immunoassay were used for
SPR analysis. The SPR biosensor was prepared with appropriate antibody, and
immersed in sterile BPW for two minutes as a zero control, then into sterile M-
broth for two minutes to identify the background response (Table 2.5). The
inoculated M-broth sample was then assayed for 3 minutes, followed by a pure
cUlture Of S. Typhimurium in M-broth (the positive control).

Table 2.5. SPR biosensor procedure for Salmonella spp. assays

 

 

Step Action Time (mirl
1 SPR sensor preparation (Table 2.1) 30
2 Sterile BPW (zero control) 2
3 Sterile M-broth (negative control) 2
4 Sample in M-broth 3
4 anti-Salmonella spp. antibody (300 ugll) 2
Total time: 39

 

64

Following the inoculated M-broth sample, the sensor was immersed in
anti-Salmonella spp. antibody for a second time, to achieve ‘sandwich' binding Of

antibody (Figure 2.3). Two identical assays were performed for each sample.

        
  

 
 
 

£2332, Antibody b
was

Bacterium

  

Bacterium -
was Antibody a

 

 

SPR biosensor gold-film

(A) (B)
Figure 2.3. (A) Antibody-antigen interaction on the SPR biosensor gold-film
surface using one antibody layer a. (B) Antibody-antigen interaction on the SPR
biosensor gold-film surface in a sandwich-type interaction, with two layers Of
antibody, a and b.

The sandwich binding arrangement was designed to increase the
molecular mass of the substrate bound to the sensing film of the SPR biosensor.
This was done to increase the magnitude of the Change in index Of refraction of
the biosensor.

SPR Biosensor E. coli O157:H7 Assays
E. coli 0157:H7 was assayed on the SPR biosensor following the same

procedure as the Salmonella spp. assays, using anti-E. coli O157:H7 antibody

and sterile CT-Mac broth in place of M-broth for the negative control (Table 2.6).

65

Table 2.6. SPR Biosensor E. coli O157:H7 Assay Procedure

 

 

Step Action Time (minL
1 SPR sensor preparation (Table 2.1) 30
2 Sterile BPW (zero control) 2
3 Sterile CT-Mac (negative control) 2
4 Sample in CT-Mac . 3
4 anti-E. coli O157:H7 antibody (300 ugll) 2
Total time: 39

 

The sandwich assay was performed in the same manner as for the
Salmonella spp. assays. Two replicates were performed on each sample.

Because initial testing indicated that the samples contained negligible
amounts Of E. coli O157:H7, five samples were inoculated with both 10 ul Of a
stock solution of 10° CFU/ml E. coli O157:H7 (about 10 CFU), and 10 ul Of pre-
enriched sample from the pork production facilities to create known positives.
The samples were prepared in sterile BPW, and enriched according to the

method used for VIDAS” ECO samples.

SPR Biosensor Positive Controls

Positive controls for the Salmonella spp. and E. coli O157:H7 SPR
biosensor assays were pure cultures Of S. Typhimurium in M-broth, and E. coli
0157:H7 in CT-Mac broth, respectively. Positive controls were assayed
following the SPR biosensor assay procedures used for the samples, with pure
culture used in place of the inoculated sample. Two positive controls each were
run for S. Typhimurium and E. coli O157:H7. Viable plate counts were performed

on the control cultures.

66

SPR Biosensor Assays in TSB

To find any difference in response of the SPR biosensor to samples grown
according the VIDAST" enrichment protocols, and samples grown in TSB, the
pork industry samples were inoculated in TSB and grown for 18 hours at 37°C.
The samples were then analyzed with the SPR biosensor following the same
procedure as the assays performed on the samples grown according to the
VIDAST" methods. Sterile TSB was used as the negative control, and pure 8.
Typhimurium and E. coli 0157:H7 cultures grown in sterile TSB were used as the
positive controls. A second rinse of antibody was performed for both anti-
Salmonella spp. and anti-E. coli 0157:H7 assays to achieve a ‘sandwich’-type
antibody binding.

E. coli O157:H7 positive samples were prepared in TSB in the same
manner as the E. coli O157:H7 positive samples were prepared for the VIDASTM
enrichment method. Five samples were inoculated with both 10 ul of a stock
solution Of 10° CFUlml E. coli O157:H7, and 10 ul Of pre-enriched sample from

the swine production facilities, to create known positives.

67

REFERENCES

BioMerieux, 1998. VIDAS E. coli 0157:H7 (ECO) Manual. biOMerieux, Inc., 595
Anglum Drive, Hazelwood, MO 63042-2320

DeMarco, D., Saaski, E., McCrae, D., Lim, D., 1999. Rapid Detection Of
Escherichia coli O157:H7 in Ground Beef Using a Fiber Optic Biosensor.
Journal of Food Protection, 62(7)711-716

Ryser, ET, 2000. Strain specific typing bacterial pathogens in the pork industry
chain. Final report for NPPC project #99-026

Texas Instruments, inc. Dallas, Texas, USA. 1999. Operation Manual, Spreeta
Experimenter’s Kit

Younts, SM, 1999. Development and Evaluation Of a Gas Sensor Based
Instrument for the Detection and Differentiation of E. coli O157:H7 from
Non-O157:H7 E. coli. Thesis for the degree of MS, Michigan State
University, East Lansing, MI. .

68

CHAPTER 3

SENSITIVITY AND SPECIFICITY ASSAYS FOR SALMONELLA TYPHIMURIUM AND
ESCHERICHIA COL/0157:H7 UTILIZING A SURFACE PLASMON RESONANCE BIOSENSOR

ABSTRACT

The Spreeta.TM SPR biosensor was used to detect two foodbome
pathogens; Salmonella enterica spp. Typhimurium and Escherichia coli O157:H7.
The selectivity of the SPR biosensor was assayed using a series of antibody
concentrations and dilution series Of the two organisms. Specificity of the SPR
biosensor was demonstrated in pure and mixed cultures of S. Typhimurium and
E. coli O157:H7.

The Optimum antibody concentration for use on the SPR biosensor was
300 uglml, a result comparable to other immunosensors. The detection limit of
the SPR biosensor for S. Typhimurium and E. coli O157:H7 in pure culture was
107 colony forming units (CFU)/ml. A pure culture of these pathogens could be
detected after 5 1l2 hours Of enrichment. The SPR biosensor was specific to S.
Typhimurium and E. coli 0157:H7 in pure cultures at concentrations Of 107
CFUlml.‘ in mixed cultures, the detection limit of the biosensor was 107 CFU/ml
for the target organism, when the non-target bacterial concentration was 10°
CFU/ml or less. Concentrations Of non-target bacteria beyond 107 CFU/ml
caused a decrease in the magnitude of the sensor response to the presence Of

the target pathogen.

69

The SPR biosensor has demonstrated potential for portable, field-based,
rapid, and accurate pathogen detection. The SPR biosensor is a versatile
instrument. By changing the antibody used in the preparation phase, it can be
used to detect many different substrates, including other foodbome pathogens.
WIth further studies and refinements, the SPR biosensor shows promise to
provide a complementary detection system to standard lab-based systems

currently used in food safety and control systems.

70

INTRODUCTION

Foodborne bacterial pathogens are believed to be the most frequently
occurring hazard Of the-nation’s food supply, causing billions of dollars to be lost
in medical costs, lost productivity, and product recalls associated with outbreaks
of foodbome illness (Hui et al., 1994). In recent years, the food processing
industry has been under increasing pressure to identify and control potential food
safety hazards caused by pathogenic bacteria. Although the Hazard Analysis
and Critical Control Point (HACCP) system has reduced the need for end-product
testing, the demand for rapid and accurate methods to detect foodbome
pathogens has increased (Seo et al., 1999).

Escherichia coli O157:H7 has emerged as an important enteric pathogen
Of considerable public health significance. Illnesses caused by E. coli O157:H7
can range from mild, watery diarrhea to life-threatening conditions, such as
hemolytic uremic syndrome and thrombotic thrombocytopenic purpura (Jay,
2000; Buchanan and Doyle, 1997). The combination of the severe
consequences of infection, its low infectious dose, and its association with many
common foods make E. coli O157:H7 a bacterial pathogen of particular concern.

Salmonella enterica has been identified as one Of the most prevalent and
costly Of known foodbome pathogens (Davies, 1997). it is a main cause of
documented foodborne illness in most developed countries (Buzby et al., 1996).
Adding to the concern aboutSalmone/la infection is the existence Of several
antibiotic resistant strains, especially the multi-drug resistant strain 8.

Typhimurium DT 104. With the emergence of this strain, the incidence and

71

severity Of Salmonella-related human illness is on the rise (Middleton et al.,
1999).

Conventional detection'methods, which can take 4 to 7 days to detect and
confirm pathogenic bacteria in food, are not acceptable for monitoring critical
control points (SeO et al., 1999). A biosensor that could detect pathogens within
minutes or hours would allow processors to take quick corrective action when
pathogens are detected. I

Biosensors are analytical instruments possessing a capturing molecule as
a reactive surface in close proximity to a transducer, which converts the binding
Of an analyte to the capturing molecule into a measurable signal (Fratamico et
al., 1998). The SPR biosensor is a biosensor that monitors antibody-antigen
interaction in real-time, utilizing surface plasmon resonance (SPR) (Elkind et al.,
1998). SPR is an Optical phenomenon that occurs as a result Of total internal
reflection of light at a metal film-liquid interface. A component Of the incident light
momentum, the evanescent wave, interacts with surface plasmons (free
oscillating electrons) in a thin metal film. When SPR occurs, energy from the
incident light is lost to the metal film, resulting in a decrease in reflected light
intensity. The resonance phenomenon occurs only at a precisely defined angle
of incidence, which is dependent on the refractive index Of the medium adjacent
to the metal surface. The refractive index changes in direct proportion to the
mass and the make-up of the media present. When antibodies are affixed to the
metal surface, the angle of incidence that causes SPR depends on the amount of

antibody-antigen substrate present. By using antibodies specific to pathogens Of

72

interest, it is possible to utilize the SPR phenomenon to measure the amount of
pathogenic bacteria present in a sample by measuring the change in refractive
index.

The objective of this study was to determine the sensitivity and specificity
Of the SPR biosensor to S. Typhimurium and E. coli 0157:H7. in order to explore
the feasibility of employing an SPR biosensor as a pathogen monitoring

technique in the food industry.

MATERIALS AND METHODS

Spreeta SPR Biosensor
The Spreeta Miniature Integrated Surface Plasmon Resonance Liquid
Sensing System (Texas Instruments, Inc.) was used in this study (Figure 3.1).

Reflectin . .
..'”’MM Mirror 9 Optical Plastic

Substrate

   

Surface Plasmon
Layer

  

 

I ”I; #5» I cm
\ i
\, 3 ,
Light Emitting Tem eratur ‘
Diode Polarizer ssnsor ° Photodiode
Array

Figure 31. Side view of the Spreeta Sensor (courtesy of Texas Instruments.
1999, URL: httpzllwwwticomlspreeta)

Near—infrared light (840nm) from a light emitting diode (LED) IS polarized
to enhance SPR. The light beam reflects off of the gold sensing film and is
directed by the gold mirror onto a linear array of Silicon photodiodes, The active

sensing region Is the area on the thin gold film that is actively used In liqUId

73

sensing. Except for the sensing surface, the sensor is coated with an Opaque
material to block out external light (Texas Instruments, 1999).

The angle at which a ray of light is incident upon the sensing film is
directly mapped into a specific point on the photodiode array. The SpreetaTM
SPR biosensor is capable of measuring the index of refraction within the range of
approximately 1.29 and 1.42 (Elkind et al., 1998). The computational software
records the index of refraction of each sample approximately every 4.8 seconds.
Depending on the refractive index of the liquid next to the gold film, at some
angle the reflected light intensity experiences a minimum corresponding to where
SPR occurs. This minimum is then processed by the Spreeta computational
software, and plotted as the change in index Of refraction over time (Texas
Instruments, 1999).

Preparation and Biotinylation of Antibodies

Lyophilized affinity purified antibody to Salmonella common structural
antigens (CSA-1)Land to Escherichia coli 0157:H7 were purchased (Kirkegaard
& Perry Laboratories lnc., Gaithersburg, MD). All antibodies were stored at 4°C
until rehydrated. Antibodies were rehydrated according to the manufacturer’s
inStructiOns (Kirkegaard & Perry, lnc.). Biotin labeling was done according to the
method of DeMarco et al. (1999), and the biotinylated antibody was diluted with

sterile phosphate buffered saline (PBS) (pH 7.2).

74

Biosensor Preparation and Antibody Attachment

The total time to prepare the SPR biosensor for an assay was 30 min.
The SPR biosensor was assembled according to the Spreeta Operation Manual
(Texas Instruments, 1999). Neutravidin binding and attachment of biotinylated
antibodies to the surface was done following the method of Spreeta’s Application
Brief 004 (Texas instruments, 1999).
Bacteria

Characterized strains of S. Typhimurium and E. coli 0157:H7 were
Obtained from Michigan State University collections. E. coli O157:H7 was
verified in a previous study by the Bacteriology Laboratory at the Veterinary
Diagnostic Center of the University of Nebraska, Lincoln (Younts, 1999). All
culturing was done in a certified Biological Safety Level 2 environment at the
Michigan State University (MSU) Meat Microbiology Laboratory. Cultures were
maintained in tryptic soy broth (TSB) (Becton Dickinson) at 4°C. Cultures for
viable cell counts and SPR assays were grown in TSB at 37°C for 18-20 hours.
Cultures were serially diluted in sterile buffered peptone water (BPW) for use in
the experiments.
SPR Biosensor Signal Normalization

The output of the biosensor was the index of refraction at which surface
plasmon resonance occurred. TO normalize the output, the index Of refraction for
the zero controls was subtracted from the responses. The resulting biosensor
output was then the magnitude Of the change in index of refraction (the increase

in index of refraction above the index of refraction Of the zero control).

75

Comparisons of Antibody Concentration
Anti-Salmonella and anti-E. coli O157:H7 biotinylated antibodies Were prepared
at 3 concentrations in sterile PBS; 300 uglmi, 30 ug/ml, and 3 ug/ml. To analyze
the difference in response Of the SPR biosensor to differing antibody
concentrations, 3 ten-fold dilution series of S. Typhimurium and E. coli 0157:H7
were made in sterile BPW, and run on each of the 3 antibody concentrations.
Viable plate counts of the serial dilutions were performed to verify the
concentration Of the organisms. Results were compared to see if there was any
difference in response of the SPR biosensor to increasing concentrations of
anfibody.
Background response.

The response of the SPR biosensor to assays performed without the use
Of an antibody was monitored to establish the background response Of the
system. Two separate experimental runs were performed on the sensor using
serial dilutions of S. Typhimurium, E. coli 0157:H7, and uninocuiated TSB.
Viable plate counts of the serial dilutions were performed tO verify the
concentration of the organisms. The SPR biosensor was prepared as for a
normal assay, but with sterile PBS introduced in place of the antibody. For each
bacterium, ten-fold dilutions of 10° CFU/ml cultures grown in TSB were prepared
in sterile BPW. For the TSB, serial dilutions were made in sterile BPW, and the
series run as if they were inoculated samples. Sterile uninocuiated BPW was

used as the zero control. The results of the assays were analyzed to see if there

76

was any significant difference in response Of the SPR biosensor between
inoculated and sterile TSB.
Negative Controls

The negative control used for the serial dilutions of S. Typhimurium and E.
coli 0157:H7 was TSB diluted in BPW. From a trial experiment conducted
earlier, the SPR biosensor background signal showed that the biosensor was
sensitive to the presence of sterile TSB. Since the organism was grown in TSB,
this sensitivity could potentially be misleading, so assays Of sterile TSB on
prepared biosensors were performed to find the contribution of TSB in. the
sample to the SPR response. Six 10-fold serial dilutions of TSB in sterile BPW
were performed, 3 assays using anti-Salmonella spp. antibody, and 3 with anti-E.
coli O157:H7 antibody. The results were analyzed to find the amount of SPR
biosensor response that was due to TSB. Results were also compared to the
background signal data (assays perfOrmed without antibody), to see if there was
a statistical difference between the background response and an antibody-
prepared SPR biosensor-response to sterile TSB. The SPR biosensor response
to sterile TSB was then used as the negative control, in order to account for the
TSB effect on the biosensor response.
SPR Biosensor Sensitivity Assays

SPR biosensor sensitivity assays were performed on a series of dilutions
of S. Typhimurium and E. coli O157:H7 in sterile BPW, from cultures inoculated

to approximately 10° CFU/ml (Table 3.1). Three assays were performed for each

77

organism. Viable plate counts of the serial dilutions were performed to verify the

concentration of the organisms.

Table 3.1. A typical dilution series experiment. The dilution series consists of
SPR analysis of 10 varying bacteria concentrations (CFU/ml).

 

 

Step Action Time (min)

1 SPR sensor preparation (Table 2.1, Chapter 2) 30
2 Immerse sensor in sterile BPW (zero control) 2
3 Immerse sensor in sample 2
a. repeat for 9 additional bacterial concentrations 18

4 Rinse gold surface with PBS 1
Total time: 53
Time for one biosensor assay: 35

 

Assays were prepared on each biosensor with the lowest concentration
(10° CFUlml) first, and the highest concentration (10° CFUlml) last. The index of
refraction for each dilution was computed as the average of the readings taken
over each 2 min period. The total time for a dilution series experiment, with 10
samples Of increasing concentration (10° — 10° CFU/ml) was 53 min. The time
for one SPR biosensor sample assay (procedure without step 3a) was 35 min.
Growth Curves

Growth Of S. Typhimurium and E. coli 0157:H7 was monitored by
performing a series of three growth curves for each bacterium. In each growth
curve, 10 ul of 10° CFU/ml stock solution of S. Typhimurium or E. coli O157:H7
(about 10 CFU) were introduced to 100 ml of sterile TSB. The inoculated
samples were incubated at 37°C. At half hour intervals for ten hours, aliquots Of

1 ml were removed from the vials, serially diluted, and viable plate counts

78

performed. The results from the plate counts were plotted over time to discover
the growth curves for each bacterium.
Specificity Assays

Specificity assays were done to establish the ability of the SPR biosensor
to select for the target antigen. For each bacteria, the SPR biosensor was
prepared with the respective antibody, then exposed to increasing concentrations
of three sample dilutants:

1. Sterile TSB (negative control).

2. Bacteria that the antibody was not specific for:

2.1. E. coli 0157:H7 dilutions for anti-Salmonella spp. prepared
SPR biosensor.

2.2 S. Typhimurium dilutions for anti-E. coli 0157:H7 prepared SPR
biosensor.

3. The bacteria of interest:

31 E coli 0157: H7 dilutions for anti-E. coli 0157: H7 prepared
SPR biosensor.

3.2 S. Typhimurium dilutions for anti-Salmonella spp. prepared
SPR biosensor.

Viable plate counts of the serial dilutions were performed to verify the
cOncentration of the organisms. Three replications were done of each dilution
series for 'both anti-Salmonella spp.—prepared SPR biosensor and anti-E. colI

0157:H7-prepared SPR biosensor, making a set of 18 assays.

79

Specificity in Mixed Cultures

Sample cultures composed of mixtures Of S. Typhimurium and E. coli
0157:H7 were analyzed on the SPR biosensor to establish the ability of the SPR
biosensor to select for the target antigen in the presence Of other bacteria. Five
cultures of target bacteria at concentrations 0, 10°, 10°, 104, 10°, 10°, 107, and
10° CFU/ml were prepared, respectively. Non-target bacteria at concentrations
of o, 102, 104,106, and 108 CFUImI were added to the five initial cultures at each
target bacteria concentration. The sample cultures were prepared from stock
solutions of known concentration, on which viable plate counts were performed to
verify the concentration Of the organisms.
Statistical Analysis

Statistical analysis was conducted using a Single factor ANOVA (analysis
of variance) function on experiment replications. For all studies, the SPR
biosensor average responses were considered statistically different when the P-

value was less than 0.05 (95% confidence level).

RESULTS AND DISCUSSION

The SPR biosensor was both sensitive and specific to the presence of S.
Typhimurium and E. coli 0157:H7 in solution. Although-the SpreetaTM SPR
biosensor has not been used previously as an immunosensor for pathogen
detection, initial results show promise for its application for food safety
monitoring. The first goal of this study was to establish the optimum antibody
concentration for preparation of the biosensor. Anti-Salmonella and anti-E. coli

0157:H7 biotinylated antibodies at concentrations of 300 ug/ml, 30 ug/ml, and 3

80

ug/ml in sterile PBS were assayed (Figures 3.2 and 3.3). The responses Of the
SPR biosensor to each of the three antibody concentrations were linear, with R2
values of 0.998 to 0.999. The magnitude of the biosensor response increased as
the antibody concentration increased. ANOVA revealed that the SPR biOsensor
responses to the different antibody concentrations were significantly different at
bacterial concentrations Of 107 CFU/ml and above for both S. Typhimurium and
E. coli 0157:H7. Because the magnitude of the sensor response for an antibody
concentration of 300 uglml was the most significant, it was chosen as the
concentration to be used in all further studies. This was done in an effort tO
increase the sensitivity of the biosensor, by utilizing an antibody concentration
that returned the largest magnitude of response. This antibody concentration is
comparable to concentrations used in other immunosensors, which typically
range from 50-400 uglml (Elkind et al., 1998; Woodbury et al., 1998; Watts et al.,
1994; Faegerstam et al., 1992), and is. the same concentration used for the
detection Of E. coli 0157:H7 by the BlAcoreT" SPR biosensor (Fratamico et al.,
1998)

Without the attachment of antibodies, the SPR biosensor responds to the
incidental contact with any proteins present in solution (Elkind et al., 1998).
Therefore, the SPR biosensor without the attached antibody Should have a
similar response for samples in TSB with or without bacteria present, as the
bacteria would not attach to the sensing surface without the antibody. In fact, the
responses of the biosensor to dilutions of S. Typhimurium, E. coli 0157:H7, and

sterile TSB were not statistically different at any dilution level (Figure 3.4). The

81

 

_

_

.0308 new .maom .m: n O cozmzcoocoo
Congo dam m‘LocoEEmrscm 2 .8583 mew o5 06 8.69.8 mmmco>< mm 8:9...

.338. 5.3.588 8..
8+3 8+3 8+3 8+3 8+3 8+3 8+3 8+3 6+3 8+3 53

 

 

 

aeovcv
65:8
o>zmmoc
:8. E286
2.8885

 

 

_

 

 

 

60.0 A .0
EEOC
o>=maoc
can. EEOEO
2.88.55
62

 

 

 

888 I. am Ea: 8m IT
mmmmd .I. mm :50: om Isl

888 I am Ea: mil

. 0

. 000..

I 000m

.- 000m

L ooov

fl 0000

 

0000

uogoeueu to xepuI u! efiueuo

82

.msoom Dam .maom .0:
m 00 cozgcoocoo Seance 2.3.90 .66 .m-=cm O. 62085 mew OE 00 $538 oam6>< .m.m 8:9...

..ESEO. coach—Loosen. mo.—

 

 

 

 

 

 

 

 

 

 

 

00+m4 . 00+w... v0+m._. N0+me 00+m... N00...
. _ - . . O
.. \
- 000.. O
k u.
. /<\l .m
8.0 A a: - 88 m.
_O._EOO W
.\, o>=mmoc W
:9: E8226 , 008 m
2.888%. m.
. I E s 8
88 v av 62 88 o I am _ an 8min: 89 m.
.088 . 888 I NE _E\ s on If m
o>=mmoc _ mmmmd u um EB: m I W.
:9: E836 u
2.8885 . 88

 

 

 

 

 

 

0000

83

succeed .85.; m9 2.8.» new 5+6 6 ...8 .m

.EESEEQE. .0 Lo mcozazu fired 9 .8533 mam Co concedes. 9.39962 mmEO>< in 8:9“.

Econ no... .6025
com: EIme Nome mom; vows mom: Sue Rom; mom: meme

bI _ p L

b _ p P

 

02 x as ESSESES 8.... 5.66 ..8 .m
...2 x 3 8.8.588 3...... EsEEEefi .9

 

 

30.0 A n: EEOC 9.609. cmE
E286 2.8.8.5.... 62 u a.

   

888 u N”. mme e....sw mohair?
888 u am EH55 ..8 m 899T?
38.0 n ma Eurasia: .w 3805+

 

_.

O
‘—

 

.. 000—.

 

 

0000 F

manager; in xepui u! afiueqo 601

84

similarity Of the linear increase Of the SPR biosensor to increasing concentrations
of samples indicated that it was the concentration of the background matrix Of
TSB that caused the change in index Of refraction, and not the presence Of
bacteria.

Because sterile TSB had an effect on the response of the SPR biosensor,
assays were performed with TSB dilution series on biosensors prepared
following the same method as for bacterial assays with antibody attachment.
Three assays were completed using anti-Salmonella antibody, and 3 with anti-E.
coli 0157:H7 antibody (Figure 3.5). ANOVA was performed on the background
signal and negative control data to assess the difference between sterile TSB
dilutions assayed with an SPR biosensor prepared without antibody (the
background signal) and SPR biosensor prepared with antibody. The response to
sterile TSB was not statistically different (P > 0.5) from the background response
of the biosensor at any dilution. The antibodies used in preparation of the SPR
biosensor did not respond to the presence of TSB in the samples, and again it
was the protein matrix of the TSB that the SPR biosensor sensed.

To determine the sensitivity of the SPR biosensor to S. Typhimurium and
E. coli 0157:H7 in solution, dilution series were assayed. Figures 3.6 and 3.7
Show the average SPR response for S. Typhimurium and E. coli 0157:H7 for
dilutions between 10° and 108 CFUImI. The limit of detection of the SPR
biosensor was taken to be the bacterial concentration at which there was a
statistically significant variance in mean of the response of the SPR biosensor to

the target organism versus a negative control. ANOVA revealed that the level Of

85

00+m. _.

8.88% $65 ...8 made as... dam 8888.28
9.7.: mm» 2.57. .6 96.5.6 .88 o. emcoammm - Lomcmmofi mam Lo .988 9:89. ounce}. .m.m 8:9”.

030w 00.. Aim c. 032:. 60320» mm... 300—. .535 E95 c0325 mm...

P00... Norm... norm... worm... 00¢; worm... norm... worm... morwé

» b b p — b L h

 

 

 

60.0 A A: 65:8 0268: :9:
E288 2.8.3.63 .02 u e..

   

888 u A”. $85 ..8 weenie.
88.0 ... am Es..sEEa¢ 8-..ch

 

 

 

 

r

O
‘—

. 000—.

0000 w

uogoeuea to xepuI ur afiueuo 601

86

.EESEEOC .0 L2 8888 .8883 mam omEo>< .0.m 830....

 

.858. 8.22.888 8..
mom: 00m: 3+0... 00m: mom: 3+0; m0+me mom: 5+0... 00+m._. 5m: mom:
P p ) _ _ p b I o
m m m m m m
. 009
. 000w
m 60.0 v A: 6.2.8 9.88: can. E8920 2.8.8250 u n... .. coon
60.0 A a: .2E8 9.80m: 85 E8606 2.8.8.68 .02 n m...
.ooov
88.0 u Am. .2080 9882 IT
0000.0 .I. am 825E203. .0 I w 0000

 

 

 

0000.

uogoeuau ,jo xepuI u! efiueqa

87

KIMBFO ...8 .w .o. 8:098. .8535 mam m0m.m>< Sm.” 930.“.

.533. 555850 5..

mo+m._. mo+m._. New... 09¢... m0+mé 3+“? mo+wé No+m._. ..0+w._. 00+m._. ..0-m.F mom...

 

 

 

_ p . p p F u _ K &

    

.8... v a. 558 9,55: 55 55.5 25.55% u a.
.80 A n: 6.58 955.. 55 55.5 2.8.5.55 .oz ... 9

00000 H mm 3:50 $5.32 Ial
00000 n Na nIKmFO ._8 .m Iol

 

0

000 _.

.- OOON

. 88

. 000v

uogaoeuaa 4o xapul u! afiuaqg

000m

 

 

0000

88

00+w. _.

02.0w co.......0 E:..:E.:a>... .0 .0 5.00. .09.... .00 0.30....
255.0. 5.5.2.850 5..
00+m. F ~0+we 00+m. F

 

 

_ p F

0..

1 Dow

 

knmmd H mm— mmm..m><lxl y 000—.
m €252.39... .w 4
N 82.38.20: .0 a
F 62.35.29... .0 o

 

 

0000 F

uogmeuaa go xepul us efiueuo 601

89

00+m. _.

.850 5:25 2.55 ...8 .m .0 5.5m .55.. an 25..

255.0. 55.5850 5..
mom... 8+3

_ _

00+m...

 

 

 

 

0500.0 u um 000.05 le
m 0.305 ..00 .w a
N $330 ..00 .w. n
F 51505 ..00 .m o

_.

-0_.

.. 00F

- 000w

 

 

0000 _.

uogweuaa 4o xapuI ug efiueua 601

90

detection of the SPR biosensor was 9.1 x 106 CFUImI for S. Typhimurium, and
8.7 x 106 for E. coli O157:H7. These limits were rounded to the nearest power,
giving a limit of detection for both 8. Typhimurium and E. coli O157:H7 of 107
CFU/ml.

Figures 3.8 and 3.9 show the linear correlation between the biosensor
response and the bacterial concentration. An R2 of 0.998 for S. Typhimurium,
and an R2 of 0.997 for E. coli 0157:H7 in the region above the limit of detection
indicates the ability of the SPR biosensor to predict the concentration of the
target organisms.

A limit of detection of 1 x 107 CFU/ml is similar to the detection limits of
other SPR biosensors in the literature. Fratamico et al. (1998) reported, a
detection limit of 1.7-2.1 x 106 CFUImI using the BlAcoreT“ SPR biosensor for E.
coli O157:H7 detection. Seo et al. (1999) reported a detection limit in the range
of 1 x 105 to 1 x 107 CFU/ml for SalmdneI/a spp., using an SPR biosensor they
constructed. A resonant mirror sensor, similar in concept to the SPR biosensor,
had a detection limit of 8 x 106-8 x 107 CFUImI (Watts et al., 1994). The ELISA
system, another popular immunosensor, has a detection limit of 105-106 CF U/ml
(Seo et al., 1999), and the VIDASTM immunoassay has a detection limit of 105
CFU/ml (Cohen and Kerdahi, 1996). Although several of these other
immunosensors report lower detection limits, a limit of 1 x 107 CFU/ml is within
the range of currently accepted methods, especially for other sensors based on
the SPR principle (Elkind, 1998). Additionally, because this was the first study

done using the Spreeta SPR biosensor as an immunosensor for pathogen

91

detection, further experimentation for the refinement of the initiation and sampling
protocols could potentially lead to a significant increase in selectivity.

However, it may be argued that there is no need to increase the selectivity
of the SPR biosensor. Representative growth curves of S. Typhimurium and E.
coli 0157:H7 (Figure 3.10) show that concentrations of S. Typhimurium and E.
coli O157:H7 reach the detectable level of the SPR instrument (1O7 CFUImI)
roughly 5 V2 hours after inoculation. This is much shorter than the time required
by many sensing methods that involve pre-enrichment, such as standard plate
counts. Currently, other biosensing methods such as VIDAS” immunoassays
and the ELISA protocol require pre—enrichments that take 18-32 hours (Grif and
Allerberger, 1998; bioMerieux, 1998). Standard plating methods, moreover,
often take 4-7 days for pathogenic detection (Jay, 2000; Seo et al., 1999). If the
SPR biosensor can detect. pathogens in samples incubated for 5 ‘/2 hours, it is
significantly more rapid than these methods.

Specificity assays were conducted to determine whether the SPR
biosensor could detect the target organism in samples that have not undergone
extensive pre-enrichment. Assays of pure cultures of target and non-target
bacteria were first conducted to determine the overall specificity of the SPR
biosensor. 'The average magnitude of response of a SPR biosensor assay on
target bacteria was greater than the magnitude of response to a pure culture of
non-target bacteria. (Figures 311-312). In fact, the response of the SPR
biosensor to pure cultures of non-target bacteria was not significantly different

from the SPR biosensor response to the negative control at all dilution levels.

92

$5.0 ...8 .m .05 22.35.53 .0 .o 55.0 5...... .2... 9a....
3.30.... 05F

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0_. m m N m w v m N _. 0
Trill. . . ..- . _ . 8&3
.00. 000.? ll
._ m. < 5&3
53.56.50»... .0 000.03.. ll
m__oo.m o . N0+m0._.
N__oo. x
m T mo+mo._.
—. =00 .m x
mE_mw 4 3+mo...
«6.00 ...
. m0+w0._.
30.00 0
. % ®0+w0._.
5.5.85.5... - 8&3
. 00+w0..
mo+w0._.
.

 

 

 

 

 

 

 

 

 

0r+m0._.

(um/nae) uouenueouoo 601
93

.E:..:E.ca>._. .0 .0. 3.00.0000 000.03.. ...fim 0.30.“.

253.0. 55.2850 5..

00+mé 00+m... 50.0w... 00+m._. 00+m._. v0+m._. 09m... N0+m._. P0+wé 00+m... 50... Nolme

—

p

_ P — — _ _ b h

w

 

 

.000 v .. .950 02.009. 5... 55...... 2.8.05.0 n P
.000 A a. .258 2.65: 5... 55...“. 2.8.35.0 52 u m.

   

008.0 u «m. 52.35.53 .0 LT
080.0 u .0 007?
38.0 n «Ergflo ..8 m not

 

-0?

,, 00..

000 F

 

0000..

 

uonoeuaa toxapu' u! afiueuo 60']

94

5.5.0 ...8 .m 5. 0.2.080 85:. 0.... 28..
083.0. co..0....00:00 00..

00+w... 00+m... ~0+wé 00+m... 00+m... v0+m... 00+m... N0+m.w ..0+m... 00+m... ..0.m_... N00,?

.

p

. b _ _ p . p »

 

 

 

.000 v a. .o....8 9.58.. 5... 55...“. 2.8.05.0 ... ....
.000 A a. .o....8 858.. 5... 55...... 2.8.85.0 52 u 0..

  

«08... u a”. ..8 .01?

88... n .0 007...:
0000.0 u «m. 0:05.500 IT

 

 

..

0..

000 _.

0000 ..

uouoeuaa ;0 xepU| u! eBueuo 601

95

Conversely, for both anti-Salmonella and anti E. coli O157:H7-prepared
biosensors, the response of the SPR biosensor to target bacteria was
significantly different than the negative control at concentrations of 107 CFU/ml
and above. This result verifies that the detection limit of the biosensor was 107
CFUImI, and indicates that the antibody-prepared biosensor was specific to the
target of interest, and not cross-reactive to the non-target bacteria. This may be
because S. Typhimurium and E. coli 0157:H7 are not closely related, and have
low numbers of similar proteins. Their respective antibodies, therefore, would
display little cross-reactivity to the other pathogen.

To assess the specificity of the SPR biosensor in mixed cultures, several
assays were conducted on dilution series with varying concentrations of non-
target bacteria (Figures 3.13-3.14). The magnitude of the SPR biosensor
response increased as the concentration of target bacteria in solution increased.
For Salmonella spp. specificity assays, 'at a S. Typhimurium concentration of 107
CFUImI, the SPR biosensor response to samples with E. coli O157:H7
concentrations of 107 CFUImI and above was statistically the same as the
response to the negative control. At S. Typhimurium concentrations 0f 108
CFUImI, adding 108 CFUImI of E. coli O157:H7 caused the biosensor response
to be statistically the same as the negative control (Table 3.2). Likewise, for E.
coli O157:H7 specificity assays, ANOVA showed that at an E. coli-O157:H7
concentration of 107 CFU/ml, the SPR biosensor response to samples with S.
Typhimurium concentrations of 107 CFU/ml and above was statistically the same

as the response to the negative control. At E. coli O157:H7 concentrations of 108

96

0.3.30 00...... 0. 800000... «.00 00.0090 .. .000 0..000E.00-..00 .0. 2.00.0000 . 0 F .0 0.30. “.

00+m€

 

 

 

Geoxa.
.0..000
0>=0000
5... .555
0.8.55.»
.02

 

 

 

 

 

.000 X...
.0..000
02.0000
5... .550...
0.8.55.0

 

 

 

 

0.000 00... ...—.500. 00000000000 0.3.5.0203. .0
005.... 3+0... 00$:

 

00+”? v0+m€ 00+w...

_

N0+w... F0+me

 

 

.lll

 

8.2.30 0.30. cm. 0.3..3E...0>._. .0 LII

\ N0+m00. F Ill
00+w00. P IT
v0+w00. .. IOI

 

 

00.3.0.
55.0.8.8 0.00.0 ...8 .m

A 00+000. . I..T
00.500; IXI

 

 

 

0.0000. F Lcl
f 00+m.00. F lll
.2050 02.0002 IOI

 

 

..

00+m...

 

 

 

0

. 000w

- 000w

. 0000

- 000v

. 0000

0000.

uouoeuau ;o xepm ug efiueua

97

0.3..30 00x.0. 0. 5000.005 ”.00 02000.0 - 0.390 ...00 M000 .0. 3.0.0.0000 .30 0.00.”.

00+m. _.

 

 

 

GQOAa.
.00000
0>=0000
:0... 090......
2.8.0.0.0
.oz

 

 

 

 

 

8Q0v0.
.0008
0>=0000
cm... .0200...
2.88.0.0

A

 

 

 

 

 

00+m. ..

_

50.7w. P

0.80 03 055.0. 5:20.880 0.0.20 .3... .0

00+m.—.

» —

00+“? 0.0+m._. 00+”?

 

 

0

 

 

 

96500. .
00000000000 003038.00»... .0 A

 

 

 

F

 

N0+m. _.

F0+w._.

 

a lrliil. I! x

3.3030 0.30.0 00.. 5...—Km ..0 =00 .m III

~0+M00. .. III:
00+m00. F |.TI
3+m00. p IQI
m0+w00. _. Ill
00+m00. w [XI
~0+m00. F Idl
00+M00. F lfll

.0008 02.0002 IOI

 

00+w...

o
89
3
w
7 080 .w
a
......
m.
p
. 080 m
0..
U
m,
88 m
0
U
0000

 

0000

 

98

Table 3.2. ANOVA results for S. Typhimurium specificity in mixed cultures. P values.

.00!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O157:H7
(CFU/lnl) S Typhimurium
1048 10‘7 10*6 10‘5 10*4 10*3 10‘2 10‘8
8. (Pure Culture)
Typhimurium '
(CFUImI)
0 0.650 0.045 0.135 0.362 0.247- 0.076 0.214 0.305
1 .1OE+01 0.077 0.702 0.206 0.292 0.321 0.386 0.165 0.118
1.10E+02 0.149 0.194 0.390 0.137 0.148 0.086 0.170 0.057
1 .10E+03 0.161 0.454 0.481 0.094 0.067 0.092 0.188 0.094
1 .1OE+04 0.353 0.087 0.157 0.053 0.170 0.108 0.092 0.442
1 .10E+05 0.089 0.208 0.362 0.119 0.082 0.053 0.334 0.188
1 .10E+06 0.445 0.257 0.157 0.257 0.249 0.327 0.239 0.125
1 .10E+07 0.099 0.046 0.043 0.038 0.035 0.020 0.014 0.012
1 .1 OE+08 0.189 0.108 0.021 0.018 0.014 0.009 0.006 0.001

 

 

Table 3.3. ANOVA results for Ecoli O157:H7 specificity in mixed cultures. P values

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Typhimurium
(CFUImI) E. coli O157:H7
10"8 10‘7 10*6 10‘5 10"4 10‘3 10‘2 1.005+8
E. coli (Pure Culture)
O157:H7
(CPU/ml)
0.00E+00 0.201 0.072 0.078 0.048 0.259 0.099 0.194 0.1 12
1 .1OE+01 0.829 0.121 0.153 0.180 0.269 0.445 0.304 0.465
1 .10E+02 0.350 0.203 0.264 0.431 0.321 0.047 0.009 0.351
, 1 .10E+03 0.992 0.443 0.398 0.365 0.922 0.088 0.802 0.219
1.1OE+04 0.321 0.149 0.418 0.391 0.387 1.000 0.181 0.216
1.10E+05 0.165 0.780 0.066 0.175 0.201 0.171 . 0.353 0.248
1 .10E+06 0.566 0.722 0.062 0.326 0.609 0.117 0.092 0.113
1 .10E+07 0.337 0.249 0.042 0.016 0.016 0.023 0.006 0.005
1 .10E+08 0.059 0.030 0.009 0.007 0.009 0.013 0.004 0.002

 

Statistically significant response (P < 0.05 )

99

 

 

CFUImI, adding 108 CFUImI of S. Typhimurium caused the biosensor response
to be statistically the same as the negative control (Table 3.3).

These results indicate that the limit of detection of the SPR biosensor can
be affected by the presence of non-target bacteria in the sample. The presenCe
of 107 CFUImI or more of non-target bacteria in the mixed samples caused the
SPR biosensor to undergo a reduction in selectivity. However, up to levels of 106
CF Ulml of non-target bacteria in the sample, the SPR-biosensor remained able
to detect the target pathogen at 107 CF U/ml.

This reduction in sensitivity may be due to physical ‘crowding out’ of the
target bacteria in proportion to non-target bacteria, causing a low number of
target antigens to come into physical contact with the antibody. As the sensing
surface is immersed into the sample, the antibodies contact only those bacteria
closest to them. At higher numbers of non-target bacteria, the target bacteria are
spread more thinly through the interfacial layer of bacteria at the sensor surface.
This causes the antibody to come into contact with a lower number of bacteria,
possibly fewer than the detection level of 107 CFUImI. This would cause a
negative response, even though the total number of target bacteria in the sample

is above the limit of detection.

CONCLUSIONS

The SpreetaTM SPR biosensor used in this investigation is a new,
experimental product that has not before been tested or used for pathogen
detection. As this is the first investigation into the use of this sensor as an

immunosensor for the detection of S. Typhimurium and E. coli O157:H7, the

100

results of this study are promising. This investigation demonstrates that the SPR
biosensor has the potential for on-Iine, rapid, portable, and accurate pathogen
detection. The SPR biosensor is capable of reporting the presence of these
bacteria at levels of 107 CFUImI, and has shown specificity to the target
organism. With additional investigation and refinements, there may be many
uses for the SPR biosensor in food safety and control applications.

Another factor adding to the potential value of the SPR biosensor is its
versatility. By using antibodies specific to different biochemicals, the biosensor
can be made specific for many different target analytes. This expands the
potential applications of the SPR biosensor to detection of other forms of
pathogenic bacteria in foods, as well as raises the possibility of utilizing the SPR
biosensor in applications in fields such as medicine, biochemistry, environmental
treatment and remediation, and others.

Finally, the SPR biosensor could have a very positive economic benefit for
the food safety industry. According to the Food Safety Research Workgroup, 21
federal agencies are spending $200 million per year on food safety research.
State and industry officials match these funds, resulting in a total of at least $400
million spent on food research every year (Forsythe, 1996).

Many current commercial and experimental biosensors, e.g. the BlAcore,
IBIS, and Raptor biosensors, cost tens of thousands of dollars to purchase and
operate (Alocilja, 2000), while the SpreetaTM SPR biosensor used in this study
cost less than $3,000. Furthermore, commercial biosensors are bulky and lab-

based, whereas because of its small size, the Spreeta SPR biosensor can be

101

easily transported to the field. If this biosensor can be optimized for use in the
food safety industry, it will provide an affordable, portable, and field-based
alternative to other, more expensive and bulky biosensors. This could make food
safety inspection much cheaper and accessible, and make it easier for food
producers to implement HACCP protocols. USDA’s Economic Research Service
(ERS) estimates the cost of the HACCP to the meat and poultry industry
regulations to be $1.0 to $1.2 billion per year over the next twenty years (Roberts
et al., 1996). The SPR biosensor tested in this investigation shows the potential
to reduce this cost, by providing an economical alternative or complement to
current pathogen monitoring techniques that is sensitive, specific, accurate and

easy-to-use.

102

REFERENCES

Alocilja, E. C., 2000. Personal communications.

Buchanan, R.L., and Doyle, MP. 1997. Foodborne Disease Significance of
Escherichia coli O157:H7 and other Enterohemorrhagic E. coli.
’ F oodTechnology 51 :69-76

Buzby, J., Roberts, T., Lin, C.-T., MacDonald, J., 1996. Bacterial Foodborne
Disease: Medical Costs and Productivity Losses. USDA Agricultural
Economic Report No. 741

Cohen, A.E., Kerdahi, K.F., 1996. Evaluation of a rapid and automated enzyme-
linked fluorescent immunoassay for detecting Escherichia coli serogroup
0157 in cheese. Journal of AOAC International. 79(4)858—860.

Davies, P. 1997. Food safety and its impact on domestic and export markets.
Swine Health Product 5213—20

DeMarco, D., Saaski, E., McCrae, D., Lim, D., 1999. Rapid Detection of
Escherichia coli O157:H7 in Ground Beef Using a Fiber Optic Biosensor.
Journal of Food-Protection, Vol. 62, No. 7, 1999, Pages 711-716

Elkind, J.L., Stimpson, D.l., Strong, A., Bartholomew, D.U., Melendez, J.L. 1998.
Integrated Analytical Sensors: The Use of the TlSPR-1 as a Biosensor.
Sensor and Actuators, 11/98

Faegerstam, L.G., FrostelI-Karlsson, A., Karlsson, R., Persson, B., Roennberg, l.,
1992. Biospecific interaction analysis using surface plasmon resonance
detection applied to kinetic, binding site and concentration analysis.
Journal of Chromatography 597:397-410

Forsythe, RH. 1996. Food Safety: A Global Perspective. Poultry Science.
75:1448-1454

Fratamico, P.M., Strobaugh, T.P., Medina, M.B., Gehring, A.G. Detection of
Escherichia coli O157:H7 using a surface plasmon resonance biosensor,
Biotechnology Techniques, Vol. 12, No. 7. July, 1998. pp. 571-576

Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, eds. 1994. Foodborne
Disease Handbook, Volume 1: Diseases Caused by Bacteria. Marcel
Dekker, Inc. New York, NY

Jay, J. M, 2000. Modern Food Microbiology. (Sixth ed.) Aspen Publishers Inc.
Gaithersburg, MD

Middleton, D., Ciebin, B., Michel R, Ford, M. 1999. Salmonella Typimurium
definitive type 104 in Ontan'o,1997-1998 One example of an
antimicrobial resistant organism. Available at URL: http:l/www.gov.on.ca/
health/english/program/pubhealthlphero/phero_19991 1 .html

103

Roberts, T., Buzby, J.C., Ollinger, M., 1996. Using benefit and cost information
to evaluate a food safety regulation: HA CCP for meat and poultry. J. Agr.
Econ. 78:1297-1301

Seo, K. H., Brackett, R. E., Hartman, N. F. ”Campbell D. P., 1999. Development of
a rapid response biosensor for detection of Salmonella Typhimurium.
Journal of Food Protection. 62(5)431-437

Texas Instruments, Inc. Dallas, Texas, USA. 1999. Operation Manual, Spreeta
Experimenter’s Kit

Watts, H., Lowe, C., Pollard-Knight, D., 1994. Optical biosensor for monitoring
microbial cells. Anal. Chem 66:2465-2470

Woodbury, R.G., Wendin, C., Clendenning, J., Furlong, C.E., Melendez, J.,
Elkind, J., Bartholomew, D., Brown, S., 1998. Construction of biosensors
using a gold-binding polypeptide and a miniature integrated surface
plasmon resonance sensor. Biosensors and Bioelectronics (1998)00-00

Younts, SM, 1999. Development and Evaluation of a Gas Sensor Based
Instrument for the Detection and Differentiation of E. coli O157:H 7 from
Non-O157:H7 E. coli. Thesis for the degree of MS, Michigan State
University, East Lansing, MI.

104

CHAPTER 4

DETECTION OF SALMONELLA TYPHIMURIUM AND ESCHERICHIA COLI 01 57: H7 IN THE
PORK PRODUCTION CHAIN USING A SURFACE PLASMON RESONANCE BIOSENSOR AND A
VIDAsT" IMMUNOSENSOR '

ABSTRACT

A SpreetaTM SPR biosensor was evaluated as a diagnostic tool for
detecting Salmonella spp. and E. coli O157:H7 in samples Obtained from swine
farms and slaughterhouses. A VIDAS” immunosensor was used as a standard
for’comparing the accuracy Of results Obtained from the SPR biosensor. Field
samples were obtained from representative small and medium-sized pork
production facilities and analyzed using both. the SPR biosensor and the
VIDASTM immunosensor. Results Of this study indicate that with further
refinement the SPR biosensor could be useful as a pathogen monitoring system
within the pork industry Chain. The average SPR biosensor response was
greater in magnitude for S. Typhimurium- and E. coli O157:H7-positive samples
than the average response to negative samples. However, the difference was
statistically significant for only one sample. With further modifications of the
initiation and sampling protocols, the SPR biosensor has potential tO increase its
ability to detect target organisms in field-samples. The SPR biosensor shows
promise to provide a complementary detection system to standard lab-based

systems currently used in food safety.

105

INTRODUCTION

Salmonella enterica spp. Typhimurium and Escherichia coli O157:H7 are
foodbome pathogens Of major concern in pork production. In the United States,
foodbome diseases cause an estimated 76 million illnesses, 325 thousand
hospitalizations, and 5 thousand deaths annually, with an associated cost of
$56-$94 billion (Buzby eta/., 1996). Pork products are growing in popularity
among consumers Of meat, and continue to expand their share of local and
global markets (U.S. Meat Export Federation, 1997). Every time an outbreak
occurs in the pork product category, consumer health is put at risk, and the
viability Of the pork industry is threatened. .

SalmOne/la enterica has been identified as one Of the most prevalent and
costly of known foodbome pathogens (Davies, 1997), and currently 11% Of all
outbreaks of salmonellosis in humans have been associated with pork (lsaacson
et al., 2000). Adding to concern about Salmonella infection is the existence of
several antibiotic resistant strains, especially the multi-drug resistant strain 8.
Typhimurium DT 104. \NIth the emergence Of this strain, the incidence and
severity of Salmonella-related human illness is increasing. Salmonella is a
particular problem in the pork industry, because of the ability Of pigs to become
long-term carriers of the organism, with the potential of spreading the pathogen
to other hogs through a variety of transmission routes (Letellier et al., 1999).

Escherichia coli O157:H7 is a major cause Of serious outbreaks and

. sporadic cases of hemorrhagic colitis and hemolytic uremic syndrome (HUS)

106

(Grif and Allerberger, 1998). E. coli O157:H7 has been identified in many
commonly consumed foods, including pork and pork products (Jay, 2000). The
combination Of the severe consequences of infection, its low infectious dose, and
its association with many common foods make E. coli O157:H7 a bacterial
pathogen of particular concern (Verozny-Rozand et al., 1998).

Fecal shedding and growth of pathogens on the farm can lead to
subsequent contamination Of the slaughterhouse environment and finished
product. Infected animals and contaminated feed are supposedly the primary
routes for introducing these pathogens into herds (Ryser, 2000).

In an effort to reduce the occurrence and numbers of pathogens on meat
and poultry products, recent food safety initiatives have emphasized the
monitoring and control Of foodbome pathogens. The Hazard Analysis Critical
and Control Point (HACCP) system has reduced the need for end-period testing,
however the demand for rapid and accurate methods to detect foodbome
pathogens has increased (SeO et al., 1999).

Because Of the ‘zero-tolerance’ status of S Typhimurium and E. coli
O157:H7, testing must be performed that is accurate and sensitive at very low
levels (Jay, 2000). Classic approaches to microbiological quality control have
relied heavily on microbiological determinations of both raw materials and end
products, but the time required for results is too long for many products (Jay,
2000). There is a growing need in the food industry for pathogen detection
systems that are sensitive to low levels Of bacteria, specific to the target

organisms, inexpensive, and capable Of running at or near real-time.

107

The properties of surface plasmon resonance (SPR) are used in the SPR
biosensor to monitor antibody-antigen interaction in real-time (Elkind et al.,
1998). SPR, an optical technique for surface and interfacial studies, is widely
used in the biosensor, pharmaceutical, and analytical Chemistry communities
(Salamon et al., 1999). Surface plasmons in a thin gold film can be propagated
by incident light, which in turn excite an evanescent wave that can probe the
optical properties Of materials in direct contact with the sensing surface (Salamon
et al., 1999). When labeled with antibodies, the SPR biosensor can directly
analyze the binding of an antiserum to the immobilized ligand. The binding
interaction generates a signal resulting in changes in refractive index on the gold
surface and the matrix. This Change is proportional to the Change in adsorbed
mass (Medina, 1997). By using antibodies specific to pathogens of interest, it is
possible to utilize the SPR phenomenon to quantitate pathogenic bacteria in a
sample by measuring the change in refractive index,

Consumers are increasingly aware of the risk foodbome pathogens pose
to human health, and as such demand a safe, high quality, and nutritious food
supply. This means that having a reliable system for monitoring the quality and
safety of foods is of increasing importance. The SPR biosensor shows promise
to fulfill the food industry’s need for a fast, reliably sensitive screening method for
pathogen monitoring. The objective Of this research was to explore the feasibility
of employing a SPR biosensor to detect Salmonella Typhimurium and E. coli
O157:H7 in environmental and meat samples from representative pork packing

facilities. The SPR biosensor was compared to the VIDAS“ immunoassay as a

108

diagnostic tool for detecting Salmonella spp. and E. coli O157:H7 in samples

Obtained from the pork industry.

METHODS AND MATERIALS

Spreeta" SPR Biosensor

The SpreetaTM Miniature Integrated Surface Plasmon Resonance Liquid
Sensing System (Texas Instruments, Inc.) was used in this study. The output of
the sensor is the index Of refraction at which a ray of light incident upon the
sensing film experiences a minimum corresponding to where SPR occurs.
Preparation and Biotinylation of Antibodies

Lyophilized affinity purified antibody to Salmonella common structural
antigens (GSA-1) and to Escherichia coli O157:H7 were purchased (lGrkegaard
& Perry Laboratories Inc., Gaithersburg, MD). All antibodies were stored at 4°C
until rehydrated. Antibodies were rehydrated according to the manufacturer's
instructions (Kirkegaard 8. Perry Laboratories, Inc.) BiOtin labeling was done
according to the method of DeMarCO et al. (1999). The biotinylated antibOdy was
then serially diluted with sterile phosphate buffered saline (PBS; pH 7.2), to
achieve the Optimum concentration of 300 .4ng (Chapter 3).
Biosensor Preparation and Antibody Attachment

The SPR biosensor was assembled according to the Spreeta Operation
Manual. Neutravidin binding and attachment of biotinylated antibodies to the

surface was done following the method of Spreeta’s Application Brief 004, with

109

modifications (Texas lnstmments, 1999). The total time .to prepare the SPR
biosensor for an assay was 30 min (Chapter 3).
Bacteria

Characterized strains Of S. Typhimurium and E. coli O157:H7 were
obtained from Michigan State University collections. E. coli O157:H7 was
verified in an earlier study by the Bacteriology Laboratory at the Veterinary
Diagnostic Center Of the University of Nebraska, LincOIn (Younts, 1999). All
culturing was done in a certified Biological Safety 'Level 2 environment at the
Michigan State University Meat Microbiology Laboratory. Cultures were serially
diluted in sterile buffered peptone water (BPW) for use experiments.
SPR Biosensor Signal Normalization

The output of the biosensor was the index Of refraction at which surface
plasmon resonance occurred. To normalize the output, the index Of refraction for
the zero controls was subtracted from the response. The resulting biosensor
output was then the magnitude of the increase in index of refraction above the
index of refraction of the zero control (Chapter 3).
Field Samples

Samples were obtained from a previous study, in which site visits were
made to small-sized (~200 heads/week) and medium-sized (>1000 heads/week)
commercial meat packing facilities, as well as to the farms that supplied them
with hogs (Ryser, 2000). The samples were pre-enriched in lactose broth Within

24 hours, and stored at refrigeration temperatures. Twenty-four samples were

110

analyzed by mini-VIDAS and SPR biosensor for Salmonella spp. and E. coli
O157:H7 (Tables 4.1 and 4.2).

Table 4.1. Farm samples (Ryser, 2000)

 

 

Number
Of
Site Sample Method and Quantity samples .
Pen Alleyway Fecal material, Scoop, 509 1
floor
Hogs back Composite Swab with neutralizing buffer 2
solution

Fecal matter Composite Scoop, 50g 1
Feed Composite Scoop, 509 2
Water Nozzle Composite Swab from nozzle tip 1

 

Table 4.2. Environmental samples from slaughterhouse (Ryser, 2000)

 

 

Number
Method and Of
Site Sample Quantity sample's
Pen Alleyway Fecal material, floor Scoop, 509 1
Prechill Swab from chillroom wall Swab 7
Dehairing Hairs scooped from dehairing 259, hair scoop 6
machine machine
Drain Drain near evisceration Swab 3

 

Field Sample Analysis

Salmonella spp. and E. coli O157:H7 were assayed using both the mini-
VIDAST" system and the SPR biosensor:
VIDAS" Salmonella

Recovery of Salmonella spp. using the VIDAS” system was done
according to the BAM/AOAC method (bioMerieux, 1998). The samples were pre-
enriched in lactose broth, then 1 ml of the pre-enriched sample was incubated in
9 ml of buffered peptone water (BPW) for 18 hours at 37°C. Following this, 1 ml.

Of the BPW enrichment was transferred into 10 ml of selenite cystine broth and

111

10 ml tetrathionate broth. The selenite cystine and tetrathionate broths were
incubated for 8 hours at 37°C and 42°C, respectively. After enrichment, 1 ml of
selenite cystine broth and 1 ml Of tetrathionate broth were transferred to separate
tubes containing 10 ml Of M-broth and incubated 18 hours at 42°C. After
incubation, the M-broth enrichments were mixed and 1 ml Of each was .
transferred into a hermetically sealed tube. The tubes were heated for 15
minutes in a water bath at 100°C. After cooling to room temperature, the
samples were screened on the mini-VIDAS system (bioMerieux, St Louis, MO).
Positive samples were biochemically confirmed by the Food Microbiology Lab Of
Michigan State University (Ryser, 2000).
VIDAS" E. coli O157:H7

Recovery of E. coli O157:H7 was done using the VIDAS” ECO assay
(bioMerieux, 1998). One ml of each sample was added to 9 ml of m-TSB with
novo-biocin, then incubated at 41°C for 6 hours. One ml Of the enriched culture
was then transferred into 9 ml MacConkey broth with cefixime and potassium
tellurite (CT-Mac) (Mast Diagnostics, Merseyside, UK), and incubated for 18
hours at 37°C. After incubation, the CT-Mac solution was mixed, and 1 ml of
each suspension transferred into a hermetically sealed tube. The tube was
heated for 15 min in a 100°C water bath, and allowed to cool. The samples were
then screened for E. coli O157:H7 with the VIDAS” ECO assay (bioMerieux,

1998)

112

SPR Biosensor Salmonella Assays

The same enriched samples from the VIDAS immunoassay were used for
SPR analysis. The SPR biosensor was prepared with apprOpriate antibody, and
immersed in sterile BPW for two minutes as a zero control, then into sterile M-
broth for two minutes to identify the background response (Table 4.3). The
inoculated M-broth sample was then assayed for 3 minutes, followed by a pure
culture of S. Typhimurium in M-broth (the positive control).

Table 4.3. SPR biosensor procedure for Salmonella spp. assays

 

 

Step Action Time (min)
1 SPR sensor preparation (Table 2.1, Chapter 2) 30
2 Sterile BPW (zero control) 2
3 Sterile M-broth (negative control) 2
4 Sample in M-broth 3
4 anti-Salmonella Spp. antibody (300 jig/I) 2
Total time: 39

 

Following the inoculated M-broth sample, the sensor was immersed in
anti-Salmonella spp. antibody for a second time, to achieve ‘sandwich’ binding of

antibody (Figure 4.1). Two identical assays were performed for each sample.

Bacterium
Antibody a

w/

SPR biosensor .qud-film

 

 

 

 

 

 

 

 

 

 

.J
1'

 

(A) (3)

Figure 4.1. (A) Antibody-antigen interaction on the SPR biosensor gold-film
surface using one antibody layer a. (B) Antibody-antigen interaction on the SPR
biosensor gold-film surface in a sandwich-type interaction, with two layers Of
antibody, 3 and b.

113

The sandwich binding arrangement was designed to increase the
molecular mass of the substrate bound to the sensing film Of the SPR biosensor.
This was done to increase the magnitude Of the change in index of refraction of
the biosensor.

SPR Biosensor E. coli O157:H7 Assays

E. coli O157:H7 was assayed on the SPR biosensor following the same
procedure as the Salmonella spp. assays, using anti-E. coli 0157:H7 antibody
and sterile CT-Mac broth in place of M-broth for the negative control (Table 4.4).

Table 4.4. SPR biosensor E. coli O157:H7 assay procedure

 

 

Step Action Time (min)
1 SPR sensor preparation (Table 2.1, Chapter 2) 30
2 Sterile BPW (zero control) 2
3 Sterile CT-Mac (negative control) 2
4 Sample in CT-Mac 3
4 anti-E. coli O157:H7 antibody (300 jig/I) 2
Total time: 39

 

The sandwich assay was performed in the same manner as for the
Salmonella spp. assays. Two replicates were performed on each sample.

Because initial testing indicated that the samples contained negligible
amounts Of E. coli O157:H7, five samples were inoculated with both 10 pl Of a
stock solution of103 CFU/ml E. coli O157:H7 (about 10 CFU), and 10 “I or pre-
enriched sample from the pork production facilities to create known positives

(Table 4.5).

114

Table 4.5. Samples inoculated with E. coli O157:H7

 

 

Number Of
Site samples
Prechill 2
Hogs Back 1
Drain 1
Pen Alleyway 1

 

The samples were prepared in sterile BPW, and enriched according to the
method used for VIDASTM ECO samples.
SPR Biosensor Positive Controls

Positive controls for the Salmonella spp. and E. coli O157:H7 SPR
biosensor assays were pure cultures Of S. Typhimurium in M-broth, and E. coli
O157:H7 in CT-Mac broth, respectively. Positive controls were assayed
following the SPR biosensor assay procedures used for the samples, with pure
culture used in place Of the inoculated sample. Two positive controls each were
run for S. Typhimurium and E. coli O157:H7. Viable plate counts were performed
on the control cultures.
SPR Biosensor Assays in TSB

To find any difference in response Of the SPR biosensor to samples grown
according the VIDAST'“ enrichment protocols, and samples grown in TSB, the
pork industry samples were inoculated in TSB and grown for 18 hours at 37°C.
The samples were then analyzed with the SPR biosensor following the same
procedure as the assays performed on the samples grown according to the
VIDASTM methods. Sterile TSB was used as the negative control, and pure S.
Typhimurium and E. coli O157:H7 cultures grown in sterile TSB were used as the

positive controls. A second rinse Of antibody was performed for both anti-

115

Salmonella spp. and anti-E. coli 0157:H7 assays to achieve a ‘sandwich’-type
antibody binding.

E. coli O157:H7 positive samples were prepared in TSB in the same
manner as the E. coli O157:H7 positive samples were prepared for the VIDAS”
enrichment method. Five samples were inoculated with both 10 pl of a stock
solution of 10‘3 CFUImI E. coli O157:H7, and 10 pl of pre-enriched sample from
the swine production facilities, to create known positives. The same samples
from Table 4.5 were used to inoculate the E. coli O157:H7- positive samples in

TSB.

RESULTS AND DISCUSSION

The results of this study compared the SPR biosensor to the VIDASTM
immunosensor for detection Of Salmonella spp. and E. coli O157:H7 in samples
from swine farms and slaughterhouses. Three Of 24 (12.5%) samples tested
positive for Salmonella spp., and 0 Of 24 (0%) samples tested positive for E. coli
O157:H7 (Table 4.6). Confirmation Of positive Salmonella spp. results was done
by the Food Microbiology Lab Of Michigan State University (Ryser, 2000).
Suspected Salmonella colonies were purified and biochemically confirmed using
APl 20E strips (bioMerieux, St. Louis, MO), followed by serological identification
and pulsed field gel electrophoresis (PFGE) analysis by the Michigan
Department Of Community Health (Ryser, 2000).

All five samples inoculated with a stock solution Of E. coli O157:H7 tested
positive on the VIDASTM ECO assay. This method of spiking samples to create

. positive samples for the VIDASTM ECO assay is commonly used (Grif and

116

 

 

 

ea o 333 m . .93
m o o 590
m o o 2:59: ocEmceo
n o m . . =EOOE
r .o o >mie=< com mcEEmw Omaoficfiaagw
o o x
F o o 0.552 h2e>>
N o 0 been.
F o o ..csz ECO...
N o o xomn moo:
P o 0 >912? com mcaEmm EEO“.
meEEOw . . . .
Co Loeeez REED L28 m age SocoELem 8:83

 

5380558. E95.) 05 CO.— 3531
- Emco :28:an {on c5 2 RIKmFO .30 .m can dam 33:25.8 DO 3520:. we cam...

117

Allerberger, 1998). Positive results Of the VIDAS assay indicated that the E. coli
0157:H7 had grown to a population of at least 105 CFUImI — the detection limit Of
the VIDAS sensor (Cohen and Kerdahi, 1996). These results were used to verify
the presence Of E. coli O157:H7 in these samples for the SPR biosensor.

Results fer the SPR biosensor assays were found by calculating the _
change Of index of refraction Of the inoculated sample above the negative
control, as well as the change in refractive index of the biosensor reading for the
second application Of antibody above the negative control. This was done to see
if a second application of antibody caused an increase in magnitude of the
sensor response due to ‘sandwich’ antibody binding.

Analysis Of the SPR biosensor Salmonella spp. assays for samples grown
according to the VIDAS” enrichment method revealed that the average
response Of the SPR biosensor to the positive samples was greater than the
average response to the negative samples. However, ANOVA revealed that
there was no statistical difference between the positive. and negative samples
(Table 4.7). Growing the samples in TSB did not affect these results, again the
positive samples were not significantly different from the negative samples both
before and after second antibody application (Table 4.8).

The'average response of the SPR biosensor to the positive samples was
greater than the average response to the negative samples. However. analysis
Of the positive and negative results for E. coli O157:H7 show that response of the
SPR biosensor to the forced-positive samples was significame larger than the

negative sample in only one case (Table 4.9). Notably, the significant difference

118

Table 4.7. P-Values Of positive Salmonella spp. samples in
VIDAS-enriched method, after application Of first and second
antibody compared to n ative samples.

 

 

 

 

 

 

irst : Se—Eond
Sample source antibody antibody
Prechill 3 0.56 0.97
Prechill 4 0.36 0.26
Prechill 5 0.46 0.40

 

 

 

 

Table 4.8. P-Values Of positive Salmonella spp. samples in
TSB, after application Of first and second antibody, compared to
negative samples.

 

 

 

 

 

 

 

 

. First Second
Sample source antibody antibody
Prechill 3 0.97 0.95
Prechill 4 0.26 0.60
Prechill 5 0.40 0.96

 

*Statistically significant difference when P < 0.05

119

was for the sample before the application Of the second antibody. After
application Of antibody b, there was no significant difference Of the known-
positive sample responses and the negative responses. Fratamico et al., (1998)
reported the opposite phenomena: application Of a second ‘sandwich’ antibody
caused a significant increase in biosensor response. Other assays, including
VIDASTM and a sandwich-ELISA also use similar sandwich-type binding to
increase the magnitude of sensor response (Cohen and Kerdahi, 1996; Jay
2000). Further experimentation needs to be done to analyze the kinetics Of the
antibody response of the SPR biosensor used in this study, to see if the second
antibody is binding adequately.

Results for the SPR biosensor assays in TSB were calculated in the same
way as for the assays performed on the VIDAS-method enriched samples. Again,
the inoculated E. coli O157:H7 samples in TSB were used as positive controls for
the E. coli O157:H7 SPR biosensor asSays. The SPR biosensor responses for
the TSB samples designated positive for E. coli O157:H7 by the VIDASTM
immunosensor were larger than the responses for the VIDASTM negative
responses. However, ANOVA revealed that the positive sample responses were
not statistically different than the negative sample responses, both before and
after application of the second antibody (Tables 4.10).

Though the difference was significant in only one case, in a Pen Alleyway
sample, the average response of the SPR biosensor to positive samples was
larger than the response to negative samples (Table 4.11). This shows the

potential Of the biosensor for use in pathogen detection. If the magnitude Of the

120

Table 4.9. P-Values Of known-positive E. coli O157:H7
samples in VIDAS-enriched method, after application of first
and second antibody, compared to nega_tive samples.

 

 

 

 

 

 

 

 

Sample _ First Second
source Antibody Antibody
. Prechill 2 0.83 0.08
, Hogs Back 0.68 0.87
Prechill 4 0.39 0.27
Drain 0.71 0.10
Alley ‘0.04 0.12 '

 

 

 

Table 4.10. P-Values Of known-positive E. coli O157:H7
samples in TSB, after application Of first and second antibody,
compared to negative samples.

 

 

 

 

 

 

 

 

 

Sample First Second

source Antibody Antibody
Prechill 2 0.17 0.50
; mags Back 0.34 0.98
Prechill 4 0.25. ' 0.29
Drain 0.08 0.33
Alley 0.26 0.17

 

 

*Statistically significant difference when P < 0.05

121

 

cmccamce c>=mmcc

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmd wmdv mfmv omxm - Omcoamm: Ozzmom

mfmm 8.0m 2.8 , mmdm cozmScp Panama cameo:
umwomm _ wvdmmv medmmm _ 8va gnome: RIMES .EOO .m
K.Frmm _ 918v 3&me _ 8.2? c238 2.3.90 .EOO .m

3:032 @258:

no.9 050m 8.3 8.3. - cwcoame m>=_mOn_

«Dav 8.2. 53” mo. 5 5:936 Emucmfi mmEc><
8.2mm mmdwov 055% 3.3? m>=mmmc gocogmm
8.58 8.59» 8.0me 8.53 9:68 gocoesmm

882$ 88% 2893 En. 383$ 88% 3835. mi 2.5%
m9. COEOE EOEEOCCO 933

 

 

.co_§>mu Emucmz cmmcc>m c5 53, 92 on. Co cocmzetfi
o5 8:888 .mcaEmm c268: new 9558 .2 $582 .8585 mam cmEc>< a; can...

122

difference in biosensor response between positive and negative samples could
be increased, the SPR biosensor could be used as a diagnostic instrument in
pathogen detection in samples of this sort.

There are several. possible reasons that the SPR biosensor response is
not at this point as Of yet. As shown in Chapter 3, the specificity Of the SPR
biosensor decreases as the amount Of background bacteria increase. Early
plating of selected samples Obtained from the pork production plants revealed a
large number Of Ubiquitous bacteria. These organisms could have caused the
‘noisy’ response Of the SPR biosensor tO the target bacteria. The pre-enrichment
Of the field samples in lactose broth, which is not selective for Salmonella or E.
- coli, could have allowed the growth Of many types Of bacteria (Jay, 2000).
Additionally, cross-reactivity Of the antibodies with non-target Organisms could
have led to the decrease in sensitivity Of the biosensor. In a similar study Of E.
coli O157:H7, Grif et al. (1998) found that the number Of ubiquitous organisms
other than E. coli O157:H7 that adhered tOhthe antibodies of an immunomagnetic
biosensor had an affect on the results. Seo et al., (1999) reported that anti-
Salmonella spp. antibody interaction with non-target bacteria caused a decrease
in Sensitivity Of a rapid response biosensor.

Another factor could have been the decrease in contact between the
target organism and the antibody because Of the physical presence of other
bacteria ‘crowding out’ the target organism. As the sensing surface is immersed
into the sample, the antibodies contact only those bacteria closest to them. At

higher numbers Of non-target organisms, the target bacteria are spread more

123

thinly through the interfacial layer at the sensor surface. This causes the
antibody to come into contact with a lower number of bacteria, possibly fewer

than the detection level Of 107 CFUImI.

CONCLUSION

Although statistically significant only in one field sample, the SPR
biosensor response was higher for positive samples than for negative, Showing
the potential for the instrument in real-world applications. Sample preparation
involved a one-step process, which took a third or less of the time required by
other sensors. This short time required by the SPR biosensor for sample
preparation is favorable for field adaptation, as the assays can be easily
Completed in one working day.

It is evident that modifications must be made to the biosensor. Potential
modifications include (1) investigations into increasing the sensitivity and/or
Specificity of the biosensor, (2) assays to discover a‘more suitable media for use
in growing the sample cultures prior to analysis, (3) investigating new methods Of
affixing the antibody to- the SPR biosensor sensing surface, (4) obtaining
antibodies more specific tO the target antigens, and (5) investigating a more
suitable method Of sample preparation, among others.

Should these modifications be implemented, the SPR biosensor shows
potential for deteCting Salmonella spp., and E. coli O157:H7 in the pork industry.
and shows versatility that may allow it to be used on other foodbome pathogens.

VVIth further study, the SPR biosensor shows promise to provide a

124

complementary detection system to standard lab-based systems currently used

in food safety.

ACKNOWLEDGEMENTS .

We would like tO acknowledge the technical help of DrS. Ryser and
Sharma, Department of Food Science and Human Nutrition, and Dr. J. Prater,
Department of Animal Science, Michigan State University. This project was

funded by the National Pork Producers Council.

125

REFERENCES

BioMerieux, 1998. VIDAS E. coli O157:H7 (ECO) Manual. bioMerieux, Inc., 595'
Anglum Drive, Hazelwood, MO 63042-2320

Buzby, J., Roberts, T., Lin, C.-T., MacDonald, J., 1996. Bacterial Foodborne
Disease: Medical Costs and Productivity Losses. USDA Agricultural
Economic Report NO. 741

Cohen, A.E., Kerdahi, K.F., 1996. Evaluation of a rapid and automated enzyme-
linked fluorescent immunoassay for detecting Escherichia coli serogroup
0157 in cheese. Journal Of AOAC International. 79(4)858-860.

Davies, P. 1997. Food safety and its impact on domestic and export markets.
Swine Health Product 5: 1 3-20

DeMarco, D., Saaski, E., McCrae, D., Lim, D., 1999. Rapid Detection Of
Escherichia coli O157:H7 in Ground Beef Using a Fiber Optic Biosensor.
Journal Of Food Protection, Vol. 62, NO. 7, 1999, Pages 711-716

Elkind, J.L., Stimpson,,D.l., Strong, A., Bartholomew, D.U., Melendez, J.L. 1998.
Integrated Analytical Sensors: The Use of the TISPR-1 as a Biosensor.
Sensor and Actuators, 11/98

Fratamico, P.M., Strobaugh, T.P., Medina, M.B., Gehring, A.G. Detection of
Escherichia coli 0157:H7 using a surface plasmon resonance biosensor,
Biotechnology Techniques, Vol. 12, NO. 7. July, 1998. pp. 571-576

Grif, K., Allerberger, F. 1998. Evaluation Of a rapid test for Escherichia coli
O157:H7 with confirmation using immunoconcetration (VIDAS ECO plus
VIDAS ICE). bioMerieux, Inc.

Grif, K., Dierich, M.P., Allerberger, F. 1998. Dynabeads plus 3M petrifilm HEC
versus Vitek Immunodiagnostic Assay system for detection of E. coli
O157:H7 in minced meat. Letters in Applied Microbiology. 26: 1 99-204

lsaacson, R.E., Firkins, L.D., Weigel, R.M., Zuckermann, F.A., DiPietrO, J.A.,
1999. Effect of transportation and feed withdrawal on shedding Of
Salmonella Typhimurium among experimentally infected pigs. AJVR,
60:1155-1158

Jay, J.M, 2000. Modern Food Microbiology. (Sixth ed.) Aspen Publishers, Inc.
Gaithersburg, MD

Letellier, A., Messier, S., Quessy, S., 1999. Prevalence Of Salmonella spp. and

Yersinia enterocolitica in finishing swine at Canadian abattoirs. Journal Of
Food Protection, 62222-25

126

Medina, MB, 1997. Hygromycin B Antibody Production and Characterization by
a Surface Plasmon Resonance Biosensor Journal Of Agricultural Food
Chemistry 45:389-394

Ryser, E. R., 2000. Strain specific typing bacterial pathogens in the pork industry
chain. Final report for NPPC project #99026, August, 2000

Salamon, 2., Brown, M.F., Tollin, G., 1999. Plasmon resonance spectroscopy:
probing molecular interactions within membranes. TIBS 24 - June
1999:21 3-219

Seo, K.H., Brackett, R.E., Hartman, N.F., Campbell, DP, 1999. Development Of
a rapid response biosensor for detection Of Salmonella Typhimurium.
Journal of Food Protection. 62(5)431-437

Texas Instruments, Inc. Dallas, Texas, USA. 1999. Operation Manual, Spreeta
Experimenter’ s Kit

U. S. Meat Export Federation, 1997 Total U. S. Pork Exports. Available from
URL: http: l/www. usmef. org/expstat/PORK8796. html

Verozny-Rozand, C., Mazuy., C., Ray-Gueniot, S., Boutrand-Loei, S., Meyrand,
A., Richard, Y., 1997. Detection of Escherichia coli O157:H7 in French-
food samples using an immunomagnetic separation method and the
VIDASW' E. coli 0157. Letters in Applied Microbiology, 25:442-446

Younts, SM, 1999. Development and Evaluation of a Gas Sensor Based
Instrument for the Detection and Differentiation Of E. coli 0157:H7 from
Non-0157:H7 E. coli. Thesis'for the degree of MS, Michigan State
University, East Lansing, MI.

127

RECOMMENDATIONS FOR FUTURE RESEARCH

The SPR biosensor (Spreetam) used in this investigation is a new,
experimental product that has not before been tested or used for microbial
detection. AS this was the first investigation along this application, there are
many optimizations to be made before the biosensor is perfected. Following are I
recommendations for future investigations to be conducted on the SPR biosensor
tO overcome initial limitations.

First, preliminary experiments were conducted to establish operating
procedures based upon. procedures suggested by the vendor (Spreetam)
manual (Texas Instruments, 1999). Several modifications were made to these
procedures, due to the effect Of the growth media used for the target pathogens
on the SPR biosensor response. Research needs to be done to explore the
effect Of different media and operating procedures on the biosensor response.
Experiments to find a culture medium with a matrix that does not interfere with
the biosensor response could increase both sensitivity and specificity of the SPR
biosensor. Investigating new sampling methods that minimize the amount Of
nOn-target bacteria could also improve the SPR biosensor response.

Results Of SPR biosensor assays on cultures compOsed Of a mixture Of E.
coli O157:H7 and S. Typhimurium Showed a decrease in Specificity when
compared tO assays performed on pure cultures. Future studies need to be
carried out using cultures composed Of additional types and quantities of

bacteria, to establish more precisely the specificity limit of the SPR biosensor.

128

Another recommended area for future research is in the antibody selection
and attachment methods. Research needs tO be conducted utilizing other
brands and types of antibody, to discover whether the degree Of Specificity Of the
antibody to target-antigen is a limiting factor in the‘ SPR biosensor. Also, a new
method of affixing the antibody to the sensing surface might improve the
biosensor responses. Currently, the antibody is affixed via avidin-biotin
interaction. This method may not be Optimal for use on the SPR biosensor.
Future investigations with other methods of antibody attachment Should be done
to study the effect Of the degree Of antibody attachment to the sensing surface.

Another factor to be studied is the incorporation Of a flow-cell arrangement
with the SPR biosensor. Most other types Of commercially available SPR
biosensors incorporate a flow-cell arrangement. A constant flow Of sample
solution over the sensing surface could increase the sensitivity and specificity Of
the SPR biosensor, because more Of the target organism would come into direct
contact with the antibOdy-prepped SPR surface. For this reason, further
investigation with this biosensor Should be carried out to develop and perfect a
flow-cell assembly. I

Furthermore, the SPR biosensor is currently an Open system, with the
samples and preparation solutions Open to the air. . This can lead to
contamination of the sample with any contaminants present in the air or surfaces
Of the laboratory setting. Creation of a closed system, such as a flow-cell
assembly, that brings the sample in contact with the biosensor sensing surface

without exposure to background contaminants such as dust, dirt, or other

129

contaminants could minimize noise in the SPR biosensor response due to
contamination Of the samples and preparation solutions.

Finally, it is recommended that research be done to investigate whether it
would be possible to prepare the SPR biosensor a longer period Of time before it
is used for sampling. If the SPR biosensor could be prepared ‘in bulk’ or ahead
Of time, it would increase the possibility Of using the biosensor in the field.
Prepared biosensors could be stored until needed, and sampling could be done
immediately (2 min), rather than waiting the time needed to initialize the SPR
biosensor before sampling (30 min).

Results Of this study are promising for converting an Off-the-Shelf SPR
biosensor into a diagnostic tool for pathogen detection. With additional
investigation and refinements to perfect the methods and experimental set-up Of
the SPR biosensor, it Shows the potential tO fulfill the food industry’s need for an

on-Iine, rapid, accurate, and portable pathogen detection system.

130

APPENDIX A

RESEARCH PROCEDURES AND PROTOCOLS

1. Creating stock cultures and determining concentrations... ..
2.Antibody Preparation
3. SPR BiosensorPreparation.......................................
4. Buffer preparation............. . ..... ..
5. VIDASTM Salmonella Enrichment Preparation......................................

6. VIDASTM E. coliO157zH7 Enrichment Preparation..............................

131

132

133

.134

135

.136

137

APPENDIX A.1: Creating stock cultures and determining concentrations

TO create stock cultures of E. coli O157:H7 and S. Typhimurium

1.
2.

3.

Vortex sample

Aseptically transfer isolate to 10 ml Of Tryptic Soy Broth (TSB) (Difco
Laboratories) in a sterile 14 ml glass vial

Incubate inoculated vial for 18-24 hours at 37°C.

Determination Of culture concentrations

91:59)

Ammse

. Vortex sample

Prepare 4 vials with 10 ml Of sterile Buffered Peptone Water (BPW) (Becton
Dickinson and CO., Sparks, MD)

Transfer 100 pl Of stock culture into first vial (1:100 dilution)

Vortex vial and transfer 100 pl to second vial (1 :10,000) dilution)

Vortex second vial and transfer 100 pl to third vial, and 1 ml to each Of 2
plates (3M Petrifilm, St. Paul, MN) (1:1,000,000 dilution)

Vortex third vial and transfer 100 pl to fourth vial, and 1 ml to each Of 2 plates
Vortex fourth vial and transfer 1 ml to each Of 2 plates (1: 10: 000: 000 dilution)
Incubate plates at 37°C for 18-24 hours

Count plates

0. Back calculate 10-fold dilutions for every step, to determine original culture

concentrations.

132

APPENDIX A.2: Antibody Preparation

1.

2.

Preparation Of Buffers
1.1. 0.01 M Acetic Acid Solution
1.1.1. Mix 25 pl Glacial Acetic Acid (Spectrum Products, Inc, Gardena,
CA) with 40 ml reagent quality water
1.2. 0.177 M Carbonate-Bicarbonate Solution
1.2.1. 1.09 g Na2C03 and 0.63 g NaHC03 dissolved in 100 ml reagent
quality water.
Antibody Rehydration
2.1. Add 100pl 0.01 M Acetic Acid to 1 mg Of antibody
2.2. Rotate vial until completely dissolved
2.3. Add 100pl Carbonate-Bicarbonate Solution
2.4. Rapidly mix until solution is clear or opalescent
2.5. Immerse in 37°C waterbath for 30 minutes
2.6. Allow to cool Slowly to room temperature
Biotin Labeling
3.1. Add 600pl of PBS (Phosphate Buffered Saline) (Pierce Chemical,
Rockford, IL) to rehydrated antibody
3.2. Dissolve 2mg EZ-Link NHS-LC-Biotin in 1m| N, N-Dimethylfonnamide
(HCON(CH3)2 (DMF) (Aldrich)
3.2.1. Add 75pl Of this mixture to the antibody solution
3.3. Place on ice for 2 hours
Antibody Dilution
4.1. Add 9 ml PBS to Antibody solution
4.1.1. Antibody concentration is now at approximately 300 mg/ml

133

APPENDIX A.3: SPR Biosensor Preparation

1.

Initialize sensor

1.1. Immerse gold surface in 10 ml of 0.1 N NaOH in 1% Triton X-100 Solution
(Sigma, St. Louis, MO) for 3 minutes.

1.2.Then immerse in 10 ml PBS for 3 minutes, until a steady PBS baseline is
established '

1.3.‘lmmerse in 10 ml 100 pI/ml neutravidin (Pierce Chemicals)in PBS

1.4. Immerse in 10 ml Of PBS

1.5. Immerse in 10 ml Of 300 pglml Salmonella, or E. coli O157:H7 antibody
(Kirkegaard & Perry Laboratories, Gaithersburg, MD) for 10 minutes, to
bind the antibody to the avidinated surface via avidin-biotin interaction

1.6. Rinse unbound antibody by immersing the sensor in 10 ml PBS

1.7. Immersed biosensor in NaOHfI’riton in PBS for two minutes

1.8. Immerse in Bovine Serum Albumin (BSA) (Pierce Chemicals) for two
minutes

1.9. Rinse remaining BSA with NaOHfTriton in PBS for two minutes

Sampling

2.1. Immerse sensor in the negative control for 3 min.

2.2. Immerse in 10 ml Of the sample for 3 min.

2.3. Immerse in positive control (Pure culture) for 3 min.

Sensor cleaning

3.1. Immerse sensor in PBS

3.2. Immerse sensor in 10 ml 0.1 N NaOH-Triton X-100 Solution

134

APPENDIX A.4: Buffer Preparation

1. Dilution Blanks ,
1.1.0.1% Buffered Peptone Water (Difco Laboratories)
1.1.1. Mix 19 of Buffered Peptone in 1 L purified water
1.1.2. Distribute 9.0 ml or 9.9 ml per test tube
1.1.3. Autoclave 15 min at 121°C

135

APPENDIX A.5: VIDAS” Salmonella Enrichment Preparation

BAM/AOAC Procedure

1.

Pre-Enrichment

1.1 . Add 1 ml Of sample to 9 ml buffered peptone water

1.2. Blend and incubate for 18 hours at 37C

Enrichment

2.1 . Transfer 1 ml suspension into 10 ml Tetrathionate broth

2.2. Incubate 6-8 hours at 42C

2.3. In parallel, transfer 1 ml suspension into 10 ml Selenite Cystine broth.

2.4. Incubate 6-8 hours at 35-37C .

Post-Enrichment

3.1 . Transfer 1 ml from Selenite Cystine broth into 10 ml M broth

3.2. Transfer 1 ml from Tetrathionate broth into 10 ml M broth.

3.3. Re-incubate both broths for 18 hours at 420 for use in later confirmation

3.4. Incubate both M-broth samples for 18 hours at 420

After Incubation

4.1. Mix each M broth and transfer 1 ml from each into a tube.

4.2. Cap tube tightly, heat for 15 min in a water-bath at 100C

4.3. Perform VIDAS test

4.4. Store remaining M broths at 2-80 for confirmation

Confirmation Of Positive Results

5.1 . Streak positive Selenite cystine or Tetrathionate broths and remaining M-
broth onto XLD plate (specific agar for Salmonella) following standard
plating procedures

5.2. Incubate plates at 37C for 18-24 hours

5.3. Confirm suspect colonies

136

APPENDIX A.6: VIDAST'“I E. coli O157:H7 Enrichment Preparation
Sample Preparation

1. Pre-Enrichment
'1.1. Dilute 1 ml sample into 9 ml m-TSB with novobiocin
1.2. Incubate samples 6-7 hours at 41 C
2. Enrichment
2.1.Transfer 1 ml enriched culture into 9 ml MacConkey broth with cefixime
and potassium tellurite (CT-Mac)
2.2. Incubate 18 hours at 35-37C
3. After Incubation
3.1.Mix each M broth and transfer 1 ml from each into a tube.
3.2. Cap tube tightly, heat for 15 min in a water-bath at 1000
3.3. Perform VIDAS test
3.4. Store remaining M broths at 2-80 for confirmation
4. Confirmation Of Positive Results
4.1. Make successive 10-fold dilutions Of positive samples in tryptone salt
4.2. Spread 100 pl Of the 10:2, 10:3, and 10“ dilutions on the surface of mm
and CT-SMAC plates
4.3. Incubate 18-24 hours at 370

Media preparation

1. m-TSB with novobiocin
1.1. Mix:
30 g dehydrated Trypcase' Soy broth base,
1.59 Bile Salts #3,
1.25 g Anhydrous NazHPO4,
.4. 10 g Casaminoacids,
Autoclave at 121 C for 15 min.
Add 1 ml Of 20 mg/ml novobiocin
M

1.1.1.
1.1.2.
1.1.3.
1.1
1.2.
1.3.
2. CT- AC — MacConkey broth with Cefixime and potassium Tellurite
2.1.Mix:
2.1.1. MacConkey broth, prepared and autoclaved at 121 C for 15 min.

2.1.2. 250 pl Of 1% potassium tellurite solution / liter Of medium
2.1.3. 1 ml of 50 mg/I stock solution Cefixime / liter Of medium

3. CT-SMAC - MacConkey Sorbitol Agar with Cefixime and potassium Tellurite
3.1. MacConkey sorbitol agar, prepared according to manufacturers
instruction
3.2.Autoclave at 121 C for 15min.
3.3. COOI to 45-50C,
3.3.1. Add:
3.3.1.1. 250 pl of 1% potassium tellurite solution / liter Of agar

137

3.3.1.2. 1 ml Of 50 mg/l stock solution Cefixime I liter Of agar

4. Cefixime stock solution at 50 mg/I
4.1. Mix
4.1.1. 5 mg Of Cefixim
4.1.2. 100 ml of 100 mmolll sodium bicarbonate buffer, pH 8.0
4.2. Sterilize, store at 2-80

138

APPENDIX B

RESULTS FOR CHAPTER 3

SENSITIVITY AND SPECIFICITY ASSAYS FOR SALMONELLA SPP. AND
ESCHERICHIA COLI 01 57. H7 UTILIZING A SURFACE PLASMON
RESONANCE BIOSENSOR

Figure B.1. Background responses Of the SPR biosensor to serial dilutions Of
S. Typhimurium, E. coli O157:H7, and sterile TSB without

anfibody ............................................................................................. 140
Figure 82. Anti-Salmonella antibody comparisons... ..141
Figure 83. Anti-E. coli O157:H7 antibody comparisons... ..142

Figure B. 4. Negative control Of SPR biosensor — response tO serial dilutions
Of sterile TSB using anti- Salmonella and anti-E. coli 0157: H7

antibodies... 143
Figure 8.5. SPR biosensor response for S. Typhimurium... ...145
Figure 86. SPR biosensor response for E. coIiO157zH7... ......146
Figure 87. Timed growth Of S. Typhimurium and E. coli O157:H7... ...147
Figure 88. Specificityfor S.Typhimurium..............,............................148
Figure 3.9. Specificity for E. COIiO157:H7..................... ....149

Figure 810. Specificity for anti-Salmonella spp. sensor in mixed culture... 150

FigUre B.11. Specificity for anti-E. coli O157:H7 sensor in mixed culture...‘.151

139

goon 84-8.. Jacobean song; mwp gees bee .EREO Eco

.m .EESEESC. .m LO 295:... 3:3 2 .8505 mam 9.: CO momcoamc. ccaocoxomm ....m 939“.

oo+w. _.

cos—:5

Slug. Norm... worm... «scum... norm; moimé helm... worm... morme

r p — p P L P b

 

 

 

:2 x 3 8:25:88 fies EEO :8 .m
...9 x 3 58.5528 as; Eeeeezafi .wv

 

 

 

P mwhlll

~ 535 =8 .mleT .
. $55 =8 .m Idl
m genera: .w lei

. Easszefi .m led

 

 

 

 

e

.9.

009

000 _.

ooooe

uoltoeueu to xeput u! a6ueu3

140

mo+m. F

mCOmtmaEOO Conzcm gccoEmeéce. .N.m 9:9...

 

 

 

 

 

 

    

3.23.50an
mo+wF B+mF 8+m.F mo+m.F 3+mF mom—F No+m.F Fo+mF oo+m_.F Fo-m.F
_ _ F _ 'L _ _ F , O
. 009
m EB: oomHHI F ooow
m Ea: oomll
E a
F. a Ball . 88
m Eta: om II
N EB: omlul
:Ea: omlxl . oooe
m Eta: m lel
o.
N _E\ : mlbi . 88
F EB: mlel

 

 

 

 

 

Doom

uopoeuau go xaput u! afiueua

141

wo+w. F

_

28838 ESE: FIHSFO :8 .m-E< Om ESE
FESLOF eoeceeeoeoo 8..

 

 

 

 

 

 

8m: 3m: 8m: 8+mF momF
. o
. 82
O
m
m ESE S81: T o8~ w
N ESE o8 II M
F ESE 08 IT - 88 m
m ESE 8 r? w
N ESE 8 IT 8:: W.
F ES: 8 If m.
.0. ES: 8 tel m
m ES: mini , 88
F ES: 8 r1

 

 

 

Doom

 

142

28¢. 8.78.. .3898 58 F0 :8 .mEn :8 SSEESm
4ch 9.6: mm... 286 LC 825.6 .mtow 2 8:83. - 58805 mam FO EEOC c>=momz Sm 939m

Sim E c825 .5335» mm... 32:. SEE 69.: c0325 mmh
oo+m.F Fem: No.3 $3 49m: 3m: 8m: 5m: 8.3

»

worm. F

 

 

 

 

8 $85 __8.m-EeI_T-
m E:_L:E_8>F.w-_.enloi
N 285 :8 8.8+
F ES 5 :8 wees. IT
N 83:38.83. .3ch IT

 

 

 

F EBSEEQC. .2ch II

 

 

F

O
F

 

 

ooF

 

. oooF

 

 

coco F

uoltoeueu jo xapur Ul e6ueu3

143

2R8 m3 .EESEEQF... .m .8 8:859. .8885 mam mm 939”.

FESLOF eoegeeceoo 8..

mom—F mo+m.F mom: 8+wF mo+m.F vo+w.F mo+m.F New... Fo+w.F oo+w.F Form: Noer

F F F F F F F F F F

 

   

 

m EESEEEC .m IeI
N E:_5E_F._a>._. .w lat
F EBB—EEG? .m Isl

 

F

O
‘—

FooF

. oooF

 

 

0000 F

uopoeuau to xepul u! e6ueu3

144

0.08 00.00.. .FIKmFO :00 .m .8 08089 08805 mam .mm 059...
EBB. 8.88880 8..
mo+w.F .mo+m.F Fo+m.F mom: m0+m.F 3+wF m0+m.F No+w.F Fo+m.F 00+m.F F0.w.F No..m.F

F F F F F F F F P r P

 

soF

- ooF

 

m 2.85 .8 m. If.
N 2.85 ..8 midi
,F FIHFmF,C,...8 mile»- . 08F

UOltoeuea to xepul ul efiueug

 

 

 

0009

145

or

m

F

EBB .008 .m 05 easgfifi .m 8 5565 85: .E 9:9“.

w

b

N

p

5:05 05:.

m

m

o

 

L———.y_

0

_

_

 

 

83:38.59»... .w 092024 I

:00 .w 00903. ....I

m zoom
m =8 .m
r =8 .m
m 530
N 230

OIQXX.

 

 

 

 

 

 

 

 

T

T

Y.

YI

 

 

 

 

AV ....

 

 

 

 

 

 

c2598 3 =5:

 

 

 

 

 

oo+mo. F

5+wo...

No+mo. _.

mo+mo. w

3+mo. _.

mo+mo. w

8+wo. r

~o+wo. _.

mo+mo. _.

mo+wo. _.

o_.+mo.w

(IN/(HO) uouenueauoo 601

146

0.000 00.00.. .EESEEQE. .w .0. 3.0.0.0000 0.0 030....

225000 85.8080
8+3 8+3 3+3 8+3 8+3 8+3 8+3 8+3 5+3 8+

_ _ h h .

w... Pom... N0-w...

b

 

 

m mwhlal
m E.mw [T
N =00 .m lol
.. :00 .mlIl
N E.mw+
.. E_mw lal

 

 

 

 

 

 

F

0..

- 00..

000 ..

0000 ..

uogweuaa ;0 xepul u! efiueuo

147

03m...

28» 02-03 318.080 .0 818.880 0.0 2:9“.
25.300. 8:98.880

00+m... 00+m._. v0.7”? N0+m... 00+m... Now...

 

 

m :00 .mlll . ..

m 09'
0 5.3+
N :00 .m ...T
F =00 .m IT
a 2:00 IT
3 EfimLT
N 03:01

m mmhlol

-0?

 

-00..

 

 

 

f 000?

 

 

0000 F

 

uogpeuea ;0 xepul u! afiuqu

148

0.3.30 005.: :_ .0m:0.00_0 mam 00.0008. - .000 0.03.02.00-80 .0. 3.050000 . 03.0 0.20.“.

0.000 00.. ..ESuo. :0_.0..:00:00 52.35303. .0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00+m€ 00+m... v0+m._. 00+me N0+wé w0+me
. IIL .
30.0 A a. .--..- 6.3.30 0.3... amp 52.3620: .wIIH.
63:00 . \ N0+m00._.|..|.
02.000: 8+083|T
:05 E8950
>__00..0_.0.0 3&8. 3 IOI
.02 25500. . A 8883+
co_.m.«cwocoo 51$er .30 m 00.5004 Ix...
Amod Vn: h0+w00...|{|
0.600 r 0908.. Ial
02.000... .2.:00 02.0002 ...OI ;
Cw...“ “€050.96 - x -r! - u ----i. -.Innllitiil! ,. 1i.-. ,
28.8000

 

 

 

 

 

 

 

00+wé

o
. 82

O

u-

e

u

. 88 w

m.

m.

p.

k 88 m

m.

m.

88 w

m.

m.

u

, 0000

 

 

0000

149

0.3.30 005:. :_ .00:000.0 mam 02000.0 - 51.3.0 ...00 .m-..:0 .0. 3.00.0000 3.0 0.30.“.
0.80 03 .833. 8:82.880 2.8.0 .80 .0
«0+0;

00+w. ..

00+m. P

00+m... v0+m...

 

 

 

 

GQOAQ.
.0..:00
02.000:
85 88......
38.0.0.0
82

 

 

 

 

 

Gaovav
.0..:00
0>=000:
85 .888
2.8.08.0

A

 

 

 

 

 

00+w. ..

 

F0+mé

 

 

 

a

 

 

283.00 .

5303:0050 E35820: .m A

 

 

 

 

 

/

8.3.30 0.3% 00.. BIKmFO :00 .m I

N0+m00. P I
00+M00. F I
3+m00. — IOI
00+m00. F III
00.500.— IXI
50.500. w IQI
00+M00. P Ill

.0..:00 020002 I

 

00+w...

0

- 000w

- 000m

.. 0000

.. 000v

0000

 

 

0000

uonoeuau :0 xapul u| afiueuo

150

APPENDIX 0

RESULTS FOR CHAPTER 4

DETECTION OF SALMONELLA SPP. AND ESCHERICHIA COLI O157:H7 IN
THE PORK PRODUCTION CHAIN USING A SURFACE PLASMON
RESONANCE BIOSENSOR AND A VIDAS IMMUNOSENSOR

Table 01. SPR biosensor results for Salmonella spp. assays in samples
enriched according to the VIDASTM protocol... 152

Table 0.2. SPR biosensor results for E. coli 0157:H7 assays in samples
prepared according to the VIDASTM protocol....... 153

Table C. 3. SPR biosensor results for E. coli 01 57: H7 assays of samples
inoculated with E. coli 0157: H7 prepared according to the VIDAST'“ -
enrichmentmethods... . ..... .......154

Table C. 4 SPR biosensor results for Salmonella spp. assays in samples
enriched In TSB... ....155

Table C.5.-SPR biosensor results for E. coli O157:H7 assays in samples
enriched in T88 W156

Table C.6. SPR biosensor results for E. coli O157:H7 assays of samples
inoculated with E. coli O157:H7, prepared in sterile TSB... ...157

151

:0..0.>00 :0_.0.>00
0.00:0.0 N 5.5003. — :0.5000~. 0.0.30.0 N 5.5000”. 3 5.5000”.
3 88.35.. 8 3 88.3.2 8 2:80 8. 2980 8.
5.00000”. .0 0.03:. 8:00:00 .0 x00:_ 5.60.00". .0 .805 8.80.000 .0 x02.
:. 00:0:0 000.024 :. 00:0:0 000.0>< :_ 00:0:0 :. 00:0:0

 

152

:000300
0.00:0.0

N :000000m . 50.30000

a €8.22 .0 a 885. .o
5:00.300 .0 0.00:. 5.30000”. .0 080:.
:. 00:0:0 000.0>< :. 00:0:0 000.0><

:000300
28520

N 83000”.
0.0500 .0.
3002.0”. .0 x00:_

00:0:0 000.0><

— 5.5000”.
0.0500 .0.
5.00000". .0 .805

00530

 

153

 

 

 

 

 

 

 

«0.00 50.0.00 00.0.80 00. 3 5.02.3 00.0000 >0..< 0v~
N0. .0 «0. Ram 00.0000 00.00 3.030 00. .0: :.0.0 00.
N». 3 «0.0000 00.0000 00.: 00.300 3.0va v ...:00.0 03
3.3 00. F .00 00.0300 No.00 00.0 2.0 0.. .0? ..000M001 00
00.0. 10.900 2.60m... 00.00 «I000: 00.030 N ...:00.0 0c
8.0.50 .92 x. .02 x. 8.0.50 .02 x. .02 x. 8.0880 .0..
0:00:20 N :0....000m . :0....000m. 0.0.50.0 N 5000001 . :0....000m. 0.0500
3 88...... a 83...... 2980 2980
.0 5.60000 .0 5:05.00. .0. 8:08.00 .0. 5:05.00
.0 x00:. :. .0 x00:. :. .0 x00:. :. .0 .005 :.

 

 

 

 

 

 

 

 

 

000:.05 50550.50 30(9) 0:. 0. 050.0000 00000.0

.0.... $0 ...00 .m. .....s 00.0.3005 00.0500 .0 0.0000 01.000 ...00 m .0. 0.300. .00:000.0 «.00 ..m 0 0.00...

154

02.0300
0.3520

000000000 . 8.50000

a €8.02 3 0 .0805. .0
02.05.00 .0 x000. 00:05.0". .0 .800.
0. 000000 000.02.. 0. 000000 000.0><

0000500
23520

«5.50000 . 8.50000
0.0500 .0. - 03500 .0.
00:05.01 .0 x005 00005.0”. .0 x005
0. 000000 000.0>< 0. 000000 000.03..

 

155

00.00300
0.00030

N 00.50001
0 0.8.05.
00.000001 .0
0. 000000 000.0><

0000300
. 00.000001 0.00030
0 .6805. .0
02.00001 .0 .000.
0. 000000 000.03..

N 00.000001
0.0500 .0.

000000

.0 .800.

. 00.0.0001
0.0500 .0.

000000

.0 000.

 

0000300
200020 P 02.00000

F 8000000
0 380.2 0 >825
00 00000001 00 00 00000001
0_ 00 x000. 0_
00000><

0000300
000006 N 00000001 P 00000001
20600 200000
000 00000001 000 00000001
00 x000. 0_ 00 x000. 0_
00000>< 000000 00000><

 

000 .m 0:3 00003005 00.0E00 00 20000 0.035 .000 .m 000 02:00. 00000005 100 .00 0.00...

157

BIBLIOGRAPHY

158

BIBLIOGRAPHY

Alocilja, E.C., Ryser, E., Osburn, W., 2000. Distribution and Prevalence of
Foodborne Pathogens in the Pork Industry Chain. Project proposal to
USDA-National Research Initiative Competitive Grants Program
(NRICGP)

Anderson, A. 1994. What is the HACCP? Journal of Human Nutrition and
Dietetics 7(1):53-60

Besser, R.E., Lett, S.M., Weber, J.T., Doyle, M, P., Barrett, T.J., Wells, J.G.,
Griffin, PM. 1993. An Outbreak of Diarrhea and Hemolytic Uremic
Syndrome from Escherichia coli O157:H 7 in Fresh-Pressed Apple Cider.
Journal of the American Medical Association 269:221722220

BioMerieux, 1998. VIDAS E. coli O157:H7 (ECO) Manual. bioMerieux, Inc., 595
Anglum Drive, Hazelwood, MO 63042-2320

Bryan, F .L. 1990. Application of HA CCP to ready-to-eat chilled foods. Food
Technology 44(7):70-77

Buchanan, R.L., and Doyle, MP. 1997. Foodborne Disease Significance of
Escherichia coli O157:H7 and other Enterohemorrhagic E. coli.
F oodTechnology 51:69-76

Buzby, J., Roberts, T. 1996. Microbial foodbome illness. Choices; First
Quarter: 14-1 7

Buzby, J., Roberts, T., Lin, C.-T., MacDonald, J., 1996. Bacterial Foodborne
Disease: Medical Costs and Productivity Losses. USDA Agricultrual
Economic Report No. 741

Canh, TM, 1993. Biosensors. Chapman and Hall, London, UK

CAST- Council for Agricultural Science and Technology, 1994. Foodborne
Pathogens: Risks and Consequences. Task Force Report No. 122,
Washington DC.

Cohen, A.E., Kerdahi, K.F., 1996. Evaluation of a rapid and automated enzyme-
linked fluorescent immunoassay for detecting Escherichia coli serogroup
0157 in cheese. Journal of AOAC International. 79(4)858-860.

Cranfield Biotechnology Centre, 2000. Available at URL: httpzllwwwcranfielci
ac.uklbiotechlspr.htm -

159

Davies, P. 1997. Food safety and its impact on domestic and export markets.
Swine Health Product 5: 1 3-20

DeMarco, D., Saaski, E., McCrae, D., Lim, D., 1999. Rapid Detection of
Escherichia coli O157:H7 in Ground Beef Using a Fiber Optic Biosensor.
Journal of Food Protection, 62(7)711-716

Deshpande, 8.8., 1994. Immunodiagnostics in agricultural, food, and
environmental quality control. Food Technology, June 1994;136:141

Doyle, MP, and Schoeni, J.L. 1987. Isolation of Escherichia coli O157:H7 from
retail fresh meats and poultry. Applied Environmental Microbiology
532394-2396

Doyle, M.P., Beuchat, L.R., Mountville, T.J., 1997. Food Microbiology:
Fundamentals and Frontiers. ASM Press. Washington DC.

Earp, R.L., Dessy, RE, 1998. Surface Plasmon Resonance. In Ramsey, G.,
(ed) Commercial Biosensors, Applications to Clinical, BiOprocess, and
Environmental Samples. Wiley-Interscience Pub., pp 99-164

Elkind, J.L., Stimpson, D.I., Strong, A., Bartholomew, D.U., Melendez, J.L. 1998.
Integrated Analytical Sensors: The Use of the TISPR-1 as a Biosensor.
Sensor and Actuators, 11/98

Epling, L.K., and Carpenter, J.A., 1990. Antibiotic Resistance of Salmonella
isolated from pork carcasses in Northeast Georgia. Journal of Food
Protection, 532253-254

Faegerstam, L.G., Frostell-Karlsson, A., Karlsson, R., Persson, B., Roennberg, I.,
1992. Biospecific interaction analysis using surface plasmon resonance
detection applied to kinetic, binding site and concentration analysis.
Journal of Chromatography 597:397-410

Forsythe, RH. 1996. Food Safety: A Global Perspective. Poultry Science.
75:1448-1454

Foster, EM, 1997. Historical Overview of Key issues in Food Safety. Emerging
Infectious Diseases, Special Issue. 3(4). Available from URL:
httpzllwwwcdcgov/ncidodlElD/vol3no4/fosterhtm

Fratamico, P.M., Strobaugh, T.P., Medina, M.B., Gehring, AG, 1998. Detection
of Escherichia coli O157:H7 using a surface plasmon resonance
biosensor. Biotechnology Techniques, Vol 12, No. 7, pp 571-576

Grif, K., Allerberger, F. 1998. Evaluation of a rapid test for Escherichia coli
O157:H 7 with confirmation using immunoconcetration (VIDAS ECO plus
VIDAS ICE). bioMerieux,- Inc. '

160

Grif, K., Dierich, M.P., Allerberger, F. 1998. Dynabeads plus 3M petrifilm HEC
versus Vitek Immunodiagnostic Assay system for detection of E. coli
O157:H7 in minced meat. Letters in Applied Microbiology. 262199-204

Griffin, PM. 1995. Escherichia coli O157:H7 and other enterohemorrhagic
Escherichia coli. In Infections of the Gastrointestinal Tract. Blaser, M.J.,
Smith, P.D., Ravdin, J.l., Greenberg, H.B., and Guerrant, R.L. (eds).
Raven Press Ltd., New York, NY

Griffin, PM, and Tauxe, R.V. 1991. The Epidemiology of Infections Caused by
Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the
associated hemolytic uremic syndrome. Epidemiology Review. 13:60-98

Harris, R.D., Luff, B.J., Wilkinson, J.S., Wilson, R., Schiffrin, D.J., Piehler, J.,
Brecht, A., Abuknesha, R.A., Mouvet, C., 1996. Integrated optical surface
plasmon resonance biosensor for pesticide analysis 1995/6 Research
Journal, Optical Fibres Group, University of Southampton, contact
riournal@ecs.soton.ac.uk

He, X., Shen, L., Malmborg, A.C., Smith, K.J., Dahlbaeck, B., Linse, S., 1997.
Binding site for C4b-binding protein in vitamin k-dependent protein 3 fully
contained in carboxy-terminal laminin-g-type repeats. A study using
recombinant factor IX-protein S chimeras and surface plasmon resonance.
Biochemistry 36:3745-3754

Hoffmann, S. L. 1993. Southwestern Internal Medicine Conference: Shiga-Like
Toxins in Hemolytic-Uremic Syndrome and Thrombotic Thrombocytopenic
Purpura. American Journal of Medical Science 306:398-406

Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, eds. 1994. Foodborne
Disease Handbook, Volume 1: Diseases Caused by Bacteria. Marcel
Dekker, Inc. New York, NY

IDEXX Food Safety Net, 1998. The Comprehensive Hazard Analysis Critical
Control Point (HACCP) Course Participant’s Guide. IDEXX Food Safety
Net Services, Westbrook, ME

IFST (Institute of Food Science and Technology), 1997. Salmonella
Typhimurium DT104 Position Statement 11 Nov. 1997. Available at URL:
httpzllwww. ifst. org/hottop20. htm

lsaacson, R.E., Firkins, L.D., Weigel, R.M., Zuckermann, F.A., DiPietro, J.A.,
1999. Effect of transportation and feed withdrawal on shedding of
Salmonella Typhimurium among experimentally infected pigs. AJVR,
60:1155-1158

Jay, J.M, 2000. Modern Food Microbiology. (Sixth ed.) Aspen Publishers, Inc.
Gaithersburg, MD - -

161

Jonsson, U., 1998. Optical Affinity Biosensors. In in AD. Scott (ed.) Biosensors
for Food Analysis, Athenaeum Press Ltd, Gateshead, Tyne and Wear, UK,
pp. 37-46

Kajiwara, H., 1998. Motif 2 in adenosine kinase homologous hinseng polypeptide
showed affinity to D-ribose by capillary zone electrophoresis and surface
plasmon resonance Journal of Chromatography 817:173-179

Karlsson, R., Michaelsson, A., Mattsson, L., 1991. Kinetic analysis of monoclonal
antibody-antigen interactions with a new biosensor based analytical
system Journal of Immunological Methods, 145(1991)229-240

Karpiskoval, R., Benes, D., Dedicova, D., 1999. Emergence of a multidrug
resistant Salmonella Typhimun‘um DT 104 in the Czech Republic.
Eurosurveillance 4:56-58

Kress-Rogers, E. (ed) 1997. Handbook of Biosensors and Electronic Noses.
CRC, Boca Raton, Florida

Kudva, I T Hatfield, P (3., Hovde c. J. 1997 Characterization of Escherichia coli
0157: H7 and other Shiga Toxin-Producing E. coli Serotypes Isolated from
Sheep. Journal of Clinical Microbiology, 35(4)892-899

Letellier, A., Messier, S., Quessy, S., 1999. Prevalence of Salmonella spp. and
Yersinia enterocolitica in finishing swine at Canadian abattoirs. Journal of
Food Protection, 62222-25

Liedberg, B., Nylander, C., Lundstroem, I. 1983. Surface plasmon resonance for
gas detection and biosensing Sens. Actuat. 4:299-304

Lin, J.S., Tsen, H.Y., 1999. Development and Use of Polymerase Chain Reaction
for the Specific Detection of Salmonella Typhimurium in Stool and Food
Samples. Journal of Food Protection, Vol. 62, No. 10:1103-1110

Malmborg, A.C., Michaelsson, A., Ohlin, M., Jansson, B, Borrebaeck, CK, 1992.
Real time analysis of antibody-antigen reaction kinetics Scandinavian
Journal of Immunology 35:643—650

Masson, L., Mazza, A., Brousseau, R., 1994. Stable immobilization of lipid
vesicles for kinetic studies using surface plasmon resonance Analytical
Biochemistry 218: 405-412

Masson, L., Mazza, A., Brousseau, R., Tabashinik, B., 1995., Kinetics of Bacillus
thuringiensis toxin binding with brush border membrane vesicles from
susceptible and resistant larvae of Plutella xylostella. The Journal of
Biological Chemistry 270(20):11887-1 1896

162

Mead, P.S., Slutsker, L., Dietz, V., McCaig, L., Bresee, J., Shapiro., Griffin, P.,
and Tauxe, R., 1999. Food-related illness and death in the United States
Emerging Infectious Disease, Vol. 5

Medina, MB, 1997. Hygromycin B Antibody Production and Characterization by
a Surface Plasmon Resonance Biosensor Journal of Agricultural Food
Chemistry 452389-394

Middleton, D., Ciebin, 8., Michel, R, Ford, M. 1999. Salmonella Typimurium
definitive type 104 in Ontario, 1997-1998: One example of an
antimicrobial resistant organism. Available at URL: http:llwww.gov.on.ca/
heaIthlenglishlprogramlpubhealthlpherolphero_19991 1 .html

Monk, J.D., Beuchat, L.R., Doyle, MP, 1995. Irradiation inactivation of
foodbome microorganisms. Journal of Food Protection 58(2)197-208

Morse, D.L., Birkhead, G.S., Guzewich, J.J., 1994. Investigating foodbome
disease, in Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, (eds)
Foodborne Disease Handbook, Volume 1: Diseases Caused by Bacteria.
Marcel Dekker, Inc: New York, NY, pp. 547-603

Moutville, T.J., 1987. Food Microbiology. CRC Press, NY p. 643

Natsume, T., Koide, T., Yokota, S., Hirayoshi, K., Nagata, K., 1994. Interactions
between collagen-binding stress protein HSP47 and collagen The Journal
of Biological Chemistry 269(49)31224-31228

Neill, M. A., Tarr, P.|., Taylor, D.N., Trofa, AF, 1994. Escherichia coli in Hui,
Y.H., Gorham, J.R., Murrell, K.D., Cliver, D.O, (eds) Food Disease
Handbook, Volume 1: Diseases Caused by Bacteria. pp. 169-213

O’Shannessy, D.J., Brigham-Burke, M., Soneson, K.K, Hensley, P., Brooks, I.,
1993. Determination of rate and equilibrium binding constants for
macromolecular interactions using surface plasmon resonance: Use of a
nonlinear least squares analysis methods Analytical Biochemistry
21 22457-468

O’Shannessy, D.J., Brigham-Burke, M., Soneson, K.K., Hensley, P., Brooks. l.,
1994. Determination of rate and equilibrium binding constants for
macromolecular interactions by surface plasmon resonance Methods in
Enzymology, 240:323-349

Ogert, R.A., Brown, J.E., Singh, B.R., Shriver-Lake, L.C., Ligler, FS, 1992.
Detection of Clostridium botulinum toxin a using a fiber optic-based
biosensor Analytical Biochemistry 205:306-312

Orskov, I., Orskov, F., Jann, B. 1977. Serology, Chemistry, and Genetics of O
and K antigens of Escherichia coli. Bacteriology Review, 41:667-710

163

Ozeki, M., Furuichi, Y., Umekawa, H., Takahashi, T., 1998. A surface plasmon
resonance assay for the binding of Amaranthus Hypochondriacus var.
Mexico Lectin to glycoprotein. Biochemistry and Molecular Biology
International. 44(1 ):21 1 -216

Pfaff, M., Goehring, W., Brown, J. C., Timple, R. 1994. Eur. Journal of
Biochemistry 225: 975-984

Quantech, 1998. Available at URL: h_ttg:l/www.biosensor.comlguantechlwintr

Ratnam, 8., March, 88., Ahmed, R., Bezanson, S., Kasatiya, S. 1988
Characterization of Escherichia coli Serotype O157:H7. Journal of Clinical
Microbiology, OCt 1988:2006-2012

Roberts, T., Buzby, J.C., Ollinger, M., 1996. Using benefit and cost information
to evaluate a food safety regulation: HA CCP for meat and poultry. J. Agr.
Econ. 78:1297-1301

Rogers, K.R., Gerlach, CL, 1996. Environmental Biosensors, A status report.
Environmental Science and Technology. 30(11)486:491

Ryser, E. R., 2000. Strain specific typing bacterial pathogens in the pork industry
chain. Final report for NPPC project #99-026, August, 2000

Salamon, 2., Brown, M.F., Tollin, G., 1999. Plasmon resonance spectroscopy:
probing molecular interactions within membranes. TIBS 24 — June
1999:213-219

Schuster, S.C., Swanson, R.V., Alex, L.A., Bourret, R.B., Simon, MI, 1993.
Assembly and function of a quaternary signal transduction complex
monitored by surface plasmon resonance. Nature 365:343-347

Scott, A.O. (ed), 1998. Biosensors for Food Analysis, Athenaeum Press Ltd,
Gateshead, Tyne and Wear, UK

Seo, K.H., Brackett, R.E., Hartman, N.F., Campbell, DP, 1999. Development of
a rapid response biosensor for detection of Salmonella Typhimurium.
Journal of Food Protection. 62(5)431-437

Shinohara, Y., Kim, F., Shimizu, M., Goto, M., Tosu, M., Hasegawa, Y., 1994.
Kinetic measurement of the interaction between an oligosacchan'de and
lectins by a biosensor based on surface plasmon resonance European
Journal of Biochemistry 223:189-194

Simmons, NA. 1997. Global Perspective on Escherichia coli O157:H7 and Other
Verocytotoxic E. coli spp.: UK Views. Journal of Food Protection, Vol. 60,
No. 11:1463-1465

164

Stenberg, E., Persson, B., Roos, H., Urbaniczky, C., 1991. Quantitative
determination of surface concentration of protein with surface plasmon
resonace using radiolabeled proteins. Colloid Interface Science
143(2):5‘3-527

Stemesjoe, A., Mellgren, C., Bjoerck, L., 1995. Determination of sulfamethazine
residues in milk by a surface plasmon resonance-based biosensor assay.
Analytical Biochemistry 226:175-181

Stemesjoe, A., Mellgren, C., Bjoerck, L., 1996. Analysis of sulfamethazine in milk
by an immunosensor assay based on surface plasmon resonance. In
Immunoassays for residue analysis, American Chemical Society, 463-470

Sutherland, R., Daehne, C. 1987. IRS devices for optical immunoassay, in
A.P.F. Turner, Karube, L, Wilson, G.S. (eds) Biosensors: Fundamentals
and Applications. Oxford Science Publications, Oxford, pp. 655-678

Tauxe, R.V. 1991. Salmonella: A postmodern pathogen. Journal of Food
Protection 54:563-568

Tempelman, L.A., King, K.D., Anderson, G.P., Ligler, PS, 1996. Quantitating
Staphylococcal enterotoxin b in diverse media using a portable fiber—optic
biosensor Analytical Biochemistry 233250-57

Texas Instruments, Inc. Dallas, Texas, USA. 1999. Operation Manual, Spreeta
Experimenter’s Kit

Turner, A. P., Newman, J. D., 1998. An Introduction to Biosensors, in A. 0. Scott
(ed) Biosensors for Food Analysis, Athenaeum Press Ltd, Gateshead,
Tyne and Wear UK, pp. 13-27

U.S. Meat Export Federation, 1997. Total US. Pork Exports. Available from
URL: httpzllwww.usmef.org/expstathORK8796.html

USDA, 1999. Food Safety: The Agricultural Use of Antibiotics and its
Implications for Human Health. Letter Report, 4I28/99, GAO/RCED-99-
74. Available at URL: http2/lwww.texas-naturaI-beef.com/food_safety.htm

USDA-FSIS, 1998. Report to Congress, FoodNet: An Active Surveillance
System for Bacterial Foodborne Diseases in the United States.
Washington, DC.

Varnam, AH, and Evans, MG. 1991. Foodborne Pathogens. Mosby-Year Book,
Inc., St. Louis, MO

Verozny-Rozand, C., Mazuy., C., Ray-Gueniot, S., Boutrand-Loei, S., Meyrand,
A., Richard, Y., 1997. Detection of Escherichia coli O157:H7 in French

165

food samples using an immunomagnetic separation method and the
VIDASTM E. coli 0157. Letters in Applied Microbiology, 25:442-446

. Wang, T.K., Tsen, T.C., Lee, J.H. 1994. Analysis of Salmonella serovars in
Taiwan by the phase induction method. Chinese Journal of
Microbiological Immunology. 27:13-24

Watts, H., Lowe, C., Pollard-Knight, D., 1994. Optical biosensor for monitoring
microbial cells. Anal. Chem 66:2465-2470

WHO (World Heath Organization), 1997. Multi-drug resistant Salmonella
Typhimurium. WHO Fact Sheet No. 139, Available at URL:
mpzllwwwwho. int/inf-fslenlfactfl .39. html

Woodbury, R.G., Wendin, C., Clendenning, J., Furlong, C.E., Melendez, J.,
Elkind, J., Bartholomew, D., Brown, S., 1998. Construction of biosensors
using a gold-binding polypeptide and a miniature integrated surface
plasmon resonance sensor. Biosensors and Bioelectronics (1998)00-00

Wortberg, M., Orban, M., Renneberg, R., Cammann, K. 1997. Fluorimetric
lmmunosensors. in Kress-Rogers, Ed., Handbook of Biosensors and
Electronic Noses. CRC, Boca Raton, Florida, p. 369-405

WVA (World Veterinarian Association), 1997. Multi-Drug Resistant Salmonella
Typhimurium. WVA Bulletin 14(2). Available at URL: http:/lwww.who.ch

Younts, SM, 1999. Development and Evaluation of a Gas Sensor Based
Instrument for the Detection and Differentiation of E. coli O157:H 7 from
Non-O157:H7 E. coli. Thesis for the degree of MS, Michigan State
University, East Lansing, MI. I“ ' - ‘

Ziprin, R.L., 1994. Salmonella. In Hui, Y.H., Gorham, J.R., Murrell, K.D., Cliver,
D.O (eds) Foodborne Disease Handbook, Volume 1: Diseases Caused by
Bacteria. Marcel Dekker, Inc. New York, NY, pp. 253-318

166