DETECTION OF STAPHYLOCOCCAL ENTEROTOXIN B
BY’ AFFINITY RADIOIMMUNOASSAY

Dissertation for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
NARONG NIYOMVIT
1976

 

This is to certify that the

thesis entitled

Detection of Staphylococcal Enterotoxin B
by Affinity Radioimmunoassay

presented by

Narong Niyomvit

has been accepted towards fulfillment
of the requirements for

[77 D; degree in K‘ :1 / 616/ r :1 a t.

2
fl/flgn (/6;

Major professor

Date 1.4; 2% /:7;

07639

ABSTRACT

DETECTION OF STAPHYLOCOCCAL ENTEROTOXIN B
BY AFFINITY RADIOIMMUNOASSAY

BY

Narong Niyomvit

The objective of this investigation was to develop
a rapid, sensitive method for detection of staphylococcal
enterotoxin B (SEB). Antibody to enterotoxin B was cova-
lently attached to Sepharose 43 by use of cyanogen bromide
activation. The antibody gel was diluted with Sephadex
G-25, and 2 ml packed volume of the diluted gel was placed
in a small column. Assays were carried out by sequen-

12SI-labeled SEB, 0.05 M sodium

tially pumping sample,
phosphate buffer saline (pH 7.5) containing 0.15 M NaCl,
and 0.13 M NH4OH (pH 10.5) containing 0.15 M NaCl through
the column. The amount of 125I-labeled SEB removed from
the column with the pH 10.5 solution was related to the
amount of SEB in the sample. After washing with the

pH 10.5 buffer, the column was ready for reuse. A single
column could be used for over 100 determinations.

In buffer solutions, the minimum SEB concentration

which could be detected was 1.2 ng/ml. When SEB was

Narong Niyomvit

added to reconstituted nonfat dry milk and hamburger, the
minimum detection levels were 2.2 ng/ml and 6.3 ng/g,
respectively. The sample of milk required no treatment
prior to assay. Hamburger required only centrifugation
and filtration through Nucleopore filters to remove fat
from the sample. The results of this investigation
showed that the affinity radioimmunoassay technique which

was developed can be utilized for the detection of SEB.

DETECTION OF STAPHYLOCOCCAL ENTEROTOXIN B

BY AFFINITY RADIOIMMUNOASSAY

BY

Narong Niyomvit

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

1976

DEDICATION

To my parents

ii

ACKNOWLEDGMENTS

The author very sincerely wishes to express his
deepest appreciation to Dr. Roger F. McFeeters for his
excellent guidance, encouragement, and individual assis-
tance throughout the doctoral program toward the com-
pletion of this dissertation.

The same appreciation is extended to Dr. J. R.
Brunner, Dr. K. Stevenson, Dr. L. L. Bieber, and Dr. A. H.
Tucker for their assistance and suggestions.

Last but not least the author wishes to express
his gratitude to the Government of Thailand and Kasetsart
University for providing the leave of absence and financial
support for this program. The Department of Food Science
and Human Nutrition of Michigan State University has pro-
vided ample facilities and excellent cooperation for the

whole program.

iii

TABLE OF CONTENTS

 

 

Page
LIST OF TABLES . . . . . . . . . . . . vii
LIST OF FIGURES . . . . . . . . . . . . viii
INTRODUCTION I O O I O O O O O O O O O 1
LITERATURE REVIEW . . . . . . . . . . . 5
General Characteristics of Staphylococci. . . 5
Characteristics of Staphylococcus aureus. . 5
Distribution of Staphylococcus aureus in
FOOdS O O O O O O O O O O O O O 6
Staphylococcal Enterotoxins . . . . . . . 7
Types of Enterotoxin . . . . 7
Factors Affecting Staphylococcal Entero-
toxin Production in Foods . . . . . . 8
Nature of Staphylococcal Enterotoxins. . . 9
Stability of Staphylococcal Enterotoxins. . 9
Toxic and Antigenic Sites of Enterotoxins . 11
Production and Purification of Staphylo-
coccal Enterotoxins . . . . . . . . 12
Production of Enterotoxin Antibodies . . . . 14
Methods for Detection of Enterotoxins. . . . 15
Biological Methods . . . . . . . . . 15
Serological Methods. . . . . . . . . 16
Precipitation in gel . . . . . . . 16
Precipitate reaction . . . . . . . 19
Radioimmunoassay (RIA). . . . . . . 20
Radiolabelling of Staphylococcal Entero-
tOXin O I O I O I I O O O O O O 2 3
Coupling of Proteins to Agarose Gels . . . 27

iv

MATERIALS AND METHODS . . . . . . .

Materials . . . . . .
Coupling of Enterotoxin Antibody to
Sepharose 4B . . . . . . . .
Iodination of Enterotoxin. . . . .
Preparation of Antibody Column . . .
Measurement of Activity . . . . .
Foods Used in Recovery Study. . .
Inoculation of Reconstituted Nonfat Dry
with Staphylococcus aureus. . .

 

Measurement ofiEnterotox1n B in Reconstituted

Nonfat Dry Milk . . . . . . .

Milk

Measurement of Enterotoxin B in Hamburger

RESULTS 0 O O O O O O I O I O 0

Development of an Affinity Radioimmunoassay

Procedure . . . . . .

Binding of Enterotoxin, Removal of Impurities,
and Removal of Unbound Labelled Enterotoxin

Selection of Dissociating Agent. . .
Improvement of Dissociating Ability of

NH4OH . . . . . . . . . .
Optimum Concentration of NH4OH . . .
Equilibration of the Column . . . .

Time Required for Contact Between Enterotoxin

and Antibody . . . . . . .
Amount of Labelled Enterotoxin Required
the Assay Procedure . . . .

Suppression of Nonspecific Binding. .
Assay Sequence . . . . . . . .
Dose-Response curve 0 o o o o o o

for

Factors Affecting the Dose-Response Curve

Validation of the Assay Procedure . .
Detection of Enterotoxin B in Foods .

Detection of Enterotoxin B in Nonfat
Milk 0 O O O O O O O O 0

Dry

Detection of Enterotoxin B in Hamburger .

Determination of D-Value for Enterotoxin

Denaturation in Reconstituted Nonfat
Dry Milk. . . . . . . . . .
Sensitivity of the Assay . . . . .
Reproducibility . . . . . . . .

Page
30
30

31
32
33
35
35

35

36
36

38

38

41
44

46
46
49

51

52
56
57
59
60
64
65

65
69

69
72
75

Page

DISCUSSION . . . . . . . . . . . . . . 77
CONCLUSIONS. . . . . . . . . . . . . . 91
BIBLIOGMPHY . C C O O O C O O C O C 0 9 3

vi

LIST OF TABLES
Table Page
1. Equilibrating the Antibody Column with PBS . . 51

2. Effect of Contact Time on Binding of
Labelled SE3 0 O O O O O O O O O I 5 3

3. Effect of Contact Time on Binding of SEB to
Its Antibodies in the 1:15 Dilution Anti-

body-Gel Column . . . . . . . . . . 54
4. Suppression of Nonspecific Binding. . . . . 57
5. Assay Sequence for Enterotoxin B . . . . . 58
6. Validation of the Assay Procedure . . . . . 66
7. Sensitivity of SEB Assay . . . . . . . . 74
8. Reproducibility of the Assay. . . . . . . 76

9. Minimal Detectable Level and Approximate Time
for Extraction and Assay of SEB by Various
Quantitative Method . . . . . . . . . 86

vii

I, II it | IN ‘I Ill-Ii

Figure

10.

11.

LIST OF FIGURES

Column used for enterotoxin B determination

Basic concept for affinity radioimmunoassay

Effect of NaCl on elution of unbound labelled

SEB O O O O O O O O O O I 0

Comparison of 8 M urea, pH 3.0; 6 M guanidine-

HCl, pH 3.0; and 2 M NH4OH, pH 11.5 for

removal of bound 125I-enterotoxin from anti-

body column . . . . . . . . . .

Comparison of elution patterns for removal of

bound labelled SEB with pH 11.5, 2 M NH4OH
with and without NaCl added. . . .

Elution patterns for removal of bound labelled

SEB with different concentrations of NaCl

Counts of controls run repeatedly when dif—

ferent concentrations of NH4OH were used as

dissociating agent. . . . . . . .

Relationship between the amount of bound and
added labelled SEB in the column 1:15
dilution antibody-gel. . . . . . .

Standard curve of SEB in PBS, showing the
relationship between the enterotoxin level
and counts per minute. . . . . . .

Standard curve of SEB in PBS, showing the
relationship between the enterotoxin level
and percentage of initial counts . . .

Effect of gel dilution and concentration of
labelled SEB on the dose-response curve.

viii

Page

34

39

43

45

47

48

50

55

61

62

63

Figure Page

12. Standard curve of SEB in reconstituted nonfat
dry milk, showing the relationship between
the enterotoxin level and counts per
minute . . . . . . . . . . . . . 67

13. Standard curve of SEB in reconstituted nonfat
dry milk, showing the relationship between
the enterotoxin level and percentage of
initial counts . . . . . . . . . . 68

14. Standard curve of SEB in hamburger, showing
the relationship between the enterotoxin
level and counts per minute . . . . . . 70

15. Standard curve of SEB in hamburger, showing
the relationship between the enterotoxin
level and percentage of initial counts . . 71

16. Heat inactivation of SEB in reconstituted
nonfat dry milk at 121 C, pH 5.2 . . . . 73

17. Comparison of standard curves of SEB in PBS,
reconstituted nonfat dry milk and hamburger,
showing the difference when the plots are
based on counts per minute . . . . . . 82

18. Comparison of standard curves of SEB in PBS,
reconstituted nonfat dry milk and hamburger,
showing the difference when the plots are
based on percentage of initial counts. . . 84

ix

INTRODUCTION

The acute illness caused by staphylococcal entero-
toxins is fairly common in the United States and in other
countries of the world, especially in the developing
areas. Five enterotoxins, designated types A, B, C, D,
and B, have been identified. They are produced by various

strains of Staphylococcus aureus. Man is considered the

 

main reservoir of enterotoxic staphylococci. They are
frequently inhabitants of the skin and mucosal surfaces
of the upper respiratory tract. In fact, the nasal pas-
sages constitute the principle sites of bacterial multi-
plication from which the organism can be transferred
readily to food by sneezing. Food handlers with skin
lesions or contaminated hands may also innoculate food.

The number of Staphylococcus aureus organisms

 

may increase during processing, storage, and distribution
of foods. The extent of such growth depends upon a
variety of environmental conditions. Some of the more
important environmental factors are temperature, length
of time held at any given temperature, and associated

growth of other organisms. Growth of Staphylococcus

 

aureus to high populations may lead to production of
enterotoxin and subsequent food poisoning. The entero-
toxins are resistant to heat and low pH. In general,
enterotoxins formed in food products will not be inacti-
vated by subsequent processing.

Convenient methods for detection of staphylococcal
enterotoxin in food are important to regulatory agencies
and the food industry in order to prevent contaminated
food from reaching consumers. Improved detection can‘
also aid research into the mechanisms of enterotoxin
production in food and evaluation of new techniques to
prevent enterotoxin formation. Considerable efforts have
been made to develop rapid, sensitive methods for measure-
ment. The method used for detection must be sensitive
enough to assure safety. A method which can reliably
detect 0.1—0.2 pg/lOO g of food is considered adequate
since it is thought that a minimum of 1 pg of enterotoxin
is required to produce symptoms in man (Bergdoll et al.,
1976). If enterotoxin is detectable in a food at any
level, it is not considered safe for consumption.

Several methods to determine the presence of
enterotoxin in foods have been developed. The biological
assays available are not sufficiently sensitive to detect
the small amounts of enterotoxin usually present in foods.
The first successful method for detection of enterotoxin

in food was the immunodiffusion method developed by

Casman and Bennett (1965). The method has been success-
ful in determining if food contains enterotoxin. How-
ever, the sensitivity of the method is relatively low,
it is laborious, and the cost per determination is high.
Several investigators have attempted to modify this
method to shorten the time required to obtain results
and to make it equally sensitive for all types of food
(Bergdoll et al., 1976). In recent years new methods for
detection of enterotoxin have been developed. These
include solid-phase radioimmunoassay (RIA) (Johnson et
al., 1971; Collins et al., 1973), and reversed passive
hemagglutination assay (RPHA) (Silverman et al., 1968).
However, a high degree of operational skill is required
to perform these assays and the mechanical manipulation
required is not readily automated. Solid-phase radio-
immunoassays are sufficiently sensitive for detection
of enterotoxins so that the high degree of purification
and concentration required by the Casman and Bennet pro-
cedure is not necessary. However, it was thought that
a new approach to solid-phase radioimmunoassays could
provide significant improvements over previously used
procedures.

The objective of this project was to investigate
an alternative approach to radioimmunoassay of staphylo-
coccal enterotoxin B. This approach provides the

following advantages: (1) multiple use of antibody,

(2) relatively rapid analysis compared to the Casman and
Bennett procedure, and (3) a system that is better
adapted to automated analysis than conventional radio-

immunoassay techniques.

LITERATURE REVIEW

General Characteristics of
Staphylococci

 

Staphylococci have long been known to be associated
with man and animals. Although they can initiate a wide
variety of infections, food poisoning caused by staphylo-
cocci is strictly an intoxication. Symptoms of staphylo-
coccal food poisoning begin 1-6 hours after ingestion of
contaminated foods and consist initially of salivation
and nausea. This is quickly followed by vomiting,
abdominal cramping, and diarrhea. The mortality rate
is very low or nil (Jay, 1970).

Characteristics of Staphylo-
coccus aureus

 

 

The bacteria causing staphylococcal food poison-

ing are classified under the name of Staphylococcus

 

aureus. Only certain strains of Staphylococcus aureus

 

can produce enterotoxin. Cultures of Staphylococcus

 

aureus are facultatively anaerobic, gram positive, non-
smotile, nonphotosynthetic, and do not form spores. They
typically occur in grape-like cluster and the individual

cells are approximately 0.8 to 1.0 pm in diameter.

Staphylococcus aureus is mesophilic in nature and can

 

grow at a temperature as low as 6.7 C (Angelotti et al.,
1961). In general, these organisms have an optimum
growth temperature around 37 C. They are capable of
growth over a pH range of 4.5 to 8 with an optimum pH

of 4.8 to 7.6 (Nickerson & Sinskey, 1974). These
organisms also possess a high degree of tolerance to
compounds such as tellurite, mercuric chloride, neomycin
and sodium azide, all of which have been suggested agents
for isolation of staphylococci.

Distribution of Staphylococcus
aureus in Foods

 

 

Staphylococcus aureus have been found in many

 

types of commercial foods. Silverman et a1. (1961)
recovered staphylococci from commercial shrimp. Walker
et al. (1961) also isolated these organisms from cheese
and showed that they died off during the normal aging of
these products. These organisms have also been isolated
from cheddar cheese ranging from 50 to 220,000 per gram
(Donnelly et al., 1964). Coagulase-positive staphylo-
cocCi have been found in vegetables, market meat, poultry
and other foods of animal origin (Munch-Peterson, 1963;
Splittstoesser et al., 1965).

The type of foods most frequently cited as having
caused staphylococcal intoxications are ham and ham

products, chicken products, especially chicken salad,

potato salad (Munch-Peterson, 1963), vegetable products
(Splittstoesser et al., 1965) and bakery goods such as
eclairs, which have custard fillings. Milk and ice cream
have sometimes been implicated as vehicles of staphylo-

coccal poisoning.

Staphylococcal Enterotoxins

Types of Enterotoxin

 

Five immunologically distinct types of enterotoxin
have been identified and designated with letters A
(Casman, 1960), B (Bergdoll et al., 1959a), C (Bergdoll
et al., 1965), D (Casman et al., 1967), and E (Bergdoll

et al., 1971). Some strains of Staphylococcus aureus

 

produce only one type of enterotoxin such as Staphylo-

coccus aureus 196E (ATCC 12565) which produces enterotoxin

 

A (Casman et al., 1963) and Staphylococcus aureus 243

 

(ATCC 14458) which produces exclusively enterotoxin B.
These are the type strains for these enterotoxins.
Others produce two or more enterotoxins. Two different
enterotoxin C's from different staphylococcus strains
were found (Borja & Bergdoll, 1967; Avena & Bergdoll,
1967) and classed as Cl and C2 on the basis of their
different isoelectric point, 8.6 and 7.0, respectively.
In the study of staphylococci isolated from 305 cultures
from outbreaks, 49% were enterotoxigenic, with 62.4% of
these produced enterotoxin A, 17.4% enterotoxin C,

11.4% enterotoxin B, and 8.7% A, B, or C in combination

(Hall, 1968). Type B was the first to be purified
(Schantz et al., 1965). Most experimental toxicity data
have been derived from work with this enterotoxin due

to the fact that it is the one most easily obtained.
Relatively little information exists on type D and E
which were the latest to be identified.

Factors Affegting Staphylococcal
Enterotoxin Production in Foods

 

 

It has been found that large numbers of Staphylo-

coccus aureus organisms must be present in food to cause

 

enterotoxin production. It has been estimated that l to

4 ug of type A enterotoxin would be required to cause
symptoms (Casman & Bennett, 1965). Foods which caused
staphylococcal poisoning were found to have staphylococcal
counts of 50 x 106 to 200 x 106 per gram of food. The
number necessary to produce enough enterotoxin to induce
food poisoning depends on such factors as the nature of
the food substrate, its pH, holding temperature, and the
characteristics of the strain in question. The number and
type of competing microorganisms present were very impor-
tant. It was shown that staphylococci were unable to

compete well with normal food-borne bacteria in both

fresh and frozen foods (Troller & Frazier, 1963).

Nature of Staphylococcal
Enterotoxins

 

 

All enterotoxins purified to date appear to be
simple proteins that contain relatively high levels of
lysine, aspartic acid, glutamic acid and tyrosine among
the 18 different amino acids making up the peptide chain.
Enterotoxin B was found to contain 239 amino acids resi-
dues with a molecular weight of 28,366 (Huang & Bergdoll,
1970b). The sequence has been determined by Huang and
Bergdoll (1970b). The molecular weights of enterotoxin A,

C and C2 are 34,700 (Chu et al., 1966), 34,100 (Borja &

1»
Bergdoll, 1967), and 34,000 (Avena & Bergdoll, 1967),
respectively. The study of amino acid sequences of these
enterotoxins is currently in progress (Robern et al.,
1975).

Stability_of Staphylococcal
Enterotoxins

 

 

Staphylococcal enterotoxins are unique among
bacterial toxins in that they are very heat resistant.
Davison and Dack (1939) found the potency of entero-
toxin could be only gradually decreased by prolonged
boiling or autoclaving.

Purified enterotoxin A is relatively heat labile
compared to enterotoxin B. A decrease of 50% in the
reaction of enterotoxin A with its specific antibody was
observed when it was heated at 60 C in 0.05 M sodium

phosphate buffer, pH 6.85, for 20 minutes (Chu et al.,

10

1966). This result could not be compared to those
obtained by Denney et al. (1966). They found that heat
inactivation at this temperature for enterotoxin A,
based on cat emetic response, to be 11 minutes (F320 =
11 min.) or Fggo = 8 min. when monkeys were employed.
The heat inactivation at 60 C was believed to be due to
enterotoxin aggregation. Hilker et a1. (1968) reported
that crude enterotoxin A, 21 ug/ml in veronal buffer,
pH 7.2, required heating at 100 C for 130 minutes to
reduce the enterotoxin concentration to less than 1
ug/mlo

The biological activity of enterotoxin B was
retained after heating 16 hours at 60 C at pH 7.3
(Schantz et al., 1965). It could not be inactivated
by pasteurization or spray drying of milk as currently
practiced (Read & Bradshaw, 1966). Satterlee and Kraft
(1969) showed that enterotoxin B was more heat sensitive
at 80 C than at 100 or 110 C. They also showed that
thermal inactivation was more pronounced at 80 C than
at either 60 or 100 C when heated in the presence of
meat proteins. Reed and Bradshaw (1966) found that raw
milk, containing 30 ug of SEB per ml, needed 37.6 minutes
to inactivate 99% or more of SEB (F = 37.6 min.) at
121 C.

The biological activities of enterotoxin C did

1
not change when it was heated at 60 C for 30 minutes, but

11

the solution became turbid when heating was continued for
60 minutes (Borja & Bergdoll, 1967). The reaction of
enterotoxin C2 with its anti-enterotoxin was reduced to
about 20% of normal when it was heated at 100 C for

1 minute (Avena & Bergdoll, 1967).

In summary, enterotoxin A is the most sensitive
to heat, while enterotoxin B is the least sensitive.
Purified enterotoxins are more sensitive to heat than
the crude toxins.

Also, native enterotoxins were resistant to pro-
teOlytic enzymes such as trypsin, chymotrypsin, rennin,
and papain. Pepsin destroys their activity at a pH of
2, but is ineffective at higher pH values (Bergdoll,
1966; Chu et al., 1966).

Toxic and Antigenic Sites
of Enterotoxins

 

 

Enterotoxins contain only two residues of half-
cystine which are cross-linked to form a cystine residue.
The half-cystine residues in enterotoxin A and B are at
positions 92 and 112 (Huang & Bergdoll, 1970a) which
creates a cystine loop of 20 amino acid residues. It
is believed that this cystine 100p is involved in the
activity of the enterotoxin (Dalidovicz et al., 1966;
Bergdoll & Robbins, 1973). It was suggested that amino
acids adjacent to cysteine at position 112 may be

involved in the active site because these amino acids

12

are alike for enterotoxin A and B (Bergdoll & Robbins,
1973). The residues near to the half-cystine residue
at position 92 in enterotoxin may comprise the major
antigenic site since there are several tyrosyl residues.
Mbst lysyl residues, which are evenly distributed on
the surfaces of molecule, are not thought to be involved
in either the toxin or antigenic sites (Spero et al.,
1971). The biological activity of enterotoxin A was
found to be destroyed when tyrosyl residues were mod-
ified (Chu et al., 1966). The normal positive charge
of the enterotoxin is thought to play an important role
in both emesis and in the reaction of enterotoxin with
its specific antibody (Jay, 1970).

Production and Purification of
Staphylococcal Enterotoxins

 

 

Enterotoxin may be readily produced in the
laboratory by cultivation of toxigenic strains in Brain
Heart Infusion broth (Bergdoll, 1970).

Several methods have been proposed for production
of enterotoxins. These methods include the sac-culture
technique of Casman and Bennett (1963), the semisolid
plate technique of Casman and Bennett (1963), and
cellophane-over-sold-agar-plate of Hallander (1965).
Factors that affect the enterotoxin production such as

pH, incubation temperature, water activity, salt, sodium

13

nitrate, glucose, streptomycin, acriflavine and Tween
80 were reviewed by Bergdoll (1970).

The appearance of enterotoxin in a culture
medium was studied by Lilly et al. (1967). They reported
that enterotoxin B appeared in the culture medium as
early as 6 hours and increased through the stationary
phase. This was confirmed by Morse et a1. (1969) who
also observed that pH of the medium increased during the
releasing period. They suggested that enterotoxin was
not preformed and released by change in pH but rather by
the result of catabolic repression. Markus and Silverman
(1968, 1969) worked with Staphylococcus aureus and found
that 95% of enterotoxin B was released during the latter
part of the exponential phase of growth. They showed
that the presence of enterotoxin was dependent upon
dg,ppyg_protein synthesis.

Enterotoxin A was found to be a primary metabolite
for it was secreted mainly in the exponential phase of
growth (Markus & Silverman, 1969).

Purification of enterotoxin B was first done by
Bergdoll et al. (1959b). Schantz et al. (1965) developed
a simplified purification procedure which resulted in a
50—60% recovery of enterotoxin B.

The purification of enterotoxin A is somewhat

different from those used for enterotoxin B. A method

14

for purification of enterotoxin A with 30-35% yield of
toxin was developed by Chu et a1. (1966).

Casman et a1. (1967) used the method developed
by Schantz et a1. (1965) to purify enterotoxin D. The
authors suggested that highly purified enterotoxin D
was obtained, but no proof was presented. Food Research
Institute laboratories purified enterotoxins C, D, and
E by methods similar to those employed for A (Bergdoll,
1970).

Production of Enterotoxin
Antibbdies

 

 

The preparation of antibody for each enterotoxin
may be different in some detail by each investigator.
Casman and Bennett (1964) prepared antibody to entero-
toxin A by injecting the rabbits at weekly intervals,
using the following amounts: 0.18, 0.36, 0.72, 2.8,
5.6, 11.2 ug intravenously, 400 ug subcutanously, and
also 88, 260 and 1760 pg mixed with aluminum phosphate.
After an interval of 1 month, the rabbits were injected
intravenously with 500 ug of enterotoxin A and after 7
days, bled from a marginal ear vein. The injection was
repeated three times at monthly intervals, removing
approximately 60 m1 of blood at each bleeding.

Nace and Spradlin (1962) prepared antibody to
enterotoxin B by injecting a series of purified entero-

toxin B, 0.005, 0.01, 0.04, 0.2 and 1.0 mg, at

15

approximately 1-week intervals. After five weeks, 2 to

3 mg enterotoxin were injected and bleeding began 8 days
later. Four to six bleedings of approximately 50 ml each
over a 3-week period were made from an ear vein. Booster
injections of 2 to 3 mg of enterotoxin can be given

every 5 weeks indefinitely with 4 to 6 bleedings after
each injection.

Methods for Detection of
Enterotoxins

 

 

Enterotoxin detection in foods is still a problem
for food investigators because of the lack of sensitive
methods for the detection and assay. The current pro-
cedures for detection of staphylococcal enterotoxins

are laborious and time consuming.

Biological Methods

 

Kittens (Hammon, 1941) and young rhesus monkeys
(Surgalla et al., 1953) have been used successfully in
many laboratories for detecting enterotoxin in culture
filtrates. The cat method requires the inactivation of
substances that may give symptoms similar to those caused
by enterotoxin. The most reliable biological assay of
the enterotoxin is the feeding of young rhesus monkeys
because only enterotoxins cause emesis in these animals.
However, the high cost of the animals and the fact that
they become resistant to the enterotoxin after a few

tests severely limits use of this procedure.

16

A number of studies have been undertaken to
identify enterotoxin with biological activities other
than emetic action on animals. For example, Milone
(1962) investigated the effect of enterotoxin on tissue
cultures and Schaeffer et al. (1966) looked for effects
on human embryonic intestinal cells. Both studies

showed negative results.

Serological Methods

 

Enterotoxins are proteins which can induce anti-
bodies when injected into animals. The enterotoxins can
then form a precipitate when mixed with their specific
antibodies. This reaction is the basis for serological
techniques for enterotoxin detection. The precipitate
formed can be observed either in a solution (precipitation
reaction) or in a gel (gel diffusion technique). Although
the reaction of enterotoxin with anti-enterotoxin does
not necessarily indicate biological activity of entero-
toxins, in most instances the correlation between the two
is sufficient to justify using the immunological reaction

to assay for enterotoxin (Bergdoll, 1970).

Precipitation in gel. The first precipitation

 

technique for enterotoxin detection was introduced by
Surgalla et a1. (1952). However, the sensitivity of the
method, 1-2 Ug/ml, was too low for detection of entero-

toxin in foods.

17

Bergdoll et al. (1959b) introduced another method
called double gel diffusion tube. The method involves
the placing of a layer of agar between the antibody-agar
and antigen layers. The antibody and enterotoxin will
diffuse toward each other through the agar gel and form
a precipitate line when they come into contact. The
position of the line is dependent on the relative con-
centration of the enterotoxin and antibody. The assay
can detect enterotoxin as low as 0.05 ug/ml. However,
the time required to detect a precipitate with this small
amount is up to 21 days (Hall et al., 1965).

Bergdoll et a1. (1965) used an Ouchterlony Plate
technique to detect enterotoxin in unknown materials.
Wells were made 2 to 4 mm apart. Antibody is placed in
the center well and control and unknown enterotoxins are
alternated in the outer wells. The method is useful for
detection of 5 to 20 HQ/ml enterotoxin, but the smaller
amounts require 2 to 3 days to obtain precipitation lines.

Casman and Bennett (1963) developed a microslide
technique for enterotoxin detection. It required 1 to 3
days to develop precipitin lines. A minimum of 0.1 ug/ml
of enterotoxin could be detected. In order to detect
0.1 to 0.2 ug of enterotoxin per 100 g of food, entero-
toxin in the food extract must be partially purified and
concentrated. The first method for purification and con-

centration of enterotoxin from food was developed by

18

Casman and Bennett (1965). The method involved a rather
lengthy procedure requiring several days to reduce the
extract to 0.1 to 0.2 ml. However, the microslide tech-
nique was an effective procedure for detecting enterotoxin
in foods at sufficiently low concentrations which would
cause illness. Zehren and Zehren (1968), for example,
successfully identified contaminated cheese vats so that
enterotoxin-free cheese could be sold. Reiser et a1.
(1974) modified this microslide method so that the result
could be obtained on the third day. They used Amberlite
CG-SO ion exchanger to absorb enterotoxin directly from
the partially treated food extract (extraction at pH 4.5
then, fat was removed with CHC13, pH 7.5). This elimi-
nated an overnight concentration step. Trypsin was used
to eliminate the interfering contaminants. The micro-
slides were held at 37 C to reduce developing time.

In the Laurell electroimmunodiffusion procedure
(Gasper et al., 1973) antiserum was mixed with agarose
gel. Samples of 1.5 to 10 ng of enterotoxin in 10 ul
buffer (0.15 to 1.0 ug/ml) were transferred from 4 mm
wells into gel by electrophoresis. The precipitation
cones formed were made visible by staining with thiazine
red in acetic acid. The method was not sensitive enough
for quantitative measurement of enterotoxin present in
foods because it fails to produce visible precipitin lines

at sufficiently low concentrations. The method also

19

required 1 to 8 hours for electrophoresis and another 2 to

5 hours for developing a visible cone.

Precipitate reaction. Several methods have been

 

developed to detect enterotoxin directly in food extracts
so that the concentration, purification, and microslide
diffusion steps could be eliminated.

A fluorescent antibody technique was introduced
by Genigeorgis and Sadler (1966). Specific antibodies
to enterotoxin were conjugated with fluorescin isothio-
cyanate and used to precipitate the enterotoxin in
samples. The precipitate was collected on Millipore
filter membranes, washed, and impression smears were
made on a slide. The sensitivity of this method was
about 1 ug/ml.

Morse and Mah (1967) developed a hemagglutination
inhibition test. Enterotoxin was coupled with sheep
erythrocytes. The sensitized erythrocytes were added
to various combinations of enterotoxin—antibody. Hema-
gglutination occurred only if enterotoxin was insufficient.
The method was as sensitive as microslide test, but it
still required several hours for the reaction to be com-
plete.

Silverman et a1. (1968) proposed the reversed
passive hemagglutination method for determination of
enterotoxin in foods. The procedure was based upon the

adsorption of enterotoxin antibody to tannic acid treated

20

sheep red blood cells followed by agglutination of the
sensitized cells with enterotoxin. The sensitivity of
the method was 0.1 ug/100 g of food without concentration
of the extract. It required 1 to 2 days to obtain the
result. There were two major problems for this method.
First, the antibody in some preparations was not adsorbed
or coupled to red blood cells. Secondly, nonspecific
agglutination of sensitized cells by food extracts some-

times occurred, particularly in meat extracts.

Radioimmunoassay (RIA). The technique of radio-

 

immunoassay was first developed for measurement of hor-
mones (Berson et al., 1956), but its use has been expanded
rapidly during the past decade to other areas (Skelley

et al., 1973). The procedure was based on the competi-
tive inhibition by unlabelled hormone of the binding of
labelled hormone to a specific antibody (Yalow & Berson,
1959). The competition required at least two reagents,

a specific antibody and labelled antigen. The reaction
complex formed between antibody and antigen was almost
always soluble at the concentrations employed. There-
fore, it was necessary to separate bound from free antigen
after reaction. Separation techniques included electro-
phoretic and chromatoelectrophoeretic methods (Yalow &
Berson, 1961; Hunter & Greenwood, 1964), gel filtration
(Genuth et al., 1965), nonspecific precipitation of

Ag*Ab (Grodsky & Forsham, 1960; Odell et al., 1965),

21

immunoprecipitation of Ag*Ab complex (Skom & Talmage,
1958), solid-phase adsorption of Ag* (Herbert et al.,
1965; Lau et al., 1966) and solid-phase antibodies
(Isojima et al., 1970; Catt et al., 1968). For some
compounds, a number of different RIA procedures have

been developed that differed primarily in the technique
used to separate bound from free antigen. The physical
manipulations and the care required to maintain an
accurate RIA technique made it essential to have skilled
technicians for these assays. In addition, the incubation
of samples often required considerable time. Catt et al.
(1967) attached antibody to an insoluble polymer such as
polystyrene, or bromoacetyl cellulose. The use of anti-
body in this procedure allowed rapid removal of the free
radioactive antigen by washing the solid phase with water
after the completion of immune reaction. In principle,

a single separation procedure would serve for all radio-
immunoassays. The technique became known as the “solid-
phase radioimmunoassay."

Johnson et a1. (1971) used the radioimmunoassay
for quantitative measurement of crude and purified entero-
toxin B. A polystyrene tube was used as a supporting
medium. The tubes are first sensitized with antibody,
followed by incubation with purified or crude enterotoxin

125

and I-enterotoxin B mixture. After a 4-hour incubation

at 37 C, the buffer was added and the supernatant was

22

counted to determine the amount of unbound labelled
enterotoxin. From the relationship of unbound entero-
toxin and amount of standard enterotoxin, a standard
curve was obtained.

Collins et a1. (1972) used the same inhibition
test for detection of enterotoxin B, but bromoacetyl-
cellulose (BAC) was used instead of polystyrene tubes.
The method is somewhat different from those of Johnson
et a1. (1971). The antibody-BAG must be first diluted

in order to bind 50% of fixed amount of 125

I-enterotoxin,
followed by incubation with the mixture of unlabelled

and labelled enterotoxin for 15 minutes at room temper-
ature. Then, the inhibition curve is obtained. Collins
et a1. (1973) used this method for detection of entero-
toxin A. Enterotoxin A at the level of 0.01 ug/ml could
be detected in a variety of media such as ham, milk pro-
ducts, crab meat, and custard. No interference was found
with any media or food products tested.

Later, Bukovic and Johnson (1975) used solid-
phase radioimmunoassay for enterotoxin C in milk, cheddar
cheese, custard, and ham salad. The assay was sensitive
to 1 to 10 ng of enterotoxin per g of food.

A comparative study of serological detection
methods for enterotoxin was done by Bennett et al. (1973).

The result of the study is as follows: (1) The RPHA

method lacks the desired specificity and is limited by

23

its reproducibility and variations in test interpretation.
It also requires high quality antigen. (2) The micro-
slide test and radioimmunoassay are very consistent.

The microslide method does not require purified antibody,
while purified enterotoxin is important in the successful
interpretation of radioimmunoassay results.

Radiolabelling of Staphylo-
coccal Enterotoxin

 

 

The requirement for radiolabelling of enterotoxin
is that high specific activity be achieved without loss
of antigenic reactivity. This means that the labelling
procedure must be mild enough to maintain the conformation
of functional groups of enterotoxin. In addition, the
intrinsic radioactivity of the labelled antigen should
not cause rapid loss of activity. Several methods have
been introduced. Most of them used 125I or 131I. 1251
has been used extensively in labelling protein because of
its high specific activity, long half life (60 days),
and relatively low cost. It also emits low energy gamma
rays and absence of B-radiation diminishes the potential
for autodestruction of the labelled antigen (Galskov,
1972).

For iodination of enterotoxin, many factors must

be considered. These include (1) damage to the labelled

enterotoxin incured during labelling and purification

24

and (2) damage during the subsequent storage (Skelley
et al., 1973).

The iodination is easy to perform. It is usually
carried out at room temperature at pH 7.5 in phosphate
buffer. It usually requires about 20-30 minutes for
iodination, and a separation of the iodinated protein
(Skelley et al., 1973). For iodination, enterotoxin
must be highly purified. The reaction time, amount of
oxidizing agent and pH must be optimized and the separa-
tion of labelled enterotoxin must be done immediately
after labelling.

Iodine monochloride (McFarlane, 1958), lacto—
peroxidase (Marchalonis, 1969) and chloramine-T (Hunter
& Greenwood, 1962) have been used as oxidizing agents.
Iodine monochloride does not work well with less than
2-5 mg of protein (Freeman, 1959). Lactoperoxidase was
used successfully to iodinate immunoglobulin to low spe-
cific activity with radioactive iodine (Marchalonis,
1969). Thorell and Johansson (1971) used this method
to iodinate insulin and found that the damage was very
low. Chloramine-T has been widely used, especially when
high specific activities are required (Freeman, 1967).
The method is suitable for small quantities of protein
and is so rapid that the damage to protein due to

radiation is minimized.

25

Upon the addition of chloramine-T, the redox
potential rises slowly. The potential determines the
ratio of 12 to I- which in turn determines the amount
of HZOI+ ion. The available HZOI+ ion then replaces H+
ion on a tyrosine ring to form mono- or di-iodotyrosine.
This replacement occurs without the alteration of protein
(Greenwood & Hunter, 1963). At the end of the reaction
period, the potential is lowered to the starting level
by sodium metabisulphite. A carrier, bovine serum
albumin (BSA), is usually added to the reaction mixture
immediately after sodium metabisulphite.

Purification is almost always necessary in order
to free enterotoxin from unreacted iodine and from
damaged components (Skelley, et al., 1973). This step
must be done immediately following iodination, otherwise,
excessive adsorption on glassware may occur. The addition
of BSA following iodination aids in purification as well
as in preventing enterotoxin loss by adsorption. The
purification can be done by gel filtration (Johnson et
al., 1971) or by dialysis (Collins et al., 1972). The
gel filtration method has advantages of rapidity and is,
therefore, the method of choice.

Chloramine-T was used to iodinate enterotoxins
A and B successfully by Johnson et a1. (1971) and Collins
et a1. (1972). Chloramine-T was found to have an adverse

effect on the stability of some enterotoxin A preparations.

26

Collins et a1. (1972) also investigated the microdiffusion
method developed by Gruber and Wright (1967), using
gaseous 1251, and found no deleterious effects on the
enterotoxin A molecule because there was no direct con-
tact between reactants and protein. It yields a product
of lower specific activity than those iodinated with
Chloramine—T. Labelled enterotoxin could be used for
approximately 2 months without changing its antigenicity.
Kauffman and Johnson (1975) studied the stability
of 125I-labelled enterotoxins A, B, and C at two specific
activities, 4 and 40 mCi per ug of protein. They found
that enterotoxin with high specific activity showed
extensive dissociation of 1251 when stored at different
temperatures, including -23 C. In comparison, entero-
toxin with low specific activity did not show a signifi-
cant loss of 1251 when stored at -23 C for as long as
2 months. Enterotoxins with low or high specific activity
formed aggregates immediately upon labelling. Aggregate
formation increased in high specific activity enterotoxins
during storage at -23 C. Enterotoxin with low specific
activities showed no significant increase in aggregate
formation, even after 2 months at -23 C. The aggregate
forms of the enterotoxin were either devoid of antigenic
activity in solid-phase radioimmunoassay or they pos-

sessed significantly reduced antigenic activity. There-

fore, if the labelled enterotoxin is to be stored for a

27

considerable time, it should be quickly frozen and stored

at -23 C or lower.

Coupling of Proteins to
Agarose GeIs

 

 

Methods for coupling of enzymes, antigens, anti-
bodies, and other biologically active compounds to solid
supports, especially agarose, have been greatly expanded
in recent years (Cuatrecasas & Anfinsen, 1971). These
developments have made affinity chromatography a major
technique for protein purification.

The solid supports used for this purpose may be
a hydrophobic polymer, such as polystyrene, or hydro-
philic, such as cellulose or cross-link dextran. Agarose
is considered to be a good supporting medium because it
exhibits good flow rate properties which are retained
after coupling. It is also mechanically and chemically
stable to the conditions of coupling and to varying con-
ditions of pH, ionic strength, temperature, and presence
of denaturants which are needed for adsorption and elution
(Cuatrecasas & Anfinsen, 1971).

A method for coupling protein or peptides to
carbohydrates was introduced by Axen et a1. (1966).

The method involves two steps: (1) the formation of
reactive intermediate by treating a polysaccharide for
a short period with aqueous cyanogen bromide under alka-

line conditions and (2) coupling the intermediate with

28

protein or peptides in neutral or preferably slightly
alkaline aqueous medium.

Porath et al. (1967) activated agarose at pH 11
by using 25 mg of cyanogen bromide and 50 g of agarose.
The activated product was then used for coupling by
adding 0.1 M sodium hydrogen carbonate and protein.

The mixture was stored at 5 C for 20 hours with a con-
stant stirring. Cuatrecasas (1970) modified this method
by using 50 to 300 mg of cyanogen bromide per m1 of
packed Sepharose. The reaction was run for 8 to 12
minutes at 20 C, pH 11. It was recommended that the
amount of protein should be 20 to 30 times greater than
the amount which was to be coupled.

This activation procedure has some disadvantages.
It lacks a well-defined end point and is difficult to
reproduce. In addition, wide fluctuations in pH are
difficult to avoid during titration. Addition of
cyanogen bromide as a solid requires that this volatile
lacrimatore be weighed in a closed vessel in fume hood.
March et a1. (1974) activated Sepharose by adding
cyanogen bromide, dissolved in acetonitrile, to beads
suspended in a solution of sodium carbonate. The neces—
sity for titration and the use of a pH meter during the
reaction were eliminated. Activation for 1 minute
resulted in coupling capacities comparable to those

obtained with the titration method.

29

Porath et a1 (1967) found that a protein-agarose
gel was very stable. No loss in activity of an enzyme
occurred after storage in 0.01 M acetate buffer, pH 4.1
for 2 months at room temperature. This gel also tolerated
relatively extreme conditions such as 0.1 M NaOH and
1 M HCl for 2-3 hours, 6 M guanidine-HCl and 8 M urea
solution for prolonged periods (Cuatrecasas, 1970).

This degree of stability allows the use of denaturants
to elute specifically bound protein or to wash a column

in preparation for reuse.

MATERIALS AND METHODS

Materials

 

Staphylococcal enterotoxin B (SEB) was obtained
commercially from Makor Chemicals Ltd., Jerusalem, Israel.
The powder contained about 2 mg sodium phosphate per mg
of protein and was readily soluble in aqueous solutions.
Polyacrylamide electrophoresis at pH 4.5 showed a major
band that accounted for about 95% of the total protein
and one minor band. The supplier claimed that the SEB
was at least 95% pure.

The antibody to SEB was also obtained from Makor
Chemicals Ltd. It was in a powder form, which was reported
to contain 0.25 mg of antibody per mg of solid.

Other chemicals used were all reagent grade.

NaCl, KSCN, KI, NaZSZOS’ and urea were obtained from
Mallinckrodt Inc., St. Louis, Mo. Bovine serum albumin
(BSA), Sephadex 43-200 and Sephadex G-25-150 were obtained
from the Sigma Chemical Co., St. Louis, Mo. Chloramine-T

was purchased from Eastman Kodak Co., Rochester, N.Y.

30

31
Coupling of Enterotoxin Antibody
to Sepharose 43

The coupling of SEB antibody to Sepharose 4B was
done by the method of March et a1. (1974). Five grams
of Sepharose 43 were washed with deionized water 2 to 3
times to remove preservative. The washed Sepharose was
then mixed with 5 ml of water and 5 ml of 2 M sodium
carbonate. The mixture was stirred slowly at the begin-
ning but more vigorously when 0.25 ml of an acetonitrile
solution of cyanOgen bromide was added (2 g CNBr in 1 ml
of acetonitrile). The slurry was stirred for 2 minutes
and filtered under vacuum on a coarse-sintered funnel.
The activated gel was washed in succession with 50 to
60 ml each of 0.1 M sodium bicarbonate, pH 9.5, water
and 0.05 M sodium bicarbonate buffer, pH 7.5. The washed
gel was transferred into a plastic bottle containing
100 ug of SEB antibody dissolved in 5 ml of phosphate
buffer. The mixture was stored at 4 C for 20 hours for
the coupling reaction. After coupling, the beads were
washed with dissociating agent (0.13 M NH4OH containing
0.15 M NaCl) and 0.05 M phosphate buffer containing 0.15 M
NaCl (PBS). This antibody gel was stored at 4 C for over
18 months in 0.05 M phosphate buffer, pH 7.5 without

losing antibody activity.

32

Iodination of Enterotoxin

Iodination with 125I was performed by a modifi-

 

cation of the method of Greenwood and Hunter (1963).
Solutions of chloramine-T, NaZSZOS’ and KI were prepared
on the same day of iodination in 0.05 M sodium phosphate
buffer with 0.15 M NaCl (PBS), pH 7.5. Carrier-free
NalZSI, free from preservative or reducing agents, was
obtained in 0.1 N NaOH from ICN Pharmaceutical Inc.,
Irvine, Calif. at a concentration of 40 mCi per ml.

Fifty ug of SEB were dissolved in 50 ml of PBS and trans-

1251. The iodination

ferred into a vial containing 1 mCi
was begun by addition of 50 ul of 0.26 mg/ml chloramine-T.
After 45 sec., the reaction was stOpped by addition of

50 ul of 0.26 mg/ml NaZSZOS' This was followed by
addition of 50 ul of 2% KI, 100 pl of 2% BSA, and 150 pl
of transfer solution (10 g/l KI and 100 mg/l bromphenol
blue in 16% sucrose solution).

The solution was transferred to a Sephadex G-25
column. The vial was washed once with 70 ul of transfer
solution which contained 8% sucrose. This was also
placed on the column.

Separation of labelled SEB from the reaction
mixture was done by gel filtration using a 1.25 x 15 cm
Sephadex G-25 column equilibrated with PBS. Prior to
application of sample to the column, 1.0 ml of 2% BSA

solution was passed through the column to minimize the

33

adsorption of SEB to the glass. Fractions (1.0 ml) were
collected in test tubes containing 1.0 ml of 2% BSA.

The fraction with the highest count, containing approxi-
mately 0.35 ug/lO ul, was saved for further use. The
125

I-SEB was stored at 4 C. It was used for at least

6 weeks after iodination.

Preparation of Antibody Column

The antibody column was made by the Glass Labora-
tory, Chemistry Department, M.S.U. It was made of 1 cm
0.0. diameter glass tube 7.5 cm long with a 4.5 cm stem
(Figure l). The column had a 7/15 ground glass joint
for convenient connection of tubing.

Sepharose-ABE was mixed with well-washed Sephadex
G-25 in proper proportions (1:5 to 1:30) (volume by
volume). Both Sepharose-ABE and Sephadex G-25 were put
into slurry form, 1 part of gel with 1 part of PBS,
before pipetting. Gel of about 2 ml was packed into
the column (about 4 cm in the column). Before the column
was used for SEB assays, the irreversible binding sites
in the column had to be saturated. The column was washed
with 25 ml of dissociating agent, used in the subsequent
assay, 50 ml of PBS, and finally incubated overnight with
20 ug SEB. Then, the column was washed with 15 ml of
dissociating agent and 30 ml of PBS. Using the assay

sequence system (Table 5), the predetermined amount of

34

 

S 7/15
¢ 1.0 cm

 

7.5 cm

 

 

 

 

\ I

.__J_

 

 

Fig. 1. Column used for enterotoxin B determi-
nation.

35

labelled SEB was run through the column at least 4 times

or until a constant count was obtained.

Measurement of Activity

 

Radioactivity was measured with either a Packard
Model 2001 Tricarb Scintillation Spectrometer (gamma ray
counter) or a Packard Model 3310 Tricarb Scintillation
Spectrometer (liquid scintillation counter). A 0.5-ml
sample of NH4OH eluate was added to either 10 m1 of dis—
tilled water or 10 m1 of counting cocktail (Research
Products International Corp., Elk Grove Village, Ill.)
depending upon whether the gamma-ray counter or the
liquid scintillation counter was used. Samples were
neutralized with 0.1 N acetic acid before counting to

reduce chemiluminescence.

Foods Used in Recovery Study

 

Measured amounts of SEB (1.5 to 10 ng/ml of milk
or 1.5 to 12 ng/g of hamburger) were added to reconsti-
tuted nonfat dry milk and hamburger and then stored in a
refrigerator for 24 hours prior to analysis to allow
equilibration of SEB in the food samples.

Inoculation of Reconstituted Nonfat Dry
MiIk with StapEyIBcoccus aureus

 

 

Reconstituted nonfat dry milk was prepared by
dissolving 9.6 g of nonfat dry milk in 100 m1 of PBS,

then divided into two portions, 100 and 500 ml. Milk was

36

pasteurized at 145 F for 30 minutes and, after cooling to

room temperature, inoculated with Staphylococcus aureus

 

Strain 196E (producing enterotoxin A) and Strain 243
(producing SEB) to 100 and 500 ml samples, respectively.
The inoculated milk samples were incubated at 37 C for
20 hours.

Measurement of Enterotoxin B in
Reconstituted Nonfat Dry Milk

 

Reconstituted nonfat dry milk was adjusted to
pH 7.5 with 2 N NaOH. This milk was applied to the column
directly without any further treatment.

Measurement of Enterotoxin B
in Hamburger

 

 

Hamburger containing PBS was homogenized without
further dilution at high speed (setting at speed 7) in a
Sorval homogenizer. After centrifugation and removal of
fat, the supernatant liquid was examined for SEB content
directly. In the experiment, 10 grams of hamburger were
homegenized for 45 seconds in 10 m1 of 0.15 M NaCl after
adjusting the pH to 7.5 with 2 N NaOH. The homogenate
was centrifuged at 32,800 x g for 20 minutes. The super-
natant liquid was decanted into a beaker and stored in a
refrigerator for about 30 minutes or long enough to
solidify the fat. The aqueous portion was removed and

filtered in succession through Nucleopore filters with

37

0.8 and 0.4 pm pore sizes (Nucleopore Corporation,
Pleasanton, CA). The filtrate was used for SEB deter-

mination.

RESULTS

Development of an Affinity Radio-
immunoassay Procedure

 

Radioimmunoassays usually require mixing and
incubation of antigen and antibodies in a liquid phase
followed by some procedure for separation of bound and
free antigen. Some solid-phase assays have been developed
in which the antibody is attached to an insoluble support.
This simplifies separation of bound and free antigen, but
the physical operations are similar.

The objective of this investigation was to develop
a calibrated antibody affinity column which could provide
sensitivity comparable to other RIA procedures, but which
was reusable, allowed rapid assays and which could be
automated if necessary.

The basic starting concept for this assay is
shown in Figure 2. A reusable column is prepared with
antibody covalently attached to the gel (A). Enterotoxin
from the sample is selectively bound by antibody (B),
then labelled enterotoxin is put through the column in
an amount sufficient to saturate the remaining antibody

sites and the excess is washed off the column (C). The

38

39

.hmmmMOGSEEHOHoMH huwcwmmm How umoosoo venom .N .mflm

Amy

in... w .H. UQDOQ

V

as
no

no
no

hm

 

nmmmsm )w
mcfluMHoommflo

AOV

a mmooxm
an

1;]

Elam

i.

an a

k.

sign

 

 

cHxOBAW

Ame lee

mmwuHHSQEH
_I
he , an,
no no
no no
none as
sine he

 

 

 

 

 

wamfimmjv

 

40

bound SEB is removed by elution with dissociating agent
(D). The labelled SEB eluted from the column would be
inversely proportional to the amount of SEB in the original
sample.

In order to achieve this goal, six steps were

required:

(1) Application of sample to the anti-enterotoxin

column;
(2) Removal of impurities from the column;
(3) Addition of labelled SEB;
(4) Removal of unbound labelled SEB;
(5) Elution of bound SEB; and
(6) Re-equilibration of the column for another sample.

Appropriate conditions had to be developed for suppression
of nonspecific binding, specific binding of SEB, removal
of impurities, addition of labelled SEB, elution of bound
SEB, analysis of bound labelled SEB, and column regener-
ation in order to obtain a quantitative assay procedure.
The following steps were investigated in the development

of an assay for SEB.

(1) Selection of a buffer for removal of unbound

SEB and other impurities;

(2) Selection of a dissociating agent for removal

of bound SEB;

41

(3) Development of equilibration conditions for the

antibody column;

(4) Determination of the amount of labelled SEB

required for a specific column;
(5) Suppression of nonspecific binding of SEB;

(6) Evaluation of dose-response curves and the
factors involved in maintenance of consistent

assays.

When the conditions for the first step were
determined, that step was used to study the second con-
dition and so on. Agarose was used for immobilization
of antibody. However, to achieve acceptable flow proper-
ties of the antibody columns, the agarose was diluted
1:10 with Sephadex G-25. Later, a 1:15 dilution was used
to reduce the amount of labelled SEB required.

Binding of Enterotoxin, Removal of Impuri-

ties, and Removal of Unbound
Labelled Enterotoxin

 

 

A single buffer was required to accomplish these
three elements of the assay. It was necessary to mini-
mize tailing, provide conditions of pH and ionic strength
which would allow specific binding of SEB to antibody
while preventing nonspecific binding of SEB. Phosphate
buffer of 0.05 M pH 7.5 was used initially because it
had been used successfully for detection of SEB by

Casman and Bennett (1965) and Johnson et al. (1971).

42

To study the effectiveness of buffers in removing unbound,
labelled SEB 20 ul of labelled SEB was added to a 1:10
dilution antibody-gel column. The column was then washed
with aliquots of 0.05 M phosphate buffer containing 0.15,
0.30, and 0.50 M NaCl. Using a peristalic pump, a flow
rate of 150 ml/hr was used. Fifteen 1.5 fractions were
collected and 0.5 ml was taken for measurement of radio-
activity. The column was then eluted with 25 ml of

2 M NH4OH, pH 11.5 to remove the bound SEB. The column
was re-equilibrated with 50 ml of 0.05 M phosphate buffer.
Figure 3 shows the elution patterns of unbound labelled
SEB. Phosphate buffer containing different concentrations
of NaCl was used to increase the ionic strength of the
buffer to decrease tailing of the elution. Figure 3

shows that tailing was reduced as the concentration of
NaCl increased. However, 0.5 M NaCl appeared to pre-
cipitate labelled SEB because the sum of counts of all
fractions was much lower than that of the control without
NaCl. This did not occur with 0.15 or 0.30 M NaCl in

the phosphate buffer. NaCl of 0.15 M was selected for
use in the assay system because more SEB was expected

to bind to antibody in the column at this concentration.
Figure 3 shows that 30 ml of PBS was sufficient to remove
unbound, labelled SEB. Therefore, this volume was

selected for the assay.

43

 

 

 

 

 

3.0 '
V
I
O
H
X
Z
H
2
m 2.0 ~
m
m
w
E4
E
O
U
C
A O
1.0 '
RKO
k \o\o
K N
0.0 ‘ I l l
3 10 17 24 31

VOLUME OF BUFFER, ML

Fig. 3. Effect of NaCl on elution of unbound
labelled SEB. Symbols: O——O, 0.05 M phosphate buffer;
O——O, 0.05 M phosphate buffer with 0.15 M NaCl; A——A,
0.05 M phosphate buffer with 0.30 M NaCl; A——A, 0.05 M
phosphate buffer with 0.50 M NaCl.

44

Selection of Dissociating Agent
In this experiment, the dissociating agents
investigated were: 8 M urea, pH 3 and 10, 3 M KSCN

pH 8, 6 M guanidine-HCl, pH 3, 1.0 and 2.0 M NH OH

4
pH 11.25 and 11.5, respectively.

The column and method used in this experiment
were the same as those described in the previous section
except 30 ml instead of 50 m1 PBS was used to remove
impurities and unbound labelled SEB from the column.
Bound labelled SEB eluted from the column was collected
in 1.5-m1 fractions when the flow rate was less than
80 ml/hr. For flow rates greater than 80 ml/hr, 3.5 ml
fractions were collected. The radioactivity in each
fraction was determined by counting with a gamma-ray
counter. The column was re-equilibrated for the next
sample by washing the column with 50 ml of 0.05 M PBS.

Figure 4 shows the elution pattern of the three
buffers. The buffer containing KSCN was not effective
for removal of bound SEB. The greatest dissociation of
bound labelled SEB was obtained with 8 M urea, pH 3.

It also showed the least tailing. Urea in pH 10 buffer
was much less effective than in the pH 3 buffer for
removal of labelled SEB. Six molar guanidine-HCl was
only slightly inferior to urea. The greatest degree of
tailing was observed with 2 M NH4OH (Figure 4). However,

when flow rates were considered, 8 M urea and 6 M

45

 

 

 

 

 

 

 

 

 

6.0
T (3
S 4.5-1
X
2
H
2
m I
E I
a) 3.0" I
B I,
S
O
U
1.5
I
0
0 12 24

 

EFFLUENT VOLUME, ML

Fig. 4. Comparison of 8 M urea, pH 3.0; 6 M
guanidine-HCl, pH 3.0; and 2 M NH4OH, pH 11.5 for
removal of bound 12SI-enterotoxin from antibody column.
Symbols: O——O, urea; A——A, guanidine-HCl; A-—A, NH4OH;
O——O, PBS.

46

guanidine-HCl could not be pumped at rates higher than

40 ml/hr because of their high viscosity. Thus, NH OH

4
was selected for further investigation because a flow
rate of up to 75 ml/hr could be used with a 1:5 to 1:30
gel dilution.

Improvement of Dissociating
Ability of NH4OH

 

 

It was necessary to reduce the tailing of the
NH4OH dissociating buffer to take advantage of its
superior flow rate. The addition of NaCl to NH4OH was
investigated as a means for improving the performance of
the buffer. The same assay was followed as described in
the preceding section with the substitution of various

dissociating agents containing NH OH. Figure 5 shows

4
that the addition of 0.3 M NaCl increased the amount of
SEB dissociated and reduced tailing. Figure 6 shows the
effect of different NaCl concentrations on tailing. It
was found that 0.15 M NaCl gave the optimum suppression
of tailing. Figure 6 also shows that 15 m1 of 1.0 M

NH4OH containing 0.15 M NaCl was sufficient to remove the

bound SEB under the conditions used in this study.

Optimum Concentration of NH4OH

 

Using the assay sequence described above, NH4OH
of different concentrations, each containing 0.15 M NaCl,

was passed through the column repeatedly. Fifteen ml of

47

 

 

 

 

 

 

 

 

 

 

5b
T 4'
>4
E
z 3.
m
3*” I
a o
e I
8 2- II
I
I
I
I
1- I
)1
0 1
0 17.0 34.0 51.0

EFFLUENT VOLUME, ML

Fig. 5. Comparison of elution patterns for re-
moval of bound labelled SEB with pH 11.5, 2 M NH4OH with
and without NaCl added. Symbols: O——O, washing with PBS
to remove impurities and unbound labelled SEB; A——A, 2 M
NH4OH without NaCl; O——O, 2 M NH4OH with 0.30 M.NaCl.

48

 

 

3.0
v
I
c>
r—I
>4
E 2.0
2
n:
It]
D-I
U3 .
E
O
U
1.0 -

 

\o\
\ lame“.

A\A
o - M

3 10 17 24 31
VOLUME, ML

 

 

 

Fig. 6. Elution patterns for removal of bound
labelled SEB with different concentrations of NaCl.
Symbols: O——O, 1.0 M NH4OH; O——O, 1.0 M NH4OH with 0.15
M NaCl; A——A, 1.0 M NH4OH with 0.30 M NaCl; A-—A, 1.0 M

NH4OH with 0.50 M NaCl.

49

NH4OH-eluate was collected and 0.5 ml was taken for
radioactivity measurement using a gamma-ray counter.

The NH4OH concentrations were 0.13, 1.0, and 2.0 M.
Figure 7 shows the relationship between the binding
capacity and the number of cycles. It was found that
the binding capacity was not stable with 1.0 and 2.0 M
NH4OH. The loss of binding capacity of the column was
accompanied by formation of gel clumps and a decline of
the flow rate. This instability of the column may be due
to either the denaturation of the antibody in the column
or disintegration of gel. The loss of binding capacity
was not observed with 40 samples for buffer containing

0.13 M NH4OH. Therefore, this concentration was selected

for the assay system.

Equilibration of the Column

 

After elution of bound SEB with NH4OH, the column
must be prepared for the next sample by replacing the
dissociating buffer with pH 7.5, PBS. It was necessary
to determine the volume of phosphate buffer required for
equilibration. The procedure used in this experiment
was to equilibrate 1:10 dilution antibody-gel with 0.05 M
phosphate buffer pH 7.5 containing 0.15 M NaCl. Twenty
ul of labelled SEB were run through the column, followed
by washing unbound SEB with 30 ml PBS. Bound enterotoxin
was eluted with 15 ml of 0.13 M NH4OH containing 0.15 M

NaCl. The column was then washed with PBS and 3.5-ml

50

6.: mm $8312 2 o.~ .1 3a.: mm $5312 2 o; .oIo 2....3 mm 503%
z mH.o .<Ill< “maonahm .unomo mowHMHoommwo mo owns mums mOvmz mo msowuouu
Icmocoo aconommwo sons waoouoomou can maouucoo mo mucsou .n .mHm

mmmZDz MAUMU

 

 

 

 

om
. ov
d
3
m
cm m
L
V
9
H
O
3
cm M
I
I
m
m
OOH m
Tu
S

 

 

 

51

fractions were collected. The pH and the presence of
NH: were determined for each fraction with Nessler's
reagent. The results are shown in Table l. The pH
reached 7.5 by fraction 3, but ammonium ion was detected
until fraction 8. This result indicated that at least

30 ml of PBS were required to wash off NH OH.

4

Table 1

Equilibrating the Antibody Column with PBS

 

 

Volume of Buffer, ml pH NH:
3.5 10.25 ++++

7.0 7.8 ++++

10.5 7.5 +++

14.0 7.5 +++

17.5 7.5 ++

21.0 7.5 ++

24.5 7.5 +

28.0 7.5 -

 

Time Required for Contact Between
Enterotoxin and AntiBOdy

 

 

It was necessary to have a reasonable time of
contact between antigen and antibody in order to obtain
sufficient antigen binding. For the assay, it was
important to determine the contact time to have a suf-
ficient count in order to do the assay. In experiments
with 1:10 dilution-gel column, the following sequence
was used: 40 ul labelled SEB, 30 m1 of 0.05 M PBS,

15 m1 of 0.13 M NH OH containing 0.15 M NaCl and 30 m1

4
of 0.05 M PBS. Fifteen ml of NH4OH-e1uate was collected

52

and 0.5 ml was taken for a radioactivity measurement
using a liquid scintillation counter. Table 2 shows

that increasing the contact time from less than one
minute (labelled SEB was passed through the column and
washed off immediately) to 5 minutes increased the count
rate by about 39%. A further 20% increase in bound anti-
gen was observed if the contact time was increased to

10 minutes. In order to maximize the antigen binding,

10 minutes was selected as a contact time for labelled
SEB. Twenty minutes was used for samples because it was
thought that high levels of impurities in food samples
might interfere to some extent with SEB binding. Consis-
tent assays were obtained using this time for binding.
However, further investigation is required to Optimize
this factor in the assay.

Amount of Labelled Enterotoxin Required
for the Assay Procedure

 

 

The amount of labelled SEB used in the assay sys-
tem must be high enough to saturate antibody binding
sites. Using the assay sequence mentioned in the pre-
ceding section, 10, 20, 30, and 40 ul of labelled SEB
were applied to the column with a contact time of 10
minutes. Fifteen ml of NH4OH-eluate were collected and
counted. Figure 8 shows the relationship between the
amount of labelled SEB applied to the column and the

radioactivity that was subsequently eluted. There is a

Table 2

53

Effect of Contact Time on Binding of Labelled SEB

 

 

Contact Time . a Mean, Standard
Min. Count/min. Y Deviation, s
b
< 1 2,215 2,492 -
2,769
5 3,900 3,452 237
3,452
3,543
10 4,348 4,150 232
4,355
3,956
3,944

 

aThe count was done by liquid scintillation

spectrometer, Packard Model 3310, using 0.5 ml of 0.13 M

NH4OH eluate in 10 ml of counting cocktail 3a70.

b

Labelled enterotoxin was passed through the
column and washed off immediately with PBS.

54

Table 3

Effect of Contact Time on Binding of SEB to Its Antibodies
in the 1:15 Dilution Antibody-Gel Column
(The result obtained from incubation of 0.68 ml of 2 ng
SEB [in PBS] per ml with different contact time, using
40 ul of labelled SEB, 30 ml of PBS for washing impurities
and unbound labelled SEB and 15 ml of 0.13 M NH4OH [with
0.15 M NaCl] for elution)

 

Contact Time, Counts Per Min. Mean, Standard
Min. of NH4OH Eluatea X Deviation, s

 

Controlb 3,696 3,674 43.4
3,702
3,624

15 3,629 3,602 53.8
3,540
3,637

20 3,623 3,603 -
3,583

30 3,613 3,591 78.3
3,682
3,576
3,494

 

aAverage count from S-minute counting of 0.5 m1
NH OH eluate in 10 ml counting cocktail 3a70, using liquid
sc1ntillation spectrometer, Packard Model 3310.

bWithout SEB in the sample.

55

 

 

 

 

 

4 I-
v
I
o
H 3 ..
ix:
2
H
2
m
£11
0.4
U)
E4
E".
I.
8 2
1 - .
0 l J I J, 1 r
0 10 20 30 40 50 60

LABELLED ENTEROTOXIN B, pl

Fig. 8. Relationship between the amount of bound
and added labelled SEB in the column 1:15 dilution anti-
body-gel.

56

linear increase in the radioactivity bound with up to
30 ul. In the measurement of SEB in samples applied to
the column it was found that 40 ul of labelled SEB
solution was required to observe a response for samples

with low enterotoxin concentrations.

Suppression of Nongpecific Binding

 

If nonspecific binding of SEB to the gel or glass
column occurred to a large extent, it could lead to false-
positive errors with low levels of SEB. The addition of
an inert protein has been used to suppress nonspecific
binding to glass or gel (Greenwood & Hunter, 1963).

An experiment was designed to suppress nonspecific bind-
ing of SEB to the column. Two ml of Sephadex G-25 were

packed into a column and eluted with dissociating agent

followed by PBS. One-half ml of BSA (5 to 20 mg/ml)

was passed through the column. This was followed by 10

ml of PBS, 40 ul of labelled SEB, 30 ml of PBS, and

finally eluted with 15 ml of 0.13 M NH OH containing

4
0.15 M NaCl. The activity was counted using 0.5 ml of
NH4OH-eluate.

The above experiment was repeated using 0.5 ml
of 20 mg/ml BSA up to the step before adding the labelled
SEB. Instead of adding only labelled SEB, a mixture of
labelled SEB and BSA (40 ul of labelled SEB:0.65 m1 of

BSA) was added to the column. The BSA concentrations

57

used were 5, 10, and 20 mg/ml. Then the column was
washed with PBS and eluted with dissociating agent.
The NH4OH was counted as before.

Table 4 shows that nonspecific binding of labelled
SEB is 2.28% of that of the control (column 1:15 dilution
antibody gel) when no protein was added.

It decreased to 1.4% when BSA was increased to
20 mg/ml. The suppression of nonspecific binding was
more effective when protein is also present in a sample.
It was reduced to about 1% in the sample containing

20 mg/ml BSA.

Table 4

Suppression of Nonspecific Binding

 

 

Treatment of Column Add Labelled Enterotoxin
mg/ml BSA Enterotoxin Bound
%

0 in buffer 2.28

5 " 1.97

10 " 1.76
20 " 1.44
20 in BSA, 5 mg/ml 1.23
20 in BSA, 10 mg/ml 1.12
20 in BSA, 20 mg/ml 1.01

 

Assay Sequence

 

The investigations of factors involved in this
affinity radioimmunoassay led to development of the assay
procedure shown in Table 5. A series of 7 steps were

required which took a total of 73 minutes. The column

58

Table 5

Assay Sequence for Enterotoxin B

 

 

Time
Step Required
Min.
1. Addition 0.5 ml of 2% BSA and 5
followed by washing with 10 ml
0.05 M phosphate buffer saline
2. Addition of sample and incubation 20
3. Wash off the impurities and adsorbed 2
substances with 5 ml 0.05 M phosphate
buffer saline, pH 7.5
4. Addition of 40 ul of labelled 10
enterotoxin
5. Wash off unbound labelled enterotoxin 12
with 30 ml, 0.05 M phosphate buffer
saline, pH 7.5
6. Elution of bound enterotoxin with 12
15 ml 0.1 M NH4OH, pH 10.5
7. Wash off NH4OH with 30 ml 0.05 M 12

phosphate buffer saline, pH 7.5

Total 73

 

59

was treated with 0.5 m1 of 2% BSA prior to sample appli-
cation. After a sample was applied to the column and
incubated for 20 minutes, the column was washed with

pH 7.5 PBS to remove impurities from the column without
removal of bound SEB. Labelled SEB was added so it would
bind to those antibody molecules that did not react with
SEB in the sample. The labelled SEB which did not bind
was removed by another PBS wash. The bound SEB, both
labelled and unlabelled, was then eluted with dissociating
NH4OH. The radioactivity of the eluted SEB was inversely
proportional to the amount of SEB in the sample. The
dissociating NH4OH was removed and the column prepared
for the next sample by equilibration with PBS.

In the manual procedure used during development
of the assay, steps 1 (BSA), 2 and 4 were done with
gravity flow. A flow rate of 150 ml/hr was used for
steps 1 (PBS), 3, 5 and 7, and a 75 ml/hr flow rate was

used for step 6.

Dose-Response Curve
Dose-response study was done using the procedure
outlined in Table 4. The concentrations of SEB used
were 1.5, 3.0, 4.5, 6.0, 7.5, 9.0, and 10.5 ug/ml.
These solutions were run through a column in which the
antibody-gel was diluted 1:15 with Sephadex G-25. A
sample of 0.68 ml and contact time of 20 minutes were

used for all concentrations. Fifteen ml of NH4OH-eluate

60

were collected and a 0.5 ml sample was taken for activity
measurements. The count with no SEB in the sample control
was the average of four determinations. Duplicate samples
were run for each of the other SEB samples. Figure 9 shows
the relationship between the SEB level and the counts per
minute of bound SEB. The same data are shown in Figure 10,
but the radioactivity is expressed in terms of the per-
centage of the counts bound in the control assays.

The slope, intercept, and correlation coefficient
for the least square regression of the curve plotted in
Figure 10 were calculated and found to be -0.821%/ng
SEB, 100.1%, and 0.986 respectively.

Factors Affecting the Dose-
Response Curve

Dilution of antibody gel with Sephadex G-25
reduced the antibody concentration in the column. This,
in turn, affected the dose-response curve. Figure 11
shows the effect of variation of labelled SEB concen-
tration from 20 to 40 ul when a column of 1:15 dilution
was used. At a low SEB level, no apparent response is
observed with 20 ul labelled SEB. A response is observed
when the labelled SEB is increased to 40 ul. An improved
response was obtained when the dilution of the antibody
gel with Sephadex G-25 was increased from 1:5 to 1:15.

For the effect of pH, there is little evidence

that antigen-antibody complex exhibits a significant pH

-4

COUNTS PER MIN X 10

4.2

4.1

3.9

3.7

3.6

3.5

61

 

 

01)

 

l l

3 6 9 fT-J ’

ENTEROTOXIN B CONCENTRATION, ng

b

Fig. 9. Standard curve of SEB in PBS, showing

the relationship between the enterotoxin level and counts
per minute.

PERCENTAGE OF INITIAL COUNTS

62

 

 

 

 

 

1009
‘0
‘0
0*
0'
\o
95-
C)
0'
C)
0’
0'
90b
85 l l 1 l
0 3 6 9

ENTEROTOXIN B CONCENTRATION, ng

Fig. 10. Standard curve of SEB in PBS, showing
the relationship between the enterotoxin level and per-
centage of initial counts.

63

 

 

 

 

 

\.
C)

m 99 IO
B
Z
{3
O
L)
g 98
H C)
E.‘
E o
H 97
[LI
0
[1.]
L9
¢
5
m 96 _
U
a:
[1.]
D4

95 L

94 ‘

93 A L A l 1 L

0 3 6 9 12 15 18

ENTEROTOXIN B CONCENTRATION, ng

Fig. 11. Effect of gel dilution and concentration
of labelled SEB on the dose-response curve. Symbols:
O—-O, 1:5 dilution antibody-gel column with 20 ul labelled
SEB; O——O, 1:5 dilution antibody-gel column with 40 ul
labelled SEB; A——A, 1:15 dilution antibody-gel column with
40 ul labelled SEB.

64

dependency in the range of 7 to 8 (Berson & Yalow, 1973).
But it was necessary to maintain a constant pH for both
standard and unknown samples in order to maintain equil-
ibrium of the column.

The ionic strength of the buffer system is very
important for antigen-antibody complexes. High concen-
trations of NaCl may inhibit reactions (Berson & Yalow,
1973). If the salt concentration of the unknown sample
differs from that of standards, differential inhibition
of the immunochemical reaction will give erroneous
results. In this experiment, when 0.68 ml of reconsti-
tute nonfat dry milk prepared in 0.05 M phosPhate buffer
containing 0.30 M was used as a sample, the count of the
eluate increased 5.45%.

Validation of the Assay
Procedure

 

 

To validate an assay, one must demonstrate SEB
in the sample and standard (purified enterotoxin) react
with antibody identically and that under the conditions
of assay, other substances in the sample do not affect.
the determination. Enterotoxins A and B were produced

in milk samples inoculated with Staphylococcus aureus

 

strains 196E (ATCC 13565) and 243 (ATCC 14458), respec-
tively. Authentic SEB samples were also obtained from
the Food and Drug Administration. The ability to detect

SEB from these samples was checked.

65

From the experiment, it was found that the anti-
body column can be used to detect SEB prepared by FDA
and produced in the inoculated milk. No cross-reaction
was observed with enterotoxin A from FDA. The experiment
also showed no cross-reaction occurred between substances
in the inoculated medium (milk inoculated with Strain 196E)

and antibody in the column (Table 6).

Detection of Enterotoxin B in Foods

 

Detection of Enterotoxin B
in Nonfat Dry Milk

 

 

Measured amounts of SEB were added to reconsti-
tuted nonfat dry milk. The concentrations used were 1.5,
3.0, 4.5, 6.0, 7.5, 9.0, and 10.5 ng/ml. Then, the pH
of the milk was adjusted to 7.5 with 2 N NaOH. The
samples were run using the procedure described in
Table 5 .

Figure 12 shows the relationship between the
counts per minute and the SEB concentration. In Figure 13
the data are expressed in terms of the percentage of the
control counts. The second standard curve has a slope,
intercept, and correlation coefficient of -0.793%/ng SEB,
100.2%, and 0.972, respectively.

The amount of SEB in an unknown sample can be
determined directly from this standard curve if a sample
is prepared in the same way as samples used in preparing

a standard curve.

66

.conumcfleumume as» muomon mosh» ooo.a omusflnu was gag:

.manmuoouoc uozo

Q

.ooooo mos saxouououco ozo

 

.Ibmwmma_woa«e mead chmuum
mnmuom .m nufi3 oouoasoocfl

 

ooz masocxss mo.ooa Made Baum « saxououousm
ooz coca mm.mm «am scum « saxououousm
Ilkbhhvfiluoedv new swouum
msousm .m nuw3 omumasoocw
booms mc3ocxss hm.Hm MARE Eoum m saxouououcm
H.m o.m v¢.mm éam Scum m saxouonouam
pooch ooood

 

HE\os .mmm mo cowuonusoocou

unsoo HowuwcH
mo ommusmoumm

CflXOflOHmudm MO MOUHSOW

 

ounooooum mommfi may no coauoowao>

0 mafia.

-4

X 10

COUNTS PER MIN

67

 

4.2 -

 

3.8 L

307 P

 

3.6’

3.5-

 

3.4 ‘ ‘— L J

 

 

O 3 6 9 12

ENTEROTOXIN B CONCENTRATION, ng

Fig. 12. Standard curve of SEB in reconstituted
nonfat dry milk, showing the relationship between the
enterotoxin level and counts per minute.

68

 

 

IOOOI

COUNTS

90”

PERCENTAGE OF INITIAL

 

 

 

85 I L 1 1
6

ENTEROTOXIN B CONCENTRATION, ng

Fig. 13. Standard curve of SEB in reconstituted
nonfat dry milk, showing the relationship between the
enterotoxin level and percentage of initial counts.

69

Detection of Enterotoxin B
in Hamburger

 

 

Ten grams of hamburger were mixed with 10 m1
SEB solution (SEB was dissolved in 0.05 M PBS). The
concentration of SEB was so prepared that l g of ham-
burger contained 2, 4, 6, 8, 10, and 12 ng. The pH of
the hamburger was adjusted to 7.5 with 2 M NaOH. Several
re-adjustments were required to be sure that the pH of
the mixture was 7.5. After 24 hours, the SEB in the
samples was extracted (see MATERIALS AND METHODS) and
0.68 ml was run through the column for SEB determination.
Two curves were constructed in the same way as for recon-
stituted nonfat dry milk (Figures 14 and 15). The slope,
intercept, and correlation coefficient for the linear
portion of the curve were found to be -1.568%/ng SEB,
106.5%, and 0.969, respectively. The amount of SEB in
the sample can be calculated directly by use of this
standard curve.

Determination of D-Value for Enterotoxin

Denaturation ipreconstituted
Nonfat Dry Milk

 

Reconstituted nonfat dry milk was inoculated
with Staphylococcus aureus Strain 243. After 20 hours
of growth, the whey portion (pH 5.2) was separated from
the curd (formed during incubation) and centrifuged at
32,800 x g for 20 minutes. Ten ml of this aqueous

solution were pipetted into Thermal Death Time (TDT)

70

 

4.1 _

 

-4

X 10
Us)
In
B”-r-
D’
D’
D’D'

3.7 b

COUNTS PER MIN
DD.

3.5-

 

H>/’

 

 

3 4 ‘ ‘ ‘ . -
0 2 4 6 e 10 12

ENTEROTOXIN B CONCENTRATION, ng

Fig. 14. Standard curve of SEB in hamburger,
showing the relationship between the enterotoxin level
and counts per minute.

71

 

PERCENTAGE OF INITIAL COUNTS

 

 

 

AI
1004A
A.
A| IA
95 P
90 5 4A
A.
85 I A L l 1 L
0 2 4 6 8 10 12

ENTEROTOXIN B CONCENTRATION, ng

Fig. 15. Standard curve of SEB in hamburger,
showing the relationship between the enterotoxin level
and percentage of initial counts.

72

cans. The cans were heated at 121 C (250 F) in a
miniature retort for 2, 5, 10, and 15 minutes (uncor-
rected for heating lag) and cooled immediately with
water to room temperature. The pH of heated milk was
adjusted to 7.5 with 2 M NaOH. The samples were centri-
fuged at 32,800 x g for 20 minutes and analyzed for SEB
remaining using the standard curve for milk (Figure 13).
Figure 16 shows the relationship of the amount of SEB
found and time for each treatment. The correlation
coefficient was 0.9969. The D-value, calculated as the

reciprocal of the slope, was 10.4 minutes.

Sensitivity of the Assay

 

The sensitivity of the assay was determined by
"the least-detectable dose" method (Hunter & Greenwood,
1964; Feldman & Robard, 1971). The least detectable dose
was defined as the lowest concentration of unlabelled
SEB which results in a significant change in the response.
Significant change was defined as the mean of the standard
deviation of the response at the point where the unlabelled
SEB is zero, multiplied by a "significant value" of
Student's 2,

Table 7 shows the calculated values for the
least-detectable dose of SEB in PBS, reconstituted nonfat
dry milk and hamburger. For a standard curve of SEB in
PBS using four samples, the mean value of the counts for

control (no enterotoxin was added) was 42,568, and the

73

 

10.0 .

 

VIII

6

ENTEROTOXIN CONCENTRATION, ug/ml

 

 

 

 

 

0.1 , L 1

 

 

O 5 10 15

HEATING TIME, MIN.

Fig. 16. Heat inactivation of SEB in reconsti-
tuted nonfat dry milk at 121 C, pH 5.2.

.ham>fluoommou .v u a
can m u a mom mmm.m pom omm.~ who Hw>oa mo.o um mocmowmacmwm How u moan

 

74

 

 

m.w 6m.a «as mum msm.mm 6 ummusnsmm
m.~ mh.a Hon owe NNH.ov m Raw:
~.H 6m.o mam mom www.me v nommsm

m Emu
HE\m: .umoo m
manmuomumo oOOcOOHmcoo .oofiumfl>oa x .cooz co.wwmfiom MMHHW
IUWMTQ OER. wma UM UHMUgfim I. H 33.52 U U Hm

voodoo DQMOAMHGmHm

 

mound mum mo mufl>fluwmswm

h OHQMB

75

standard deviation, 3, was 509. The significant change
at 95% confidence was 399 counts per minute (significant
change = ts//7;:1), or 0.94% reduction in count. This
corresponded to a minimum SEB concentration of 1.2 ng/ml
for detection. The least-detectable dose response for
milk and hamburger were found to be 2.2 and 6.3 ng/ml,

respectively.

Reproducibility

 

For the evaluation of reproducibility of the
assay, a toxin concentration near the midpoint of the
assay range (0 to 12 ng/ml) was selected. Enterotoxin
at a concentration of 6 ng/ml in PBS was run four times
during a two-week period when 64 samples, including meat
and milk, were measured. The average level detected was
5.9 i 0.4 ng/ml. Six consecutive runs were done with a

different batch of 125

I—SEB using the standard curve in
Figure 10. A concentration of 6.5 ng/ml of SEB was used.
The average level detected in this experiment was found
to be 6.6 i 0.4 ng/ml. When the data from these 10
determinations were combined and expressed as the per-

centage SEB recovered, the average was 100.4 i 5.7%

(Table 8).

76

 

H.5m

 

m o
«.mm m.m
m.aoa m.m
H.ma N.m
m.mcH h.m
N.HHH v.o m.m «.5 m.m
m.Hoa H.m
m.mm m.m
«.mm m.m
h.m ¢.ooa h.moH ¢.o m.m v.m o.m
woes m \m
a mod m consume: HE\ G HE :
cosumfl>mo ca soums mm @0664 cmwwmmwwm Hmmwm venom mum emcee mum
oumosoum muo>ooom mo omousmonmm mo dowuonusoocoo mo cowuouucmocoo
com:

 

hound may no wuflawnaonooumom

m manna

DISCUSSION

An affinity radioimmunoassay has been developed
for the measurement of staphylococcal enterotoxin B in
buffer solution and in food products. The principle of
the assay is similar to the solid-phase radioimmunoassay
reported by Catt et al. (1967). However, the ability to
reuse the immobilized antibody and the potential for
adapting the column technique for automated analysis
systems are significant advantages of this procedure.

SEB was used in the development of the affinity
radioimmunoassay procedure because the SEB and its anti-
toxin are commercially available. In addition, SEB is
stable under the experimental conditions which were
employed.

Agarose 4B was used to bind antibody since it
was more porous than the GB derivative and has greater
capacity than the ZB gel. Antibody was readily attached
to agarose 43 with a general cyanogen bromide coupling
procedure (March et al., 1974). The immobilized antibody
was stable to extended refrigerated storage and to a

variety of experimental conditions. The functional

77

78

stability of agarose-ABE complex allows repeated assays
with a single column with the procedure given in Table 5.
One column has been used for assay of over 100 samples
without measurable loss of its antibody activity.

Most antibody preparations which are used for RIA
show marked heterogeneity of binding affinity for the
antigen (Parker, 1976). During the initial experiments
to bind and dissociate SEB from antibody columns, it was
not possible to dissociate low levels of SEB from the
column. It was found that before a new batch of antibody
gel could be used for the column it had to be saturated
by overnight treatment with an excess of SEB. Twenty
micrograms of SEB were sufficient for saturation. Either
purified SEB or a crude SEB preparation concentrated from
a culture filtrate was suitable for this purpose. The
column was then eluted with dissociating agent to free
those antibody binding sites with sufficiently low
affinity that antigen could be dissociated in the assay.
After this procedure, which had the effect of permanently
saturating the high affinity binding sites, low levels
of SEB could be attached and removed from the column
repeatedly.

The dissociation of antigen was improved with
higher NH4OH concentrations. This had the effect of
increasing the capacity of the antibody column. However,

1 M and 2 M NH4OH resulted in significant loss of binding

79

capacity upon repeated use (Figure 7). Clumping of the
gel was observed with repeated use of these solutions.
Microscopic examination showed fragmentation of the gel
beads. The column was stable for over 40 cycles with
respect to SEB binding capacity and physical structure
when the NH4OH concentration was reduced to 0.13 M. The
fraction of the antibody binding sites which dissociated
with this lower level of NH4OH was reduced, but it was
still sufficient for the range of SEB concentrations for

which the assay was intended. Therefore, 0.13 M NH OH,

4
pH 10.5, containing 0.15 M NaCl was chosen as the dis-
sociating agent.

The buffer for equilibration and removal of
unbound impurities from the antibody column (steps 2, 4,
and 6 in Table 5) had to be chosen to provide conditions
which allow specific binding of SEB, prevents nonspecific
binding and minimizes tailing during elution of unbound
labelled SEB from the column (step 4). A pH 7.5 phosphate
buffer was chosen since this was known to provide suitable
conditions for antigen-antibody reactions in the slide
diffusion technique (Casman & Bennett, 1965) and in
solid phase radioimmunoassays for SEB (Johnson et al.,
1971). Sodium chloride was added to the phosphate buffer
to reduce tailing during elution of unbound SEB. The

results in Figure 3 show a reduction in tailing as the

NaCl concentration was increased. However, previous work

80

with affinity chromatography (Robinson et al., 1972)
indicated that improved binding of antigen should occur
at lower NaCl concentrations. Therefore, 0.15 M NaCl
was selected for subsequent experiments because it
resulted in a significant reduction in tailing. In
addition, the nonspecific binding of labelled SEB at
this salt concentration was only 2.3%.

A salt concentration of 0.15-0.20 M was found by
Casman and Bennett (1965) to be effective for extraction
of staphylococcal enterotoxins from food samples. For
radioimmunoassay, samples for analysis should have the
same pH and ionic strength as that of the equilibrated
column for maximum sensitivity and accuracy. Therefore,
an assay system with 0.15 M NaCl should minimize the
work required to prepare a sample for analysis.

The nonspecific binding of labelled SEB to the
gel and column was reduced by treatment of the column
with 0.5 ml of a 2% BSA solution prior to application
of a sample.

The flow rate for steps 1, 3, 5, and 7 (Table 5)
when PBS was pumped through the column was 150 ml/hr.

It was found with a 1:5 to 1:30 antibody gel that 150
ml/hr was the maximum flow rate which could be maintained.
Higher rates resulted in compaction of the gel and a
gradual decline in flow rate. A 75 ml/hr flow rate for

the dissociation agent was chosen to allow a longer

81

contact time between dissociating agent and the SEB-
antibody complexes. Further experiments are required
to determine whether this flow rate might be increased
without changing the elution characteristics.

Once the assay procedure described in Table 5
had been developed, it was necessary to evaluate the
specificity and accuracy of the assay. Dose-response
curves were obtained for the commercial SEB added to
PBS, reconstituted nonfat dry milk and hamburger. The
data obtained from the dose-response study were plotted
in two ways. Figures 9, 12, and 14 show the counts per
minute measured in the NH4OH-e1uate as a function of
enterotoxin concentration in the samples. The milk and
buffer curves (Figures 9 and 12) were straight lines over
the range of concentration measured. The hamburger
samples (Figure 14) did not show a significant decline
until the SEB concentration was 6 ng/g of hamburger.
However, above 6 ng/g, a linear relationship was observed
with a slope greater than that found for PBS or milk
samples. Figure 17 shows that standard curves prepared
in reconstituted nonfat dry milk and hamburger are always
lower than that prepared in PBS. The lower count of
labelled SEB bound with milk indicates some interference
by impurities in the sample. In hamburger samples, the
lack of response at low SEB levels suggests the possi-

bility that a factor(s) may be present in the sample

-4

X 10

COUNTS PER MIN

82

 

 

 

 

 

 

3.6 ’ IA
3.5 P
i I
3.4 l J l 4
0 3 6 9 12

ENTEROTOXIN B CONCENTRATION, ng

Fig. 17. Comparison of standard curves of SEB in
PBS, reconstituted nonfat dry milk and hamburger, showing
the difference when the plots are based on counts per
minute. Symbols: O——O, PBS; O——O, reconstituted nonfat
dry milk; A——A, hamburger.

83

which binds SEB so it cannot be extracted. One of the
possible factors that might bind SEB at low levels would
be antibody since Bergdoll (1972) found that enterotoxin
antibodies were usually present in animals in a minute
amount. After all these substances were neutralized with
the added SEB, the excess SEB began to appear in the
extract.

The data were also calculated in terms of the
percentage of the average counts obtained in the specific
products with no enterotoxin. These results are shown in
Figures 10, 13, and 15. This method for plotting the
results was found to be preferable when samples are done
over a period of several days or weeks because at a given
SEB concentration the counts per minute measured declined

as the 125

I decayed. However, the activity calculated

as percentage of a control sample run on the same day did
not change. Therefore, a single standard curve could be
utilized for a period of several weeks if necessary. In
addition, it can be seen in Figure 18 that the dose
response curve for the SEB in PBS and milk are nearly
coincident when expressed on a percentage basis. There-
fore, it may be possible to use a single standard dose-
response curve for several food products provided that

it is first established that the curves are linear over

the range of concentrations which are measured.

84

 

 

 

 

100
U)
H
z
D
8

95
a
c
H
e
H
2
H
[:4
o
[r]
o
4
E
z
8

m 90
[L]
m

85 ‘ PL 1 .

 

ENTEROTOXIN B CONCENTRATION, ng

Fig. 18. Comparison of standard curves of SEB in
PBS, reconstituted nonfat dry milk and hamburger, showing
the difference when the plots are based on percentage of
initial counts. Symbols: O—-0, PBS; O——O, reconstituted
nonfat dry milk; A——A, hamburger.

85

The minimum detectable concentration for SEB was
calculated according to the method of Feldman and Rodbard
(1971). The calculated levels were 1.2 ng/ml for buffer,
2.2 ng/ml for reconstituted nonfat dry milk, and 6.3 ng/g
for hamburger. These levels are within the range for
practical SEB detection in food suggested by Bergdoll
et al. (1976).

Table 9 shows data on several assay techniques
which have been used for detection of staphylococcal
enterotoxins. The minimum detection levels of the assay
system developed are lower than those found by Collins
et a1. (1973) for RIA of enterotoxin A in ham, milk
products, crab meat, and custard. Thelevels were com-
parable to the minimum detectable levels of both entero—
toxin A and B reported by Johnson et al. (1973) in a
variety of foods. When the time required for preparation
of a sample is taken into consideration, affinity and
solid-phase radioimmunoassays require the least prepar-
ation time. They did not require complicated and lengthy
purification steps like other methods. Food was extracted
with PBS at pH 7.5 and centrifuged to remove food par-
ticles. The supernatant obtained was used for SEB
determination directly without any further treatment.
Table 9 shows that both affinity and solid-phase radio-
immunoassays require approximately one hour for prepar-

ation of a sample while other methods require 2 to 48

86

.GOHummHumm>cH umocon

cam flflXOnwOHGHGO .HOhM

 

Amme ..Hm um HmnumHHmzv ummu

 

emuom m oooa was» aonsmMHuuHmm mHmcwm afloao
AmhmH ..Hm um Hmmmmuv

mum gm omH sOHmsMMHoosdEEHouuumHm HHmnamH
mH we OOH AvhmHv .Ho um HmmHmm
mauwe Nb OOH .xmmmHv pumccmm can cosmmo

umszanoou moHHmouowz
AmmmH ..Hm um Haney

and we om umm» onsu cofimammflonamm mansoo
AmmmH
..Hm um GMEH0>HHmV Ammmmv woman

vamH m m.H GOHumcHuaHmmmamn m>Hmmmm ommuo>mm
mm H H AmhmHv .Hm um com::0b
mm H on .AmhmHv .Hm um msHHHoo

”homomocsafiHOHomu mmmnmuoHHom

«\H H H N.H hummMOGSEEHOHomH huHchmé

mommfl coHuomuuxm

 

anon tomHHoUmm mEHB

HE\ms .mmm mo Hm>mq
anmuomwma HMEHGHE

wonumz

 

oosumz m>HumuHusmsa m50HHm>
an mmm mo mommd can coHuomuuxm How mEHB mmeonummd can Ho>mq mHnwuomumn HMEHst

m mHQMB

87

hours. Although RPHA method does not require concentra-
tion step, it does require a step for precipitation of
protein in a sample. A great number of false positive
assays may be obtained due to nonspecific agglutination
of the sensitized cells if protein precipitation is not
done.

Affinity and solid-phase radioimmunoassays appear
to offer higher sensitivity and shorter sample preparation
times than other techniques. When these two methods are
compared, affinity radioimmunoassay requires a shorter
time for assay of a sample. This is because the affinity
radioimmunoassay does not require the reaction to go to
equilibrium. The time required for the reaction of
labelled SEB with its antibodies is just long enough to
get a constant count under the assay conditions.

Good reproducibility of the SEB affinity radio-
immunoassay was found both with samples run consecutively,
and with samples measured over a two-week period when
other samples were run in between the samples used to
check the reproducibility. For the 10 determinations
shown in Table 8, a mean of 100.4% of the added SEB
concentration was measured. The standard deviation
was 5.7%. Once a standard curve was obtained, it could
be used for at least several weeks even when the batch
of 1251-533 was changed. Johnson et al. (1973) checked

the recovery of both enterotoxin A and B from condensed

88

milk. They used six samples for each enterotoxin, in
which 1 to 75 ng of enterotoxins were added to 1 m1 of
milk. The average percentage recovery was 87.2 with SEB
and 90.7 with enterotoxin A. Standard deviations of the
recovery were 29.8 and 20.5% for enterotoxin B and A,
respectively.

The specificity of the column was evaluated by
using purified enterotoxin A and B obtained from the
Food and Drug Administration and milk samples in which

the A and B enterotoxins were produced by Staphylococcus

 

aureus strains 196E and 243, respectively. No SEB was
detectable with either of enterotoxin A samples even
though the purified enterotoxin was used at nearly 500
times the minimum detectable level of SEB. According to
data provided with the FDA purified SEB, a concentration
of 8 ng/ml was used as a sample. Based upon the standard
curve developed with commercial SEB, a concentration of
8.1 ng/ml was measured. SEB was also detected at a level
of 7,300 ng/ml in the milk sample. These data indicated
no cross reactivity of the anti-SEB column with entero-
toxin A and ability to accurately measure SEB from dif-
ferent sources. Similar result concerning about cross
reactivity for enterotoxin A and B has been reported by
Bergdoll and Robbins (1973).

A D-value of 10.4 minutes was measured for

thermal denaturation of SEB in pH 5.2 milk at 121 C.

89

The fact that this is similar to the D-value of 9.4
minutes in pH 6.4 to 6.6 milk using slide immunodiffusion
(Reed & Bradshaw, 1966) is a further indication that the
affinity radioimmunoassay technique gives results which
are comparable to those obtained using other assay methods.

Further improvement of the affinity radioimmuno—
assay procedure should be possible. One of the important
areas of improvement would be to use only those antibodies
with an affinity for SEB in the Optimum range for this
method. If the highest affinity antibodies were removed,
it would not be necessary to presaturate new columns
with SEB. Removal of the lowest affinity antibodies
might make the point of column saturation more distinct.
In addition, it might reduce the time required for sample
incubation so that consistent binding of SEB to the optimum
affinity sites can be obtained.

To obtain only antibodies with SEB affinities in
the working range for the assay, it may be possible to
reverse the procedure used in the assay. A preparative
column with immobilized SEB could be prepared. Antibody
to SEB would be put through the column at the same pH
and ionic strength that is used for SEB binding in the
assay. After washing to remove low affinity antibodies
and impurities, the antibody could be eluted with 0.13 M

NH40H containing 0.15 M NaCl. This would remove the

90

proper antibody fraction, but leave the very high

affinity antibodies attached to the column.

CONCLUSIONS

An affinity radioimmunoassay has been developed
for staphylococcal enterotoxin B. The assay is sensitive,
reproducible, and accurate for SEB measurements in phos-
phate buffer, reconstituted nonfat dry milk, and hamburger.
A column of antibody gel has been used for over 100 assays
without detectable loss of antibody activity. The assay
is relatively simple to run and it is rapid, requiring
about 1 1/4 hr per sample, compared to the immunodiffusion
method (Casman & Bennett, 1965), or a solid-phase RIA pro-
cedure with comparable sensitivity which required an
18-hr incubation period (Johnson et al., 1973).

The procedure developed is not well adapted to
large numbers of manual assays because each column would
require at least 2 channels of an accurate peristalic
pump. However, it may be useful for small numbers of
manual assays, since it appears that a calibrated column
could be prepared and used as required. The major
potential of the procedure would appear to be the devel-
opment of an integrated autoanalyzer system in which

several columns could be processed simultaneously.

91

92

The assay does not depend upon any properties
peculiar to enterotoxin B. It should be possible to
develop similar assays for other toxins, hormones, or
other compounds for which suitable antibodies and stable

binding proteins are required.

BIBLIOGRAPHY

BIBLIOGRAPHY

Angellotti, R., M. J. Foter, and K. H. Lewis. 1961.
Time-temperature effects on salmonella and
staphylococci in foods. Am. J. Pub. Hlth.
51: 76-88.

Avena, R. M., and M. S. Bergdoll. 1967. Purification
and some physiochemical properties of enterotoxin

C. Staphylococcus aureus strain 361. Biochem.
'6: 1474-1480.

Axen, R., and J. Porath. 1966. Chemical coupling of
enzymes to cross-linked dextran (Sephadex).
Nature 210: 367-369.

Bennett, R. W., B. A. Keosyan, S. R. Tatini, H. Thota,
and W. S. Collins. 1973. Staphylococcal entero-
toxin: A comparison study of serological detec-
tion methods. Canadian Institute of Food Science
and Technology J. ‘6: 131-134.

Bergdoll, M. S. 1966. Immunization of rhesus monkey with
enterotoxin B. J. Infect. Disease 116: 191-196.

Bergdoll, M. S. 1970. Enterotoxins. In Microbial Toxins
Vol. III, edited by T. C. Montie, S. Kadis, and
S. J. Ajil. Academic Press Inc., New York, N.Y.
pp. 265—296.

 

Bergdoll, M. S. 1972. Enterotoxin. In The Staphylococci,
edited by J. 0. Cohen. Wiley-Interscience, New
York, pp. 301-331.

 

Bergdoll, N. S., C. R. Borja, and R. M. Avena. 1965.
Identification of a new enterotoxin as enterotoxin
C. J. Bacteriol. 22; 1481-1489.

Bergdoll, M. S., C. R. Borja, R. Robbins, and K. F. Weiss.

1971. Identification of enterotoxin E. Infection
and Immunity. '4: 593-595.

93

94

Bergdoll, M. S., R. Reiser, and J. Spitz. 1976. Staphylo-
coccal enterotoxin detection in food. Food
Technol. 3215): 80-84.

Bergdoll, M. S., and R. N. Robbins. 1973. Characteri-
zation of type of staphylococcal enterotoxins.
J. Milk Food Technol. 36: 610-612.

Bergdoll, M. 8., H. Sugiyama, and G. M. Dack. 1959a.
Staphylococcal enterotoxin. I. Purification.
Arch. Biochem. Biophys. 85: 62-69.

Bergdoll, M. S., M. J. Surgalla, and G. M. Dack. 1959b.
Staphylococcal enterotoxin. Identification of
a specific precipitating antibody with enterotoxin
neutralizing property. J. Immunol. 83: 334-338.

Berson, S. A., R. S. Yalow. 1973. Radioimmunoassay. In
Peptide hormones, edited by S. A. Berson and R. S.
Yalow. American Elsevier Publishing Company,
Inc., New York, N.Y., pp. 84-120.

Berson, S. A., R. S. Yalow, A. Baumann, M. A. Rothschild,
and K. Newerly. 1956. Insulin-1311 metabolism
in human subject: Demonstration of insulin bind-
ing globulin in the circulation of insulin treated
subjects. J. Clin. Invest. 35: 170-190.

Borja, C. R., and M. S. Bergdoll. 1967. Purification
and partial characterization of enterotoxin C

produced by Sta h lococcus aureus strain 137.
Biochem. 6: I4E7—IZ73.

Bukovic, J. A., and H. M. Johnson. 1975. Staphylococcal
enterotoxin C: Solid-phase radioimmunoassay.
Appl. Microbiol. 32; 700-701.

Casman, E. P. 1960. Further serological studies of
staphylococcal enterotoxin. J. Bacteriol. 19:
849-856.

Casman, E. P. and R. W. Bennett. 1963. Culture medium
for production of staphylococcal enterotoxin A.
J. Bacteriol. ‘86: 18—23.

Casman, E. P. and R. W. Bennett. 1964. Production of
Antiserum for staphylococcal enterotoxin. Appl.
Microbiol. 12: 363-367.

Casman, E. P. and R. W. Bennett. 1965. Detection of
staphylococcal enterotoxin in food. Appl.
Microbiol. ‘13: 181-189.

95

Casman, E. P., R. W. Bennett, A. E. Dorsey, and J. A.
Issa. 1967. Identification of a fourth staphylo-
coccal enterotoxin, enterotixin D. J. Bacteriol.
24: 1875-1882.

Catt, K. J., H. D. Niall, and G. W. Tregear. 1966. Solid-

phase radioimmunoassay of human hormone. Biochem.
J. 100, 31c.

Catt, K. J., H. D. Niall, C. W. Tregear, and H. G. Burger.
1968. Disc solid-phase radioimmunoassay of human
lutienizing hormone. J. Clin. Endoclinol. Metab.
28: 121-126.

Catt, K. J., and G. W. Tregear. 1967. Solid-phase radio-
immunoassay in antibody-coated tubes. Science.
158: 1570-1571.

Chu, F. S., K. Thadhani, E. J. Schantz, and M. S. Bergdoll.
1966. Purification and characterization of
staphylococcal enterotoxin A. Biochem. 5: 3281-
3289.

Collins, W. S., A. D. Johnson, J. F. Metzger, and R. W.
Burnett. 1973. Rapid solid-phase radioimmunoassay
for staphylococcal enterotoxin A. Appl. Micro-
biol. 25: 774-777.

Collins, W. S., J. F. Metzger, and A. D. Johnson. 1972.
A rapid solid-phase radioimmunoassay for staphylo—
coccal B enterotoxin. J. Immunol. 108: 852-856.

Cuatrecasas, P. 1970. Protein purification by affinity
chromatography, derivation of agarose and poly-
acrylamide gel. J. Biol. Chem. 245: 3059-3065.

Cuatrecasas, P., and C. B. Anfinsen. 1971. Affinity
chromatography. In Methods in enzymology,
vol. XXII edited by W. B. Jakaby. Academic Press,
Inc., New York, pp. 345-378.

 

Dalidowicz, J. E., S. J. Silverman, E. T. Stefanye, and
L. Spero. 1966. Chemical and biological proper-
ties and alkylated staphylococcal enterotoxin B.
Biochemistry 5: 2375-2381.

Davison, E., and G. M. Dack. 1939. Some chemical and
physical studies of staphylococcal enterotoxin.
J. Infect. Dis. 64; 302-306.

96

Denny, C. B., P. L. Tan, and C. W. Bohrer. 1966. Heat
inactivation of staphylococcal enterotoxin A.
J. Food Sci. 33: 762-767.

Donnelly, C. B., L. A. Black, and K. H. Lewis. 1964.
Occurrence of coagulase positive staphylococci
in cheddar cheese. Appl. Microbiol. 33: 311-315.

Feldman, H., and D. Rodbard. 1971. Mathematical theory
of radioimmunoassay. In Principles gf cqmparative
protein bindingassays, edited by W. D. Odell and
W. H. Daughaday. J. B. Lippincott Co., Phila-
delphia, pp. 158-173.

 

 

 

Freeman, T. 1967. Trace labelling with radioiodine.
In Handbook 93 experimental immunology, edited
by D. M. Weir. F. D. Davis Company, Philadelphia,
pp. 579-607.

 

Freeman, T., C. M. E. Mathews, A. S. McFarlane, H. Benhold,
and E. Kallee. 1959. Albumin labelled with
131iodine in an analbuminaemic subject. Nature
183, 606.

Galskov, G. 1972. Radioimmunochemical corticotropin
determination. Acta Endocrinol 69 Suppl. 162, 5.

Gasper, E., R. C. Heimsch, and A. W. Anderson. 1973.
Quantitative determination of type A staphylo-
coccal enterotoxin by Laurell electroimmuno-
diffusion. Appl. Microbiol. 33; 421-426.

Genigeorgis, C., and W. W. Sadler. 1966. Immunofluor-
escent detection of staphylococcal enterotoxin B.
II Detection in foods. J. Food Sci. 33: 441-449.

Genuth, S., L. A. Frohman, and H. E. Leboritz. 1965.
A radioimmunological assay method for insulin
using insulin-1 51 and gel filtration. J. Clin.
Endoclinol. Metab. 33: 1043-1049.

Greenwood, F. C., W. M. Hunter, and J. S. Glover. 1963.
The preparation of 1251-labe11ed human growth
hormone of high specific radioactivity. Biochem
J. 33: 114-123.

Grodsky, G. M., and P. H. Forham. 1960. An immunochemi-
cal assay of total extractable insulin in man.
J. Clin. Invest. 33; 1070-1079.

97

Gruber, J., and G. G. Wright. 1967. Iodine-131 Labeling
of purified microbial and antigens by micro-
diffusion. Proc. Soc. Exp. Biol. Med. 126: 282-
284.

Hall, H. E. 1968. Enterotoxin and enzyme production by
a selected group of Staphylococcus aureus cultures.
BaCto PIOC. ' pp. 77-7go

 

Hall, H. E., R. Angelotti, and K. H. Lewis. 1965.
Detection of staphylococcal enterotoxin in food.
Health Lab. Sci. 3: 179-191.

Hallander, H. O. 1965. Production of large quantities of
enterotoxin B and other staphylococcal toxins on
solid media. Acta Pathol. Microbiol. Scand.

33: 299-305.

Hammon, W. McD. 1941. Staphylococcal enterotoxin: an
improved cat test, chemical and immunological
studies. Am. J. Public Health. 33: 1191-1198.

Herbert, V., K. S. Lau, C. W. Gottlieb, and S. J. Bleicher.
l965.’ Coated charcoal immunoassay of insulin.
J. Clin. Endocrinol. Metab. 33: 1375-1384.

Hilker, J. S., W. R. Heilman, P. L. Tan, C. B. Denny, and
C. W. Bohrer. 1968. Heat inactivation of entero-
toxin A from Staphylococcus aureus in veronal
buffer. Appl. MICrobiol. 3E: 308-310.

 

Huang, I-Y., and M. S. Bergdoll. 1970a. The primary
structure of staphylococcal enterotoxin B. II.
Isolation, composition, and sequence of chrymo-
tryptic peptides. J. Biol. Chem. 313: 3511-3517.

Huang, I-Y., and M. S. Bergdoll. 1970b. The primary
structure of staphylococcal enterotoxin B. III.
The cyanogen bromide peptides of reduced and
aminoethylated enterotoxin B, and the complete
amino acid sequence. J. Biol. Chem. 333: 3518-
3525.

Hunter, W. M., and F. C. Greenwood. 1962. Preparation
of iodine-131 labelled human growth hormone of
high specific activity. Nature 194: 925-926.

Hunter, W. M., and F. C. Greenwood. 1964. A radio-
immunoelectrophoretic assay for human growth
hormone. Biochem J. 33: 43-56.

98

Isojima, S., O. Naka, K. Koyama, and H. Adachi. 1970.
Rapid radioimmunoassay of human luteinizing hor-
mone using polymerized anti-human chorionic
gonadotrophin as immunoadsorbent. J. Clin.
Endocrinol. Metab. 33: 692-699.

Jay, J. M. 1970. Modern food microbiology. D. Van
Nostrand Company, New York, pp. 194-207.

Johnson, H. M., J. A. Bukovic, and P. E. Kauffmann. 1973.
Staphylococcal enterotoxin A and B: Solid-phase
radioimmunoassay in food. Appl. Microbiol. 33:
309-313.

Johnson, H. M., J. A. Bukovic, P. E. Kauffman, and J. T.
Peeler. 1971. Staphylococcal enterotoxin B:
Solid-phase radioimmunoassay. Appl. Microbiol.
33: 837-841.

Kabat, E. A. 1976. Structural concepts in immunology
and immunochemistry. Holt, Rinehart and Winston.
New York, pp. 204, 223.

Kauffman, P. E., and H. M. Johnson. 1975. Stability
of 125I-1abe11ed staphylococcal enterotoxins in

solid-phase radioimmunoassay. Appl. Microbiol.
33: 776-779.

Lau, K. S., C. W. Gottlieb, and V. Herbert. 1966. Pre-
liminary report on coated charcoal immunoassay
of human chorionic "growth hormone prolactin"
and growth hormone. Proc. Soc. Rxp. Biol. Med.
333: 126-131.

Lilly, H. D., R. A. McLean, and J. A. Alford. 1967.
Effect of curing salt and temperature on pro-
ductions of staphylococcal enterotoxin. Bact.
Proc., p. 12.

March, S. C., I. Parikh, and P. Cuatrecasas. 1974. A
simplified method for cyanogen bromide activation
of agarose for affinity chromatography. Anal.
Biochem. 33; 149-152.

Marchalonis, J. J. 1969. An enzymatic method for the
trace iodination of immunoglobulins and other
proteins. Biochem. J. 113: 299-305.

Markus, Z., and G. J. Silverman. 1968. Enterotoxin B
production by nongrowing cells of Staphylococcus
aureus. J. Bacteriol. 33: 1446-1447.

 

99

Markus, Z., and G. J. Silverman. 1969. Enterotoxin B
synthesis by replicating and non-replicating cells
of Staphylococcus aureus. J. Bacteriol. 91:
506-512. _—

 

Milone, N. A. 1962. Study on the use of tissue culture
for bioassay of staphylococcal enterotoxin. J.
Food Sci. '31; 501-507.

McFarlane, A. S. 1958. Efficient trace-labelling of pro-
tein with iodine. Nature 182, 53.

Morse, S. A., and R. A. Mah. 1967. Microtiter hemag-
glutination-inhibition assay for staphylococcal
enterotoxin B. Appl. Microbiol. 33: 58-61.

Morse, S. A., R. A. Mah, and W. J. Dobrogosz. 1969.
Regulation of staphylococcal enterotoxin B.
J. Bacteriol. 33: 4-9.

Munch-Peterson, E. 1962. Staphylococci in food and food
intoxication. J. Food Sci. ‘33: 692-710.

Nace, G. W., and P. Spradlin. 1962. Bleeding rabbits.
Turtex News 33: 26-29.

Nickerson, J. T., and A. J. Sinskey. 1974. Microbiology
of foods and food processing. American Elsevier
Publishing Co., New York, p. 247.

Odell, W. D., J. F. Wilber, and W. E. Paul. 1965.
Radioimmunoassay of thyrotropic in human serum.
J. Clin. Endocrinol. Metab. 33: 1179-1188.

Odell, W. D., G. A. Araham, W. R. Skowsky, M. A. Hescox,
and D. A. Fisher. 1971. Production of antisera
for radioimmunoassays. In Principles 93 com eti-
tive rotein-binding assays, edited—by W. B.
OdeIl, W. H. DaugHaday. J. B. Lippincott Co.,
Philadelphia, pp. 57-76.

 

 

Parker, C. P. 1976. Radioimmunoassay of biologically
active compounds. Prentice-Hall Inc., New Jersey,
p. 43.

Porath, J. R. Axen, and S. Emback. 1967. Chemical
coupling of protein to agarose. Nature 215:
1491-1492.

Read, R. B., and J. G. Bradshaw. 1966. Staphylococcal
enterotoxin B thermal inactivation in milk.

100

Reiser, R., D. Conaway, and M. S. Bergdoll. 1974.
Detection of staphylococcal enterotoxin in foods.
Appl. Microbiol. 31: 83-85.

Robern, H., M. Dighton, Y. Yano, and N. Dickie. 1975.
Double-antibody radioimmunoassay for staphylo-
coccal enterotoxin C2. Appl. Microbiol. 33:
525-529.

Robinson, P. J., P. Dunnill, and M. D. Lilly. 1972.
Factors affecting scale-up of affinity chroma-
tography of B-galactosidase. Biochim. Biophys.
Acta. 333: 28-35.

Satterlee, L. D., and A. A. Kraft. 1969. Effect of
meat and isolated meat proteins on the thermal
inactivation of staphylococcal enterotoxin B.
Appl. Microbiol. 31: 906-909.

Schaeffer, W. I., J. Gabliks, and R. Calitis. 1966.
Interaction of staphylococcal enterotoxin B with

cultures of human embryonic intestine. J. Bac-
teriol. 33: 21-26.

Schantz, E. J., W. G. Roessler, J. Wagman, L. Spero, D. A.
Dunnery, and M. S. Bergdoll. 1965. Purification
of staphylococcal enterotoxin B. Biochem. 3:
1011-1016.

Silverman, S. J., A. R. Knott, and B. Howard. 1968.
Rapid, sensitive assay for staphylococcal entero-
toxin and a comparison of serological methods.
Appl. Microbiol. 33: 1019-1023.

Silverman, G. J., J. T. R. Nickerson, D. W. Duncan, N. S.
Davis, J. S. Schacter, and M. M. Joselow. 1961.
Microbial analysis of frozen raw and cooked
shrimp. I. General result. Food Technol.

33: 455-458.

Skelley, D. S., L. P. Brown, and K. P. Besch. 1973.
Review radioimmunoassay. J. Clin. Chem. 33:
146-186.

Skom, J. H., and D. W. Talmage. 1958. Nonprecipitating
insulin antibodies. J. Clin. Invest. 31: 783.

Spero, L., H. M. Jacoby, J. E. Dalidowicz, and S. J.
Silverman. 1971. Gaunidination and nitroguani-
dination of staphylococcal enterotoxin B. Biochem.
Biophys. Acta. 333: 345-354.

 

SP

Su

81

 

 

101

Splittstoesser, D. F., G. E. R. Hervey, and W. P. Wetter-
gren. 1965. Contamination of frozen vegetables
by coagulase-positive staphylococci. J. Milk and
Food Technol. 33: 149-151.

Surgalla, M. J., M. S. Bergdoll, and G. M. Dack. 1952.
Use of antigen-antibody reactions in agar to
follow the progress of fractionation of anti-
genic mixture: application to purification of
staphylococcal enterotoxin. J. Immunol. 69:
357-365. _—

Surgalla, M. J., M. S. Bergdoll, and G. M. Dack. 1953.
Some observations on the assay of staphylococcal
enterotoxin by the monkey-feeding test. J. Lab
Clin. Med. 33: 782-788.

Troller, J. A., and W. C. Frazier. 1963. Repression of
Staphylococcus aureus by food bacteria. II.
Causes of inhibition. Appl. Microbiol. 11:
16 3-165. ‘—

Thorell, J. I., and B. G. Johanson. 1971. Enzymatic
iodination of polypeptides with 1251 to high
specific activity. Biochim. Biophys. Acta.
333: 363-369.

Walker, G. C., L. G. Harmon, and C. M. Stine. 1961.
Staphylococci in colby cheese. J. Dairy Sci.
33: 1272-1282.

Weirether, F. J., E. E. Lewis, A. J. Rosenwald, and R. E.
Lincoln. 1966. Rapid quantitative serological
assay of staphylococcal enterotoxin B. Appl.
Microbiol. 33: 284-291.

Yalow, R. S., and S. A. Berson. 1959. Assay of plasma
insulin in human subjects by immunological methods.
Nature 148: 1648-1649.

Yalow, R. S., and S. A. Berson. 1961. Immunological
specificity of human insulin: Application to
immunoassay of insulin. J. Clin. Invest. 33:
2190-2198.

Zehren, V. L., and V. F. Zehren. 1968. Examination of
large quantities of cheese for staphylococcal
enterotoxin A. J. Dairy Sci. 33: 635-644.