III III" III II III III II IIIIIIIIII

WWI .. 3. 132915105

III/III

 

 

 

 

 

 

llBRARY
Michigan Stan
University

 

 

 

This is to certify that the

thesis entitled

MOLECULAR ANALYSIS OF PLASMIDS ON SPORADIC OCCURRENCE
OF HEMORRHAGIC COLITIS ASSOCIATED WITH
ESCHERICHIA COLI 0157:H7
IN WESTERN MICHIGAN

presented by

PENGCHIN CHEN

has been accepted towards fulfillment

of the requirements for
CLINICAL LABORATORY
M . S . degree in SCIENCES

 

MWW

Major professV

Date Aug. 14, 1987

 

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

 

 

 

 

 

MSU

LIBRARIES
“

 

 

 

RETURNING MATERIALS:
Place in book drop to

remove this
your record.
be charged i

checkout from
FINES will
f book is

returned after the date
stamped below.

 

(Hi i % n02

 

{ 38531131

A
q

i‘I/3,“

I"

J

 

HOLECULAR.ANALXSIS 0F FLASHIDS 0N SPORADIC OCCURRENCE

OF HEHDRRHAGIC COLITIS ASSOCIATED WITH

ammonia mi; 019%—

IN’WESTERN'HICHIGAN

By

Pengchin Chen

A.THESIS

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

MASTER OF SCIENCE

Medical Technology Program

1987

ABSTRACT

MOLECULAR ANALYSIS or PIASHIDS 0N SPORADIC OCCURRENCE
or HEMORRHAGIC COLITIS ASSOCIATED WITH
ESCHERICHIA QII 0157:H7
IN WESTERN HICHICAN
By

Pengchin Chen

Escherichia £211 0157:H7 has been epidemiologically linked to outbreaks
of hemorrhagic colitis associated with fast—food restaurants and nursing
homes. Sporadic cases now exceed those associated with outbreaks.
Seventeen strains of Escherichia 92;; 0157:H7 isolated from patients
(Michigan residents) with hemorrhagic colitis were examined for their
plasmid content. All seventeen 0157:H7 strains possessed a large 72 Md
plasmid. Three major plasmid patterns were found but did not correlate
to geographic areas of occurrence. Plasmid pattern and restriction

profile analysis indicates Escherichia coli 0157:H7 strains from

 

sporadic cases in Western Michigan are related but not identical.

This is dedicated to my family for their continuing support.

iii

ACKNOWLEDGMENTS

I would like to thank my major advisor, Dr. Robert Martin, for the
outstanding job he has done in providing me with encourgement and
advice. I am also indebted to Dr. Martin for providing laboratory space
and adequate funding to allow this project to be completed. I would
like to thank my committee, Dr. Douglas Estry and Dr. Sharon Zablotney,
for their assistance in acccomplishing the requirements for my Master's
degree.

Special thanks to Dr. Babara Robinson, Mr. William Schneider and
the staff of Diagnostic Microbiology Section in Michigan Department of

Public Health for their technical assistance.

iv

TABLE OF CONTENTS

LITERATURE REVIEW .

Characteristics of Escherichia coli .
Escherichia coli as a pathogen .
Diarrheagenic Escherichia coli .

ETEC . . . . . .

EIEC .

EPEC .

EAEC .

EHEC .
Escherichia coli 0l57: H7 .

 

Molecular genetic techniques as epidemiological tools .

Plasmid pattern analysis . .
Isolation and purification of plasmid DNA .
Size of plasmid molecules . . .
Restriction endonuclease digestion .

MATERIALS AND METHODS .

Bacterial strains and plasmids .

Chemicals and reagents .

Growth media .

Equipment . . .

Restriction endonuclease . .
Kado and Liu rapid plasmid isolation procedure .
Agarose gel electrophoresis . .
Growth media and amplification of plasmids .
Extraction and concentration of pBR322 .
Isolation of DNA from agarose gel .

Restriction endonuclease analysis

RESULTS .

Growth media and plasmid amplification .
Plasmids from various isolation steps .
Plasmid patterns of E. coli 01572H7
Restriction profiles of E. coli 0157:H7 .
Restriction profile comparison . .
Restriction profiles of 72Md plasmid .

'U
m
0’0
t-‘(D

oouxicxo‘uwuwwI-I

18

. 18
. 18
. 19

2O

. 21

. 22
24

. 24

. 25
26
27

28

. 28
28
31
31
36
37

DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Culture conditions . . . . . . . . . . . . . . . . . . . . . 40
Plasmid isolation methods . . . . . . . . . . . . . . . . . 41
Plasmid pattern . . . . . . . . . . . . . . . . . . . . . . 42
Restriction profile . . . . . . . . . . . . . . . . . . . . 43
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Appendix 1: Identification of E. coli 0157:H7 . . . . . . . 46
Appendix 2: Transformation of E. coli HBlOl by pBR322. . . . 47
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

vi

Table

LIST OF TABLES

Biochemical characteristics of Escherichia coli

0 serogroups of E. coli in normal feces and various
infections in humans

A range of plasmids in bacteria

Procedures commonly used for isolating plasmid DNA

Strains of E. ggli used in plasmid analysis

Restriction enzymes used to digest plasmid DNA

Bgl I restriction profiles of E. 99;; 0157:H7

Restriction profile comparisons of representative
strains from each plasmid group and isolated

small plasmids

Restriction profiles of strain 51560

vii

Page

12
14
19
21

34

36

37

LIST OF FIGURES

Figure

1 Flow diagram of plasmid DNA isolation by Kado-Liu
prodcedure

2 Plasmid patterns of two E. coli 0157:H7 strains in

different growth media and amplification procedure

 

3 pBR322 DNA from varying extraction and concentration
steps

4 Plasmid pattern analysis of E. coli 0157:H7

 

5 Bgl 1 restriction profiles of E. coli 0157:H7
6 Restriction profile comparisons of representative
strains from each plasmid group and isolated

small plasmids

7 Restriction profiles of 51560 and 61476 after
restriction endonuclease digestions

viii

Page

23

29

3O

32

33

35

38

 

 

 

pe
pr

me

fr

de.

gr.-
5a]

Sal

Cle.

5011

LITERATURE REVIEW

Characteristics of Escherichia coli

The organism now known as Escherichia coli was first isolated and

described by Theodor Escherich in 1885 under the name Bacterium coli

commune. Escherichia coli, which belongs to the family

 

Enterobacteriaceae is listed together with Escherichia blattae [14] in
the genus Escherichia [76].

E. coli is a short Gram-negative, non-sporeforming and, usually,

 

peritrichous and fimbriate bacillus. A capsule or microcapsule is often
present. Having both a respiratory type and a fermentative type of
metabolism, E. coli grows under facultative anaerobic condition.

Lactose-fermenting organisms with the appearance of E. coli are

 

frequently referred to as "coliforms" but they cannot, without further

definitive identification, be regarded as E.

coli (Table l) [100].

 

Colonies on nutrient agar may be smooth (S), low convex, moist,
gray, with a shiny surface and entire edge, and easily dispersible in
saline; or may be rough (R), dry and difficult to disperse well in
saline. Mucoid and slime-producing forms occur. In broth S strains of
E. gall produce a uniform turbidity, while R strains tend to leave a
clear supernant medium over a granular deposit of growth [76]. A
soluble alpha hemolysin can be demonstrated on media containing washed

1

 

 

 

 

1365'
H H
intr
aVai

{69}

2
erythrocytes and there is a cell-associated beta hemolysin which is

released from cell by lysis [95].

Table 1. Biochemical characteristics of Escherichia coli

 

Optimum growth temperature 37C
Catalase +
Oxidase -
Indole production +
Methyl red production +

Voges-Prokauer
Citrate utilization -

NO3‘ reduction +
Gelatin liguefaction -
H25 production -
Urease -
Phenylalanine -
Fermentation Mixed acid
Gas from glucose (37C) +
Acid from arabinose (May be delayed) +
lactose (May be delayed) +
malose (May be delayed) +
mannitol (May be delayed) +
sorbitol (May be delayed) +
trehalose (May be delayed) +
adonitol -
inositol -
esculin Various for different strains
salicin Various for different strains
sucrose Various for different strains
xylose Various for different strains
Mole % G+C 50-51

 

A variety of antigens may be used to type E. coli strains. The

 

best known of these are the O somatic lipopolysaccharide, K capular, and
H flagellar antigens which forms the basis of the typing scheme first
introduced by Kauffman [48]. The serotyping of E. 92;; fimbriae is also
available to expand previous typing schemes [77]. Bacteriophage typing

[69] and colicin typing [30] also have been employed but they have not

 

 

 

C01

ma

CC

CC

CC

come into general use.

Escherichia coli as a pathogen

Escherichia coli is regarded as a member of the normal flora of

 

man. Colonization by E. 99;; takes place soon after birth [5].
Although E. 99;; is a normal inhabitant of the intestinal tract of
animals and man, it is also asociated with a variety of diseases. Most
disease is associated with relatively few serological groups of E. gal;
even though a large number exist in nature. These serogroups tend to be
host-specific.

In young animals, E. 99;; infections are usually referred to as
colibacillosis. Two main forms are recognized: (I) systemic
colibacillosis caused by bacteremic strains of E. gall and (2) enteric

colibacillosis caused by enteropathogenic strains of E. coli. Coliform

 

mastitis is another form of E. 99;; infection but occurs in adult cattle
[70].

E. 92;; is a multipotent pathogen that has evolved the ability to
cause disease in several body systems of human. Diseases caused by E.
gal; in man include gastro-intestinal infections, urinary tract
infections, meningitis, wound infections and septicemia. There are a

variety of serogroups of E. coli found in normal feces and in various

 

infections (Table 2). Wound infections due to E. coli have long been
recognized. They most commonly follow surgical operations, such as
appendectomy, in the course of which the alimentary tract is entered.

Meningitis due to E. £011 is predominantly an infection of the newborn.

4
The source of septicemia may be either a discrete lesion, or a more
diffuse infected tissue lesion or, most commonly, a urinary tract
infection. Urinary tract infections may involve the kidney, ureter,
bladder and urethra. The organism most commonly isolated from all types

of urinary tract infections is E.

coli [71].

Table 2. O serogroups of E. coli in normal feces
and various infections inhumansa

 

 

 

 

 

 

 

Serogroups
Normal feces 01,02,0S,O6,07,08,0l8,020,025,045,075,081
Septicemia 01,02,04,06,o7,08,o9,011,018,o22,o25,o75
Meningitis 01,06,07,0l6,018,083
UIIb 01,02,04,06,07,08,09,011,022,025,062,075
ETECc 01,06,07,08,09,015,020,025,027,060,063,

070,078,085,088,089,099,0lOl,OlO9,0llh,
0115,0126,0l28,0142,0148,0153,0159

 

 

 

 

 

 

 

EIECd 028,0112,0115,0124,0136,0143,0144,0147,
0152,0164

EPECe 018,020,026,044,055,096,0111,0112,0114.
0119,0124,0125,0126,0127,0128,0l42,0158,
0159

EHECf 026,0111,0157

EAECg 0 Serogroups not yet defined

aLevine[54], Sussman[100] eEnteropathogenic E. coli

rinary tract infection Enterohemorrhagic E. coli
cEnterotoxigenic E. coli gEnteroadherent E. col

dEnteroinvasive E, coli

Diarrheagenic Escherichia coli

There are a number of pathogenic strains of E. 99;; that cause
distinct syndromes of diarrheal diseases. Based on distinct virulence
properties, different interactions with the intestinal mucosa, discrete
clinical syndromes, differences in epidemiology ,and distinct O:H
serotypes, diarrheagenic E. 99;; can be classified into five main

categories [56]. The five categories are: (l) enterotoxigenic coli

lm

 

[31,90]; (2) enteroinvasive E. coli [26]; (3) enteropathogenic E. coli

 

[33,74]; (4) enteroadherent E. coli [65,66]; (5) enterohemorrhagic E.

 

coli [87].

ETEC Enterotoxigenic E. coli are a major cause of infant diarrhea

 

in less developed countries [8], one of the main bacterial causes of
dehydrating infant diarrhea in developing areas [7], an infection
correlated with adverse nutritional consequences [9], and the agent most
frequently responsible for travellers' diarrhea [27,68,89]. The clinical
features of ETEC infection are watery diarrhea, nausea, abdominal
cramps, and low grade fever. ETEC infection is acquired by ingesting
contaminated food or water. The bacteria possess attachment or
colonization factors that allow them to overcome the peristaltic defense
mechanism of the small intestine. Attachment factors allow bacteria to
colonize the proximal small intestine where they elaborated heat-labile
enterotoxin (LT) or heat stable enterotoxin (ST). A limited number of
O:H serotypes occurr repeatedly throughout the world and account for a
majority of the ETEC strains. Usually these serotypes contain plasmids

which encode the proteins LT, ST, and colonization factors [53,85].

6

EIEC Enteroinvasive E. 99;; cause an invasive dysenteric form of
diarrheal illness. EIEC, which usually affect adults, has been reported
in some foodborne outbreaks. Clinically, the illness is marked by fever,
severe abdominal cramps, malaise, toxemia, and watery diarrhea followed
by gross dysentery consisting of scanty stools of blood and mucus which
contains many polymorphonuclear leukocytes. EIEC have a predilection
for colonic mucosa as the favored site of host parasite interaction.
Like Shigella, their cardinal pathogenic feature is the capacity to
invade and proliferate within epithelial cells and cause eventual death
of the cell [26]. The invasive capacity of both EIEC and Shigella is
dependent on the presence of large 140 Md plasmids [38] coding for the
production of several outer membrane proteins involved in invasiveness
[35]. The strains of EIEC are distinct serotypes from ETEC and EPEC.

EPEC Enteropathogenic E. coli strains cause a distinctive

 

ultrastructural histopathologic lesion in human intestines and cause
diarrhea by mechanisms distinct from LT, ST, and Shigella-like
invasiveness [33]. EPEC was reported in outbreaks of diarrhea in infant
nurseries and both sporadic and epidemic infant diarrhea in communities.
Clinically, the illness is characterized by fever, malaise, vomitting,
and diarrhea with predominant amount of mucus but without gross blood.

A high percentage of EPEC strains adhere to Hep-2 cells in tissue
culture, a property rare in ETEC, ETIC or normal flora strains [24].
EPEC strains from patients with diarrhea possess a plasmid about 55-65
Md in size that encode the EPEC adherence factor necessary for adherence
to Hep-2 cells [4]. A plasmid encoding 94 Rd protein [55] has been

found in all important serotypes such as those in serogroups 055, 0111,

7
0119, 0127, and 0142 but has not been found in ETEC, EIEC, EHEC or E.
gall that cause meningtitis or pyelonepritis.

EAEC Enteroadherent E. galll incriminated in travellers' diarrhea
with no recognized bacterial enteropathogens, are identified only by
their property of adherence to Hep-2 cells. EAEC can cause diarrhea
without blood or fecal leukocytes. EAEC do not elaborate LT, ST, or
elevated levels of Shigella-like toxin, neither do they invade
epithelial cells or possess EPEC adherence factor plasmids. No
serogroups have been defined in this category.

EHEC Enterohemorrhagic E. gall refers to the strains 0157:H7
which was a serotype not previously recognized as a cause of diarrheal
disease in humans till 026:H11 and 0111 were reported to be associated

with hemorrhagic colitis [57]. E. coli 0157:H7 is the causative agent

 

of hemorrhagic colitis and there is also strong incrimination of 0157:H7
as a cause of hemolytic uremic syndrome. The clinical syndrome was
notable in that bloody and copious diarrhea, unaccompanied by fecal
leukocytes, was seen in afebrile patients. These features distinguish
it from classic dysentery due to Shigella or EIEC, infections
characterized by fever and scanty stools of blood and mucus containing
many fecal leukocytes. EHEC do not elaborate LT or ST, therefore can be
distinguished from ETEC. Although the initial stages of bacterial
attachment to the apical surface of epithelium and the destruction of
microvilli caused by EHEC are similar to lesions produced by EHEC, EHEC
and EPEC infections can still be clearly differentiated by pathologic
features in gnotobiotic piglets [29,102]. EPEC involve the entire

intestine of piglet, EHEC only the cecum and colon and EPEC lesions are

 

in
he

CI

f:
at

SL

St

PE

8
generally less severe. Some infiltration by leukocytes is seen with

EPEC infection but not with EHEC infection.

Egcherichia coli OlSZ;HZ

In 1982, a report of two restaurant-related outbreaks of disease
in the United States described a distinct clinical syndrome of
hemorrhagic colitis [87]. The cases were characterized by severe,
crampy abdominal pain, little or no fever, and an initially watery
diarrrhea followed by gross blood. Gastrointestinal contrast studies
and endoscopy performed for some patients were suggestive of vascular
insufficiency. Although stool examination for known enteric pathogens
was unrevealing, E. gall 0157:H7 was isolated from most stools obtained
within four days of the onset of symptoms. Nine isolates were obtained
from 43 patients. E. gall 0157:H7 was isolated from a frozen meat patty
at a supply warehouse that shipped to the restaurant, a finding
suggesting an origin early in the food chain [87]. An outbreak of
gastrointestinal illness in a Canadian nursing home again implicated E.
gall 0157:H7 as the causative agent [18]. In this outbreak, 31 of 353
residents became ill. Eighteen of the affected patients had bloody
diarrhea, and eight had watery diarrhea without blood. The outbreak
lasted 18 days and, like the two outbreaks in the United States, was
associated with meat consumption. Unlike the outbreak in the United
States, mildly ill patients without hematochezia were identified.
Person-to-person spread was also suggested.

Subsequently, outbreaks or sporadic occurrence of hemorrhagic

9
colitis caused by 0157:H7 were recognized in the United States
[16,37,78,86,88] and Canada [l7,40,42,51,66,84,97]. E. gall 0157:H7 was
also reported to be associated with hemolytic uremic syndrome
[18,32,45,46,47,80,86,96], a triad of acute nephropathy, thrombo-
cytopenia, and microangiopathic hemolytic anemia [15], in addition to
other enteric pathogens such as Shigella [50], Salmonella [3], Yersinia
[81], Qamaylobacter [l9], and Enteroviruses [2,75,83]. Although the
initial evidence indicated that E. gall 0157:H7 was a rare human
pathogen, awareness of the unusual illness led to the detection of the
associated agent and emergence of 0157:H7 as an enteric pathogen of
public health importance in North America.

E. gall 0157:H7 grows well between 30 and 42C with 37C being
optimal for growth. The organism can survive for 9 months at -20C, the
storage temperature of meat, with little change but grow poorly in the
temperature range (44 to 45.5C) generally used for recovery of E. gall
from food [25]. The organism does not produce LT or ST and is not
enteroinvasive [104]. 0157:H7 strains from persons with hemorrhagic
colitis and hemolytic uremic syndrome develop a phage-encoded potent
cytotoxin active on Hela and Vero cells [73,94,99]. In addition,
0157:H7 possess a 72 Md plasmid [42,44,104]. Recently, Karch et al.
[44] have shown this plasmid is required for expression of a new
fimbrial antigen responsible for adhesion to epithelial cells.

Evidence has been presented to consider routine culturing of
diarrheal stools, especially those with blood, for E. gall 0157:H7
because of the mounting clinical and epidemiologic importance [78,86].

Epidemiological studies of outbreaks of infectious diseases have focused

10
on bacterial pathogens for their behavior, spread and presence in
infections. Traditionally, bacterial epidemic strains have been defined
by genus, species, biotype, antibiotics resistance pattern, phage type
or serotype. Routine serotyping of E. gall 0157:H7 is impractical for
most clinical microbiology laboratories. E. gall 0157:H7 has been
isolated based upon the characteristics that it does not ferment D-
sorbitol rapidly (about 95% of other E. gall ferment D-sorbitol) [63],
possession of H antigen 7 (about 10% of other E. gall have this
particular antigen) [28] and by detection of cytotoxic toxin (Shigella-
like toxin) in the Vero cell line [42,72]. Yet no single method based on
phenotypic characteristics has been used successfully to establish

irrefutably the identity of the bacterial strain.

Eolecular genetic technigues as epidemiological tools

Identification of epidemic strains by genetic means offers the
potential of bypassing such problems as autoagglutination in serotyping
and the necessity for selecting precise environmental conditions for
optimal toxin elaboration. The new and simple genetic technique can be
applied across genus and species barriers and they may be more direct,
more specific, and possibly more rapid than other procedures [103].

The molecular detail and diversity that can be seen in plasmids
make them discriminating epidemiological markers. Plasmid pattern
analysis followed by restriction endonuclease digestion has been used in
studying epidemiology of Gram-negative bacilli, including both

Paeudomonaa aegaginosa and various species of Enterobacteriaceae,

11
associated with a variety of outbreaks [41]. Similar molecular genetic
methods have also been used for Gram-positive cocci [91] and for Gram-
negative cocoi [12]. Several papers using molecular techniques to

investigate outbreaks or sporadic cases of E. coli 0157:H7 have also

been published [44,45].

Plasmid pattern analysis

Plasmids are extrachromosomal elements which can replicate
autonomously. They are double-stranded, closed circular DNA that range
in size from 1 Kb (0.56 Md) to greater than 200Kb (132 Md) [58]. Most
of the plasmid DNA inside bacteria is in the form of covalently-closed-
circle (CCC), meaning that there are no breaks in either of the
nucleotide strands which comprise the double helix. Most of the CCC
plasmid molecules isolated from bacteria are twisted to form supercoiled
molecules which have superhelix twists. If one of the two
polynucleotide strands in a closed-circular plasmid is broken, or
nicked, a relaxed open circle is formed. When both polynucleotide
strands are broken, a linear molecule is formed. Some of the very large
plasmids are particularly difficult to keep in the CCC form during
isolation and purification. Other forms of plasmid DNA also occur in
cell lysates, including dimers, trimers and other multimers. Catenanes
(interlocked rings) also occur. These are ususally rare and are
believed to result from errors in replication [36].

F factor (the causative agent of gene transfer between E. gall

strains) and R factors (the drug resistance transfer agents) are among

the characteristics encoded by plasmids.

coding for enzymes that are advantageous to the bacterial host.

12

Some plasmids contain genes

Among

the phenotypes conferred by different plasmids are: resistance to

antibiotics (R factors), production of antibiotics, production of

colicins, production of enterotoxins, and production of restriction and

modification enzymes [58].

3) [10].

Numerous plasmids have been described (Table

Table 3. A range of plasmids in bacteria

 

 

 

Original host Plasmid Phenotype recognized Size(Md)
E. goli F transfer 62.5
lambda dv lambda immunity 8.6
ColEl colicinogeny 4.2
Ent enterotoxin 53
15 cryptic 1.5
Shigella Collb-P9 colicinogeny 61.5
Pl phage; restriction 67
R100 drug resistance 58
Salmonella typhimurium series of drug resistance
type 29 R factors
Pseudomonas aeruginosa RPl drug resistance 38
Eseudomonas putida TOL toluene degradation 78
SLgthlococcua aureus pl258 resistance to antibiotics,
heavy metals 19
Streptomyces coelicolor SCPl drug resistance and
production ?
_Agrobacteriua tumefaciena Ti plant tumor 95-156

 

 

 

However, many plasmids are cryptic (confering no recognizable

phenotype).

of bacterial strains.

Such plasmids exist in bacteria and can be distinct markers

A simple and straightforward application of

molecular genetic techniques to epidemiological investigation is the use

of plasmid "finger printing" by electrophoresis.

The technique provides

fairly accurate estimates of plasmid size and content to identify

13
epidemic strains of bacteria and epidemic plasmids that have spread
through several different bacterial species. Plasmid pattern analysis
is rapid, relatively simple to perform, as specific as phage typing and

superior to biotyping and resistance typing.

Iaalatian and parification of plasmid DNA

Isolation of plasmid DNA is based on the two major differences
between bacterial DNA and plasmid DNA: (1) the bacterial chromosome is
much larger than the DNA of plasmids; and (2) plasmid DNA is extracted
in a covalently circular form while chromosomal DNA extracted from the
cell is in the form of broken, linear molecules. Most purification
protocols therefore involve differential precipitation steps in which
the long strands of chromosomal DNA are entangled in the remnants of
lysed cells and are preferentially removed. The protocols also take
advantage of the distinctive properties of closed circular DNA. Because
each of the complementary strands of plasmid DNA is a covalently closed
circle, the strands cannot be separated by conditions such as heating or
exposure to alkali which break most of the hydrogen bonds in DNA.

Closed circular molecules regain their native configuration when cooled
or returned to neutral pH while chromosomal DNA remains in the denatured
state [59].

A variety of methods have been developed to isolate plasmids from
bacteria. All methods involve five basic steps: (1) growth of bacteria
and amplification of plasmids; (2) harvesting and lysis of bacteria; (3)

remove of chromosome and debris; (4) concentration of the plasmid DNA;

14

and (5) purification of the plasmid (Table 4) [10].

Table 4. Procedures commonly used for isolating plasmid DNA

 

(l)Growth and

amplification

(2)Lysis of cell

(3)Removal of
chromosome
and debris

(4)Concentration
of remaining
DNA

(5)Purification

Grown in broth or on agar plate
Amplified by chloramphenicol

Lysozyme, detergent

Clearing spin to precipitate chromosome and
membrane (prior treatment with high
concentration of salt may improve separation)

Controlled shearing, alkaline denaturation,
and rapid renaturation, followed by removal of
single strained DNA with nitrocellulose or
phenol

Spin into CsCl cushion

Concentrated by butanola
Precipitated with sodium chloride
Precipitated with sodium acetate
Precipitate with ammonium acetateb
Precipitate with polyethylene glycol
Precipitate with ethanol

Precipitate with isopropanolc

Ethidium bromide/cesium chlorided
Vertical dye-buoyante
Chromatography on hydroxylapatitef
High salt sepharose chromatographyg
Dye affinity chromatographyh

RPC-S reverse phase chromatographyi

 

aManiatis [62]
bMaxam [67]
cBirnboim [6]

dClewell [21]
eStougaard [98]
fShoyab [93]

ECorneilis [23]
Buenemann [13]
iWells [105]

Elgrascale preparation as alternatives

Although the methods listed in Table 4 yield plasmid DNA which is

approximately 90% pure, they suffer from the drawback of being either

time consuming or expensive and are applicable only in research

15

laboratories. In addition, they are hampered by difficulties when many
isolates are examined at the same time. Therefore, analysis is often
performed only with representative strains and their epidemiolgical
backgrounds is not clear.

Recently, rapid microscale isolation procedures that allow plasmid
DNA material to be used directly in restriction endonuclease analysis
have been reported [43,49,101]. Such rapid procedures have several
advantages; (a) plasmid obtained can be directly used for restriction
analysis, (b) large and small plasmids are detectable, (c) the procedure
is applicable to a variety of bacteria, and (d) a large number of

bacterial cultures can be examined at the same time [101].

Size af plasmid molecules

Two methods were originally used to determine the molecular weight
of plasmid DNA. By using sucrose gradients also containing particles of
a known sedimentation coefficient (S value), the S value for a given
plasmid DNA can be determined. Empirical equations exist for both CCC
DNA and 0C DNA that relate such S value to molecular weight [22]. The
second method involves measurement of the contour length of plasmid
molecules using electron microscopy [52]. Since OC DNA molecules are
more easily measured than CCC molecules, CCC molecules are sometimes
converted to OC DNA by X-ray irradiation [20].

Determination of S values has now largely given way to sizing
using agarose gels. This can be done with intact CCC DNA or with

fragments produced by cleaving all molecules in a preparation with a

16
site specific restriction endonuclease followed by summation of the
sizes of the fragments. The technique is simple, rapid to perform and
capable of resolving mixtures of DNA fragments that cannot be separated
adequately by other sizing procedures. Furthermore, the location of DNA
in the gel can be determined directly: bands of DNA are stained with low
concentration of fluorescent, intercalating dye ethidium bromide and as
little as 1 ng of DNA can be detected by direct examination of the gel
in UV light [92]. The electrophoretic migration rate of DNA through
agarose gel is dependent upon the molecular size of DNA, the
conformation of DNA, the agarose concentration, and the applied current.
When other parameters hold constant, molecules of linear, duplex DNA
travel through gel matrices at rates that are inversely proportional to

the loglo of their molecular weights [39].

Restriction endonuclease digestion

Restriction endonucleases are enzymes, isolated chiefly from
prokaryotes, that recognize specific sequences within double stranded
DNA. In nature, they serve to restrict (cleave) foreign DNA
incorporated into bacterial genome by transformation. The bacterium's
own genome is protected by being methylated on an adenine or cytosine
residue (by modification methylase) which render it unavailable for
digestion by its own enzymes. Restriction endonucleases can be
classified into three groups. Type I and type III enzymes carry
modification (methylation) and ATP requiring restriction (cleavage)

activity in the same protein. Both types of enzymes recognized

l7

unmethylated sequences in substrate DNA. Type I enzyme cleaves DNA
randomly, whereas type III enzymes cut DNA at specific sites. Type II
enzymes cut DNA within or near their particular recognition sequences,
which typically are four to six nucleotides in length with a twofold
axis of symmetry [60]. Because of the unique properties of type II
enzymes, they are very useful in characterizing or examining DNA.

Restriction endonuclease characterization is dependent on the
presence of specific nucleotide sequences. Restriction endonucleases
cleave DNA at points on the plasmid where specific nucleotide base
sequences occur, and the lengths of the intervening fragments can be
separated by electrophoresis to produce a restriction endonuclease
profile. If two plasmids are of the same size and yield identical
restriction endonuclease profile, especially when two or more enzymes

are used, they may be assumed to be identical or nearly so.

MATERIALS AND METHODS

Eacgerlal strains and plasmids

Seventeen E. gall 0157:H7 isolates identified in the Diagnostic
Microbiology Section, Michigan Department of Public Health with
confirmation of verotoxin production by Centers for Disease Control from
June 1985 through May 1987 were collected for plasmid analysis (Table
5).

Frozen competent E. coli HBlOl cells were purchased from Bethesda

 

Research Laboratories (BRL), Gaithersburg, MD. and Plasmid pBR322 was
purchased from Boehringer Mannheim, Indianapolis, IN. A molecular
weight standard, Hind 3 digested lambda phage, was purchased from

International Biotechnologies Inc., New Haven, CT.

gagmicals and reagents

Tris-hydroxymethyl-aminomethane (Tris), bromphenol blue, ficol,
sodium acetate, boric acid, ethylenediaminetetraacetic acid (EDTA),
ampicillin, chloramphenicol, magnesium sulfate, potassium chloride,
magnesium chloride, isoamyl alcohol, and ribonuclease A were purchased
from Sigma Chemical Company, St. Louis, MO. Hydrochloric acid

(concentrated), sodium hydroxide, and glacial acetic acid were from

18

19
Mallinckrodt, Paris, Kentucky. Chloroform (HPLC grade) and isopropanol
were from Aldrich, Milwaukee, WI. D-glucose, ether, and n-butanol were
from Fisher, Fair Lawn, NJ. Other chemicals and reagents include sodium
dodecyl sulfate (SDS; Pierce, Rockford, IL), phenol (redistilled nucleic
acid grade; BRL), Tris-HCl (Boehringer Mannheim), 100% ethanol (U.S.

Industrial Chemiscals, Tuscola, IL), and agarose (IBI).

Table 5 Strains of E. coli used in plasmid analysis

 

 

 

Lab. N0. Hospital | Pt. age] Source| CDC confirmed
50681 Blodgett Hosp., Grand Rapids 37 stool Verotoxin -
51396 Blod ett Hosp., Grand Rapids l9 stool Verotoxin +a
51560 NOCH , Grand Haven 10 stool Verotoxin +
6UNCL-3 Macomb Chd Hosp., Mt. Clemens 70 bowel Verotoxin +
61167 Blodgett Hosp., Grand Rapids 19 NA Verotoxin +
61385 NOCH, Grand Haven NA NA Not reported
61409 Gerber Memorial Hosp., Fremont 25 stool Verotoxin +
61427 NOCH, Grand Haven NA stool Verotoxin +
61476 NOCH, Grand Haven 20's NA Verotoxin +
61494 Blodgett Hosp., Grand Rapids 3 stool Verotoxin +
61582 NOCH, Grand Haven 15 stool Verotoxin +
61676 Blodgett Hosp., Grand Rapids 2 stool Verotoxin +
61951 NOCH, Grand Haven NA stool Verotoxin +
62017 NOCH, Grand Haven NA stool Verotoxin +
62083 St. John Hosp., Detroit 4 stool Verotoxin +
70048 Holland Community Hosp.,Holland 56 Verotoxin +
70288 Blodgett Hosp., Grand Rapids 37 stool Verotoxin +
70593 Blodgett Hosp., Grand Rapids stool Verotoxin +

 

aNonmotile, no H determinant.
bNorth Ottowa Community Hospital.

Ggowth pgdla

Two growth media were used in this study. LB (Luria-Bertani)
medium [10g of bacto-trypton (Difco, Detroit, MI), 5g of bacto-yeast

(Difco), and 10g of sodium chloride (EM, Gibbson, NJ) per liter] and TST

20
(trypticase soy tryptose) broth [15g of trpticase soy (BBL dehyrated,
Cockeysville, MD), 13g of tryptose (Difco dehydrated), and 3g of yeast
extract per liter]. LB broth was adjusted to pH:7.5 with sodium
hydroxide. TST broth was adjusted to pH:7.2. LB agar was of the same
formulation except 15g of agar (Difco) was added to one liter of broth.
TST agar included those components listed plus 15g of agar.

S.0.C. medium (2% bacto-tryptone, 0.5% yeast extract, 10mM NaCl,
2.5mM KC1, 10mM MgClz, 10mM MgSO4, 20mM glucose) was prepared as follows
: 2g of bacto-tryptone, 0.5g yeast extract, lml 1M NaCl, and 0.25ml 1M
KCl were added to 97ml of distilled water, allowed to dissolve and then
autoclaved. The medium was cooled to room temperature, then 1 m1 2M
Mg2+ stock (1M MgC12 + 1M MgSO4, 0.45 micron polysulfone filter

steriled) and 1 ml 2M glucose (filter steriled) were added.

Balm;

Equipment used included an Eppendorf centrifuge 5415 (Brinkmann,
Westbury, NY), Vortex (Scientific Industries, Boehmia,NY), 65C water
bath (Precision, Chicago, IL), 37C shaker bath 3540 (Lab-Line, Melrose
Park, IL), 302 nm UV transilluminator (Fisher), Centrifuge IEC B-20A
(Damon/IEC Division, Needham Heights, MA), IEC 870 rotor, Eppendorf tips
Eppendorf tubes (Cole-Parmer, Chicago, IL), Eppendorf pipetters
(Brinkmann), Oakridge tubes (Nalgene, Rochester, NY), Mayer nitrogen
evaporator (Organomation, South Berlin, MA), combs, gel decks,
electrophoresis systems (H5, H6, BRL), power supply (LKB 2002,

Gaithersburg, MD), pH meter (Markson 90, Taiwan, ROG), Elutip column,

21

N0. 2 red filter.

Restrigtion endonuclease

Six restriction enzymes were used (Table 6).

0.45 micron cellulose acetate Elutip prefilter (Schleiche & Schuell,
Keene, NH), syringes, 18G 1 1/2 needles (Becton Dickinson, Rutherford,
NJ), 0.45 micron polysulfone filters (Gleman, Ann Arbor, MI), polaroid

film type 47 (Fabrigue aux EU par Polaroid Corp., Cambridge, MA), and

Table 6. Restriction enzymes used to digest plasmid DNA

 

Enzyme | Origin |Source|Activity| Recognition site

Hind 3 Hemophilus influenzae Rd BRL

lOU/ul# 5'A* AGCTT3'

 

 

 

 

 

TT GA*HOA

Bgl 1 Bacillus globigii BRL lOU/ul S'CCCNNNN* NGGC3'
CGGN*HQNNNNCGC

Pst 1 Erovidencia atuarti IBI l6U/ul 5'CTGCA* G3'
G*HQAC£TC

Ava l Anabaena variabilis IBI 10U/ul 5'C*pPyCGPuG3'
GPuGCPy*HOC

BamH l Eacillus amyloliguifaciens H IBI 24U/ul 5'G* GATCC3'
CCEAG*HOG

EcoR 1 Escherichia coli RY 13 IBI 10U/u1 5'C* AATTC3'
CTEAA*HOG

 

#One unit of restriction enzyme will digest one microgram of an
apropriate DNA in one hour at the appropriate temperature.

*Restriction site.

22

Eaga apa Liu rapid plasmid isolation procedure

Bacterial colonies were inoculated to 6ml of TSTB and shaken at
220rpm in 370 waterbath overnight. One ml of this culture was
transferred to a 1.5ml Eppendorf tube and cells were pelleted by
centrifugation at 7000rpm (RCF:4000) for 2 minutes in an Eppendorf
microcentrifuge. Pelleted cells were suspended using a Vortex in 100 ul
lysing solution (3% SDS in SOmM Tris, pH:12.6). The lysate was
subsequently incubated at 65C for 1 hour followed by extraction with an
equal volume of phenol-chloroform (1:1; chloroform contained 1/24 volume
of isoamyl alcohol). After centrifugation for 5 min at 14000rpm
(RCF:l6000), the aqueous phase was mixed with loading buffer (25% ficol,
5% SDS, 0.025% bromphenol blue, and 25mM EDTA) then loaded to agarose
gels for electrophoresis or further purified. In the latter case, the
phenol-chloroform extracted lysate was extracted twice with diethyl
ether to remove traces of phenol. The ether was removed using an
Eppendorf pipette and remaining traces of ether were evaporated under
nitrogen. After adding 3M sodium acetate, plasmid DNA was precipitated
by adding 800u1 cold (-20C) 95% ethanol ([NaOAc]-0.3M)and placed in a
-7OC freezer for 5 minutes. Precipitated plasmid DNA was immediately
pelleted by centrifugation at 14000rpm at 4C for 15 minutes. The
supernatant was discarded and the pellet was dried under nitrogen and
resuspended in 100 ul TE buffer (10mM Tris-Cl, lmM EDTA, pH:8.0).
Plasmid DNA was analyzed by agarose electrophoresis and by restriction
endonuclease digestion. A flow diagram of the K-L plasmid isolation

procedure is illustrated in Figure l.

23

Isolation of plasmid.DNA

Cell growth
(Overnight incubation of 6 m1 inoculated TSTB)

W
Spin down

(lml cell at RCF: 4000 for 2 min)

Cell lysis and alkaline denaturation
(lOOul 3% SDS in SOmM Tris, pH:12.6)
(Incubated at 650 for 60 min)

||
H
W

Removal of cell debris and denatured DNA
(Phenol-chloroform extraction, centifugation at RCF:16000 for 5 min)

II
II
W

Removal of phenol
(Repeated extraction with ether)

II
II
W

Neutralization
(100u1 3M NaOAc,pH:5.2)

Precipitation of DNA
(800u1 ethanol, -70C 5min, RCle6000 for 15min)

W
Resuspension of DNA

(Dry the pellet, resuspend in lOOul TE buffer)

Figure 1. Flow diagram of plasmid DNA isolation by Kado-Liu procedure

24

Egagaag gal gleggrophogesla

Agarose gels were prepared by suspending 150mg agarose in 25ml TBE
buffer (89mM Tris, 89mM boric acid, 2mM EDTA, pH:8.0) for a 0.6% small
(8 wells) gel (5X7 cm) or 450mg agarose in 75ml TBE for a large (14
wells) gel (11X14 cm). Agarose was melted either in a microwave or on a
hot plate, mixed well and cooled to 50C before pouring. The cooled
agarose was poured in a taped gel deck and allowed to stand 30-60
minutes for polymerization. DNA (20u1) was mixed with loading buffer (1
part loading buffer : 4 parts sample). All electrophoresis was
performed on a horizontal apparatus (BRL Model H5, H6) and carried out
at 5V/cm gel [61] to resolve DNA fragments. Higher voltages were
occasionally used for screening. Gels were stained in 3ug/ml ethidium
bromide and DNA bands were visualized over a 302nm UV transilluminator.
Polaroid type 47 film exposed through Toshiba N0. 2 red filter was used
to photograph gels. A molecular weight standard (Hind 3 digested lambda
phage) was resolved identically and consisted of eight DNA bands: 15.27
Md, 6.22 Md, 4.33 Md, 2.88 Md, 1.53 Md, 1.34Md, 0.37 Md, and 0.083 Md.
Molecular weight estimates of unknowns were made by plotting migration
distances of the standard on the abscissa and log of base pair number on

the ordinate.

i he dia a li ic tion f lasmid

E. gall 0157:H7 was inoculated to LB broth or TST broth with and

without chloramphenicol to determine the extent of plasmid amplification

25
and the effect of growth in media on plasmid concentration. Plasmid
amplification [59,79] was performed as follows: 0.6ml overnight E. gall
grown LB broth was added to 6m1 LB broth and shaken at 225 rpm in 37C
waterbath for 3 hours (late log phase). Next, 0.3 ml of late log phase
culture was added to 6ml of LB broth, incubated 2.5 hours (log phase) at
37C with shaking. Chloramphenicol (34mg/ml in ethanol) was added to the
broth at the final concentration of l70ug/ml to stop protein synthesis
within the cells. The culture was then incubated in a 37C waterbath

shaker for overnight.

Eatraction and concgntration of pBR322

Large scale extraction of pBR322 was performed to determine the
effectiveness of different extraction steps and precipitation
procedures. Procedures used were similar to those used in the Kado-Liu
procedure. pBR322 transformed E. gall HBlOl cells were inoculated to
TST broth. After amplification, 80ml was pelleted at 6000rpm (RCF:4000)
for 10 minutes using an IEC 870 rotor. Cells were suspended and 10ml of
lysing solution was added. DNA samples from the aqueous layer were
taken after the first and second phenol-chloroform extraction {15
minutes at 12000rpm (RCF: 16000)}, and subsequently extracted twice by
ether or chloroform. Chloroform or ether extracted lysates (300 ul)
were concentrated by butanol [62] to 75ul or precipitated by ethanol
[62] or isopropanol [62], with and without a 70% ethanol wash, then
resuspended in 75u1 TE buffer. Ethanol precipitation was performed as

follows: (1) 100u1 of 3M NaOAc was added to 300ul of chloroform or ether

26
extracted lysate. (2) 800ul of 95% ethanol was added and the lysate was
chilled at -7OC for 5 minutes. (3) DNA was pelleted by centrifugation
(15 min, 14000rpm). (4) pellet was resuspended in 75u1 TE buffer.
Isopropanol precipitation was performed as above except 60ul 3M NaOAc
and 360ul isopropanol was added instead of 100ul NaOAc and 800ul
ethanol. Precipitated pellets were washed with 70% ethanol A 70%
ethanol wash was accomplished by adding 800ul of 70% ethanol. The
lysate was centrifuged for 15 minutes at 14000rpm. Butanol
concentration was performed by repeatedly extracting the chloroform or
ether extracted lysate with an equal volume of n-butanol until the final
volume of the aqueous phase was 75u1. Precipitated and concentrated
samples (300u1--> 75u1) were diluted (1:4) to obtain the same
concentration as previous samples before being applied to agarose gel

for electrophoresis.

Igalagion of DEA fram agarose gel

The procedures for isolation of DNA from agarose gels were those
described by the American Type Culture Collection [1]. The lysate,
after being extracted twice by ether was precipitated by adding one
third volume of 3M sodium acetate and 2 volumes of ethanol to provide a
final concentration of 0.25M sodium acetate [62]. The mixture was
chilled to -70C for 15 minutes and centrifuged at 12000rpm (RCF:l6000)
for 30 minutes [106]. The pellet was dried under nitrogen and
resuspended in 0.75m1 of TE buffer. Loading buffer (150ul) was added to

the preparation. A preparative gel using 900u1 in a sample trough

rather
After 5
interes
needle
low sa
added

DNA we
and ge
Eluti;
(manu:
colum:
Tris-
salt

to e1
colun
elute

addir

27
rather than wells was used to resolve a large quantity of plasmid.
After staining the gel with ethidium bromide, the plasmid bands of
interest were cut out under UV light and forced through a 18 gauge
needle (with a syringe plunger) into a 50ml Oakridge tube. Ten ml of
low salt elution buffer (0.2M NaCl, 20mM Tris-HCl, lmM EDTA, pH:7.4) was
added to the tube. The tube was vortexed and shaken overnight at 37C.
DNA was eluted into the buffer during overnight incubation. The buffer
and gel pieces were transferred to a syringe and forced through an
Elutip prefilter and Elutip column. The column was prepared as follows
(manufacturer's procedure): (1) the tip of column was cut off; (2) the
column was prewashed using 2ml high salt elution buffer (1M NaCl, 20mM
Tris-HCl, lmM EDTA, pH:7.4); and (3) the column was primed using 5m1 low
salt elution buffer. The syringe with DNA and gel pieces was connected
to elutip prefilter and column and the DNA was forced slowly through the
column. The syringe and prefilter were then disconnected and DNA was
eluted using 0.4ml high salt elution buffer and concentrated after

adding 2 volumes of ethanol.

ti end ea e na is

Restriction enzyme digestion was performed using a total volume of
24ul of sample. The DNA preparation (19.6ul) was removed to a new
Eppendorf tube. The appropriate 10X buffer (2.4u1) and the desired
restriction enzyme (2ul) were added to the tube. The tube was mixed
throughly by tapping and then incubated at 37C for 2 hours. The DNA
fragments in the restriction digest were analyzed by gel electrophoresis

[34,62]. 27

28

RESULTS

fizowth pgida and plasmid aaplification

Two strains of E. coli 0157:H7 were grown in LB broth, TST broth,

 

and LB broth with chloramphenicol for amplification of plasmids and
screened for plasmid content by the Kado-Liu plasmid isolation
procedure. Cells grown in TST broth demonstrated the largest cell
pellet and those which had undergone the plasmid amplification procedure
demonstrated the smallest cell pellet after centrifugation. Plasmid
patterns are shown in Figure 2. Cells grown in TST broth had the
highest concentration of the 72 Md large plasmid in both strains (lane 4
and 7). Amplification of plasmid DNA by chloramphenicol was only

effective for the 3 Md plasmid (lane 2).

Elaaalds fgom various isolatiap atepa

Large scale preparation followed by microscale extraction and
concentration on pBR322 transformed E. gall HBlOl cells was performed to
illustrate the effectiveness of different DNA extraction steps (Figure
3). DNA bands from phenol-chloroform extracted lysate (lane 1 and 2)
were clearly shown although the intensity of staining was less than

after the precipitation procedures. The concentration of DNA was

29
increased after extracting twice with chloroform (lane 4) whereas ether
extraction (lane 3) as well as butanol concentration (lane 9,10) did not
increase the concentration of DNA. Procedures using ethanol or isopro-
panol precipitation enhance the brightness of the bands, but they also
demonstrated an extra band above the original band and picked up more
chromosomal material and small molecular weight RNA (lane 5,6,8,9).
There was no difference when the use of ethanol or isopropanol for pre-
cipitating DNA (5 vs 6,8 vs 9) was compared. Washing with 70% ethanol

did tighten the DNA bands, but the effect was not great (lane 11-14).

MW Standard:
Band 1: 15.27Md
Band 2: 6.21Md
Band 3: 4.33Md
Band 4: 2.88Md
Band 5: 1.53Md

Band 6: 1.34Md

 

Figure 2. Plasmid patterns of two E. coli 0157:H7 strains in

different growth media and amplification procedure.
Lane 1,8: molecular weight standard; lane 2: strain 61409 after plasmid
amplification; lane 3: strain 61409 grown in LB broth; lane 4:5train
61409 grown in TST broth; lane 5: 61427 after plasmid amplification;
lane 6: 61427 grown in LB broth; lane 7: 61427 grown in TST broth.

 

30

'1 2 3 4 5 6 7 8 9 101 121314

Figure 3. pBR322 DNA from varying extraction and concentration steps.

Lane
Lane
Lane
Lane
Lane
Lane
Lane
Lane
Lane

LanelO:
Lanell:
Lane12:
Lanel3:
Lane14:

\OQVONUIJ-‘WNH

P-C

"U'U’U'U'U'IU'IU’U'U'U'U'U
000000000000

: phenol-chloroform (P-C) extracted once
extracted twice

extracted twice, ether (E) extracted twice
extracted twice, chloroform (C) extracted twice

twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,

E

E
E
C
C
C
E
E
C
C

twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,
twice,

ethanol precipitated

isopropanol precipitated

butanol concentrated

ethanol precipitated

isopropanol precipitated

butanol concentrated

ethanol precipitated, 70% ethanol washed
isopropanol precipitated, 70% ethanol washed
ethanol precipitated, 70% ethanol washed
isopropanol precipitated, 70% ethanol washed

31

Elaamld pattegps of E, coli 015Z;H7

Seventeen isolates of E. gall were screened by the Kado-Liu
plasmid isolation procedure (Figure 4). Three plasmid patterns were
determined: (1) Strains with only 72 Md plasmid [51560, 61476, and 70288
(A4,10,B10)]. Strain 61494 (A 11) had a 64.9 Md band in addition to 72
Md band. (2) Five strains contained a 72 and 1.47 Md plasmid [51396,
61385, 61427, 61676, and 62083 (A2,7,9,B5,8)]. Strain 70048 (B9)
contained the same pattern as those in group 2 and additional 10.9 and
3.31 Md bands. (3) Six strains contained 72, 2.90, and 2.72 Md bands
[6UNCL-3, 61167, 61409, 61951, 62017, and 70593 (A5,6,8,B6,7,ll)].
Strain 61582 (B4) had an additional 165 Md band. All strains of E. gall
0157:H7 possessed the 72 Md large plasmid. Strain 50681 (A2), a rough
strain (serogroup untypable), possessed a large 77.5 Md plasmid band

and three other bands (10.3, 3.38, and 1.38 Md).

Eestrictlon profiles of E, goll Ql§Z;EZ

Restriction endonuclease digestion with Bgl 1 was performed on
plasmid DNA of all seventeen strains. Strains with similar plasmid
patterns were grouped for comparison (Figure 5 and Table 7). A common
restriction profile pattern was found throughout the seventeen 0157:H7
strains and within plasmid pattern groups. Plasmid pattern group 2
contained a 1.42 Md band (A:6-9;B:3-5), a 2.78 Md and a 2.31 Md (B:3-5)
band. Plasmid pattern group 3 contained a 5.36 Md (A:lO-14;B:6-8), a
3.24 Md (A:ll,12:B:6,8) and a 2.48 Md (B:6-8) band. In addition,

a number of strains contained a variety of unique bands.

32

 

sat-sagaata‘as'aaaa

 

MW Standard:
Band 1: 15.27Md Band 2: 6.21Md Band 3: 4.33Md
Band 4: 2.88Md Band 5: 1.53Md Band 6: 1.34Md

Figure 4. Plasmid pattern analysis of E. coli 0157:H7.
A: Lane 1,12:molecular weight standard, lane 2:50681, lane 3251396 lane
4:51560, lane 5:6UNCL-3, lane 6:61167, lane 7261385, lane 8261409, lane
9261427, lane 10:61476, lane 11261494.
B: Lane 1,122molecu1ar weight standard, lane 2:61476, lane 3:61494 lane
4:61582, lane 5261676, lane 6:61951, lane 7262017, lane 8:62083, lane
9:70048, lane 10:70288, lane 11270593

MW Standard:

  

Band 1: 15.27Md
Band 2: 6.21Md
Band 3: 4.33Md
Band 4: 2.88Md
Band 5: 1.53Md

Band 6: 1.34Md

Figure 5. Bgl I restriction profiles of E. coli 0157:H7
A: Lane lzmolecular weight standard, lane 2261494, lane 3:61560, lane
4:61476, lane 5270288, lane 6:51396, lane 7:61427, lane 8:62083, lane
9:70048, lane 10:6UNCL-3, lane 11:61409, lane 12:61951, lane 13:61582,
lane 14:61167.
B: Lane 1:molecular weight standard, lane 2:61476, lane 3:62083, lane
4261385, lane 5:61676, lane 6:61951, lane 7:62017, lane 8270593

 

34

Table 7. Bgl restriction.profiles of E. coli 0157:H7

 

 

 

Lane # Strains Molecular weights of digested fragments (Md)
A: 1 Std 15.27 6.21 4.33 2.88 1.53 1.34
2 61494 11.6 7.55 6.95* 6.23* 5.83 4.04* 3.71* 3.31 3.05*
2.65 2.58* 2. 25 2.18 2.12 1.85 1.56
3 51560 11.6 7.55 6. 23 5. 83 3.31 2.65 2.25 2.12 1.85
4 61476 11.6 7.55 6. 23 5. 83 3.31 2.65 2.25 2.12 1.85
5 70288 11.6 7.55 6. 23 5. 83 3.31 2.65 2.25 2.12 1.85
6 51396 11.6 7.55 6. 23 5. 83 3.31 2.65 2.25* 2.12 1.85 1.42*
7 61427 11.6 7.55 6. 23 5. 83 3.31 2.65 2.25* 2.12 1.85 1.42*
8 62083 11.6 (8 34) 7. 55 6. 23 5.83 3.31 2.65 2.25 2.12 1.85
1. 42
9 70048 11. 6 7.55 6.23 5.83 4.77 3.31 2.65 2.38 2.25 2.12
1.85 1.42
10 6UNCL-3 11.6 7.55 6.23 5.83 5.36 3.31 2.65 2.25 2.12 1.85
11 61409 11.6 (10.2) 7.55 6.23 5.83 5.36 (3.9) 3.31 3.24 2.65*
2.25* 2.12 1.85
12 61951 11.6 7.55 6.23 5.83 5.36 3.31 3.24 2.65* 2.25* 2.12
1.85
13 61582 11.6 (8.34) 7.55 6.23 5.83 5.36 4.2 3.77 3.31 2.91
2.65 2.25 2.12
14 61167 11.6 9.93 7.55 6.23 5.83 5.36 3.31 2.65 2.25
B: 1 Std 15. 27 6. 21 4. 33 2. 88 1. 53 1.34
2 61476 11. 6 7. 55 6. 23 5. 83 3. 31 2. 65 2. 25 2.12 1. 85
3 62083 11. 6 7. 94 7. 55 6. 23 5. 83 4. 24 (4.04) 3. 64 3. 31 2. 78
2. 65 2. 31 2. 25 2.12 1. 85 1. 42
4 61385 11.6 7. 55 6. 23 5. 83 3.31 2.78 2.65 2.31 2.25 2.12
1. 85 1. 42
5 61676 11.6 7. 55 6. 23 5.83 3.31 2.78 2.65 2.31 2.25 2.12
1. 85 1. 42
6 61951 11.6 9. 6 7. 55 6. 23 5. 83 5. 36 3. 31 2. 65* 2, 48 2. 25
2.12 L 85
7 62017 11.6 7.55 6.23 5.83 5.36 3.31 2.65* 2.48 2.25 2.12
1. 85
8 70593 11.6 7.55 6.23 5.83 5.36 3.31 3.24 2.65* 2.48 2.25
2.12 1.85
( ): Vague bands
Bold : bands of common pattern

*Brighter bands

35

 

MW Standard:
Band 1: 15.27Md Band 2: 6.21Md Band 3: 4.33Md
Band 4: 2.88Md Band 5: 1.53Md Band 6: 1.34Md

Figure 6. Restriction profile comparisons of representative strains from
each plasmid group and isolated small plasmids.

Lane 1: strain 61427, lane 2: Bgl 1 restriction of small plasmid of

61427, lane 3: Bgl 1 restriction on 61427, lane 4: Bgl l restriction on

51560, lane 5: Bgl I restriction on 61951, lane 6: Bgl I restriction on

small plasmids of 61951, lane 7: strain 61951, lane 8: molecular weight

standard.

36

es c o o 1 co at on

Further restriction endonuclease digestion was performed to
determine if the additional bands found in group 2 and 3 were
contributed by the small plasmids or the 72 Md large plasmid. Three
strains (61427, 51560, 61951) from different plasmid pattern group and
isolated small plasmids of 61427 (1.47 Md) and 61951 (2.9, 2.73 Md) were
digested with Bgl l to compare their restriction profiles (Figure 6 and
Table 8). Additional bands in the restriction profiles of 61427 and
61951 (when compared to that of 51560 ) were found in undigested 61427
(lane 1) and 61951 (lane 7) and restriction enzyme digested isolated

small plasmids from 61427 (lane 2) and 61951 (lane 6).

Table 8 Restriction profile comparisons of representative strains from
each plasmid group and isolated small plasmids.

 

Lane# Strains Bgl 1 Molecular weights of DNA bands (Md)

 

1 61427 No 72 1.42

2 61427* Yes 2.25 1.42

3 61427 Yes 11.2 7.60 6.66 6.08 3.14 2.38 2.25 2.08 1.92
1.72 1.42

4 51560 Yes 11.2 7.60 6.66 6.08 3.14 2.38 2.08 1.92 1.72

5 61951 Yes 11.2 7.60 6.66 6.08 3.14 2.97 2.54 2.38 2.08
1.92 1.72

6 61951* Yes 5.94 4.36 2.54

7 61951 No 72 6.66 2.97 2.54 2.45

8 Std No 15.27 6.21 4.33 2.88 1.53 1.34

 

*Isolated small plasmids Bold to indicate the differences from 51560

37

Restriction profiles of 22 Md plasmid

Because similar plasmid and restriction patterns were common among

all seventeen E. coli 0157:H7 isolates, strains with only a 72Md plasmid

 

were examined using a variety of restriction endonucleases to determine
restriction profiles of the common 72 Md large plasmid. The restriction
profiles of 51560 (A) and 61476 (B) after Hind 3, Bgl 1, Pst 1, Ava 1,

BamH l, and EcoR 1 are shown (Table 9 and Figure 7).

Table 9. Restriction profiles of strain 51560

 

Lane# Enzmes Molecular weights of digested fragments (Md) MW Summation

 

1 - (Std) 15.27 6.21 4.33 2.88 1.53 1.34 31.58
2 -(51560) 72 72
3 Hind 3 17.2 14.2 8.25 7.26 4.75 4.22 3.96 3.70 3.23

3.04 2.21 1.35 73.37
4 Bgl 1 11.2 7.59 6.47 6.21 3.23 2.64 2.18 2.08 1.91

1.84 46.75
5 Pst l 9.57 7.13 4.36 3.83 3.63 3.17 3.04 1.58 1.54

1.42 1.31 40.58
6 Ava 1 7.00 6.40 5.94 5.68 5.28 4.16 2.88 2.57 2.18

1.75 1.52 45.36
7 BamH 1 17.2 15.2 9.24 2.84 2.61 2.01 1.29 50.39
8 EcoR 1 19.1 12.5 9.24 8.58 3.76 2.38 2.21 1.98 1.62 61.37

 

38

 

MW Standard:
Band 1: 15.27Md Band 2: 6.21Md Band 3: 4.33Md
Band 4: 2.88Md Band 5: 1.53Md Band 6: 1.34Md

Figure 7. Restriction profiles of 51560 and 61476 after restriction
endonuclease digestions.
A: 51560 B: 61476; lane 1: molecular weight standard, lane 2: strain
51560 or 61476, lane 3: Hind 3 digestion, lane 4: Bgl 1 digestion, lane
5: Pst 1 digestion, lane 6: Ava 1 digestion, lane 7: BamH 1 digestion,
lane 8: EcoR 1 digestion.

DISCUSSION

Hemorrhagic colitis is a newly recognized syndrome characterized
by severe crampy abdominal pains, watery diarrhea followed by grossly

bloody diarrhea and little or no fever [87]. Escherichia coli 0157:H7

 

was first recognized as a pathogen in the United States when the
organism was incriminated as the cause of two geographically separate
outbreaks (Michigan and Oregon) of hemorrhagic colitis, both associated
with undercooked beef from one fast food chain [87].

E. goli 0157:H7 does not produce a classic heat-labile or heat
stable enterotoxin. It is noninvasive and is not a known
enteropathogenic serotype [104]. It has been found to produce a Vero
cell toxin [42,46,72,78,86] and its cytopathic effect (CPE) can be
neutralized with Shiga antitoxin [72]. Although stool filtrates may
display Vero cell toxin activity and may be of diagnostic value [78],
Vero cell toxin production is not specific for the 0157:H7 serotype
[45]. Isolation of E. 9211 0157:H7 is best accomplished when the stool
is cultured within four days of the onset of symptoms [78,84,86].
Bacterial shedding may be more prolonged in children [78].

Although E. 991; 0157:H7 can be identified serologically,

serotyping all E. coli isolates to detect E. coli 0157:H7 is

 

impractical. Morphologically and biochemically, E. coli 0157:H7 has

been indistinguishable from other E. coli serotypes except for the

39

D]
it
01

0f

40
absence of sorbitol fermentation after 24 hour incubation [104]. Doyle
[25] and Harris [37] had used this sorbitol non-fermenting
characteristic as a marker for the population containing 0157:H7
serotype. The inability to ferment sorbitol in MacConkey-based agar has

been used previously as a screening tool for enteropathogenic E. coli

 

serotype 0111 and 055 [82].

The molecular detail and diversity that can be seen in plasmids
from wild strains of bacteria make them discriminating epidemiological
markers. Plasmid pattern analysis is a technique that is rapid,
relative simple to perform, as specific as phage typing, and superior to
biotyping and resistance typing [11]. Plasmid pattern analysis followed
by restriction endonuclease digestion has been used to identify many
epidemic bacterial strains. The use of new and simple molecular genetic
methods offers the potential to overcome the drawbacks of identifying E.

coli 0157:H7 by methods based on phenotypic characteristics.

Culture conditions It has been reported that 37C is an optimal
temperature for the growth of E. 991; 0157:H7 [25]. The use of LB broth
with choramphenicol to stop protein synthesis is a technique extensively
used for amplifying plasmid DNA. Because TST broth is routinely used in
Diagnostic Microbiology Section, Michigan Department of Public Health LB
and TST broths were compared. After the same period of incubation, TST
produced more cells and a higher plasmid concentration. These results
indicated TST broth, in our hands, was more suitable for growing E. £91;
0157:H7 and for isolating their plasmids. The successful amplification

of a 3Md plasmid in these experiments revealed that this plasmid was

9!

Pl
th

re

41
under relaxed control, while the large 72Md plasmid and the 1.47Md
plasmid, on the other hand, were under stringent control. The
replication of the latter two are coupled to that of the host so that
only one or at most a few copies of each plasmid was presented in each
bacterial cell. These facts make evident the importance of high cell
number in order to obtain suitable plasmid yields. TST broth without
choramphenicol was routinely used because it provided the highest

concentration of large plasmid DNA.

Plasmid isolation.methods Despite the fact that the plasmid
concentration was lower than after precipitation procedures, phenol-
chloroform extracted lysate proved to be sufficient for plasmid
detection. Moreover, the direct use of phenol-chloroform extracted
lysate precluded precipitating and further cleaning the sample,
minimized the manipulation of DNA, and provided a clean sample without
chromosomal debris or RNA.

Although phenol-chloroform extracted lysate was useful for
demonstration of plasmid patterns, it could not be used for further
restriction endonuclease digestion and could not be stored because
traces of phenol and SDS remained in the lysate. Further cleaning was
necessary to obtain a sample that could be used for restriction
endonuclease digestion or for storage.

A number of variables for concentration and/or precipitation of
plasmid DNA were investigated. Chloroform or ether can be used to remove
the traces of phenol, but chloroform extraction suffered from a

resulting lower volume of lysate and more extensive centrifugation was

42
needed to separate aqueous and organic phases. Although the
concentration of DNA after ether extraction was lower than after
chloroform extraction, the preparation after the precipitation procedure
had the same concentration of DNA. Therefore, ether extraction was the
method of choice.

Butanol concentration was unsuitable because multiple extractions
were required and additional ether extractions were needed to remove
butanol. Furthermore, precipitation procedures were required to remove
high concentrations of salt from the preparation.

Ethanol and isopropanol precipitation were equvilent in the
recovery of DNA. Ethanol precipitation rather than butanol concentration
or isopropanol precipitation was adopted because it was easier to remove
after precipitation and the use of ethanol was more economical.

The method of Kado and Liu was modified somewhat for our use. The
organic solvent for extraction of protein was two volumes instead of one
volume because we noted better recovery of DNA. Sodium acetate was
added in one third the volume of lysate, ethanol was added in two
volumes of the lysate plus sodium acetate.

Preparative gels for the purpose of separating and then collecting
high concentration of plasmid DNA were not sucessful. Because of the
extensive mechanical manipulation of DNA, the recovery of small plasmid

from gels was low and the 72Md large plasmid could not be isolated.

Plasmid pattern ALL E. coli 0157:H7 strains examined possess a
72Md large plasmids. Although all strains contained this large plasmid

,other plasmids enabled us to place isolates in one of the three groups.

43

When grouped according to geographic area, strains 51396, 61167, 61494,
61676, 70288, and 70593 were from Grand Rapids and strains 51560, 61385,
61427, 61476, 61582, 61951, and 62017 were from Grand Haven. There was
no correlation between plasmid patterns and geographic areas. Although
all seventeen strains contained the 72Md large plasmid, strains from

different areas had different plasmid patterns. Therfore, the isolates
from sporadic cases in Western Michigan are related in that all contain

a large plasmid but these isolates are not identical.

Restriction profiles A common restriction profile with nine bands
containing 11.6, 7.55, 6.23, 5.83, 3.31, 2.65, 2.25, 2.12 and 1.85 Md

was found throughout the seventeen E. coli 0157:H7 strains (Figure 5,

 

Table 7). This common restriction profile resulted from the restriction
digestion products of the common large plasmid. Apparent small
differences in molecular weight are due to inaccuracies in migration
rates in different gel. The 1.42Md band found in strains 51396, 61385,
61427, 62083,and 70048 corresponded to the 1.47Md band of the plasmid
pattern group 2. The 2.65 Md and 2.25Md bands corresponded to the
2.90Md and 2.72Md bands of plasmid pattern group 3.

Restriction profiles of three reprentative strains of each plasmid
group and isolated small plasmids were compared (Figure 6). The extra
2.25 and 1.41Md bands in 61427 (lane 3) corresponded to the bands of
digested small plasmid of 61427. The 5.36 and 3.24Md of group 3 (Figure
5) corresponded to 5.94Md (lane 6) and 3.14Md (lane 7) respectively.
The extra 5.36Md and 3.24Md bands were a form other than CCC DNA and

were proven to be the same as the 2.90Md and 2.72Md Md CCC bands.

44

The extra 2.78 and 2.31Md bands of group 2 and 2.48Md band of
group 3 (Figure 5B) only indicate better resolution of restriction
profiles from enlarged photographs of the small gel. Comparing strain
62083 in Figur 6A and 6B provide the evidence for this conclusion.

Because of difficulties in purifying the 72Md plasmid and the
existence of common plasmid patterns and restriction profiles, the
strains with only a 72Md plasmid were used for further restriction
endonuclease analysis. The resolved fragment molecular weight summation
from Hind 3 is 73.37Md which was essentially identical to the estimated
molecular weight 72Md reported by others [104]. Summations after
digestion using other enzymes were lower than the estimated molecular
weight. Reasons for this include the possibility that the concentration
of additional fragments was too low to be visualized, masking of small
molecular weight fragments by RNA, or small fragments ran off the gels.
Variation in determining the molecular weight of large plasmid can be
large. The molecular weight of the large plasmid from E. 92;; 0157:H7
has been reported as 60Md by Karch et al. [44] and Levine et a1. [57],
65Md by Johnson et al. [42], and 72Md by Wells et al. [104]. The
adoption of 72Md as the molecular weight of the large plasmid is based
upon results of summation from restricted fragments and the use of large
molecular weight standard (62Md) in the study by Wells et al [104].

This study demonstrated that plasmid pattern analysis is a very
useful diagnostic method for identification of E. 92;; 0157:H7. Further
studies include the investigation of large plasmid encoding products and
the development of a DNA probe specific to this plasmid. Although the

plasmid has been reported to be required for expression of a new

45
fimbrial antigen responsible for adhesion to the epithelial cells [44],

in vivo studies of fimbria are necesssary to confirm that E. coli

 

strains colonize a specific site in the bowel. It would be of interest
to determine whether the plasmids in E. 92;; 0157:H7 actually carry
fimbrial structure genes, reulatory genes that act in concert with
chromosomal genes for fimbrial expression, or both. A DNA probe using
P32 labelled 2.25Md Hind 3 digested fragment has been reported by Levine
et al. [57]. However, the development of non-radioactive probes which
are more suitable for use in clinical laboratories have not been
reported.

Data collection of hemorrhagic colitis is dependent on voluntary
reporting and the incidence of this disease may be underestimated.
Seventeen E. 99;; isolates in this study are from a limited number of
geographic locations in Western Michigan. E. £91; 0157:H7 might be
missed during isolation of diarrheagenic pathogens in many laboratories.
It is important to heighten the awareness of physicians and laboratory

staffs to the existence of E. coli 0157:H7 and to understaning of

diagnostic procedures for its identification.

APPENDICES

46

Appendix I: de t c tio f 011 015 'H

After biochemical identification, E. coli strains were

 

serologically typed using 0157 antiserum and H7 antiserum. 0 antigen
suspension was prepared by inoculating a tube of TSTB, incubating at 37C
for 5 hours, heating in boiling water for 1 hour, then diluting to the
density of McFarland NO. 1 standard with formalinized saline.
Serological testing utilized a final dilution of 1:500 0157 antiserum
(0.02m1 1:10 0157 antiserum plus 1.0ml diluted antigen) in 48C waterbath
overnight. Positive reactions were further confirmed by titration of
diluted antigen with 0157 antiserum. Antigen suspension and 2 fold
serial dilution antiserum at the titers of 1:320, 1:640, 1:1280, and
1:2560 were reacted at 480 overnight. A positive reaction was indicated
by precipitation of the antigen suspension by 0157 antiserum at a titer
of 1:1280. Because the H antigen of E. 991; are often poorly developed,
all strains were inoculated to motility medium to enhance production of
flagella. Formalinized suspension were then tested using H7 antiserum in

a single tube test in which the final dilution of antiserum was 1:500.

47

Appendix II: Transformation of EI coli HBlOl by pBR322

The procedures used for transformation followed the instruction of
manufacturer (BRL). Ten nanograms of pBR322 was added to 20ul E. 92;;
HBlOl cells in an autoclaved and chilled 1.5 ml Eppendorf tube. The
mixture was incubated on ice for 30 minutes, heat shocked one minute in
a 42C waterbath, and then incubated on ice for another 2 minutes. 80u1
S.O.C. was added and eppendorf tube was shaked at 225 rpm in 37C
waterbath for 1 hour. The cells were plated on LB agar plates
(containing 100ug/ml ampicillin). Transformants were screened by

plasmid detection procedure of Kado and Liu [43].

BIBLIOGRAPHY

BIBLIOGRAPHY

1.Anonymous. Recombinant DNA workshop manual. Rockville: American Type
Culture Collection, 1986:17-9

2.Austin TW, Ray CG. Coxackie virus group B infections and the
hemolytic uremic syndrome. J Infect Dis 1973;127:698-701

3.Baker NM, Mills AE, Rachman I, et a1. Hemolytic-uremic syndrome in
typhoid fever. Br Med J 1974;2z84-7

4.Baldini MM, Kaper JB, Levine MM, Candy DCA, Moon HW. Plasmid-mediated
adhesion in enteropathogenic Escherichia coli. J Pediatr Gastroenterol
Nutr 1983;2z534-8

5.Bettelheim KA, Breadon A, Faiers MC, O'Farrell SM, Shooter RA. The
origin of O-serotype of Escherichia coli in babies after normal
delivery. J Hyg 1974;72:67-78

 

6.Birnboim HC, Doly J. A rapid alkaline extraction procedure for
screening recombinant plasmid DNA. Nucl Acid Res l979;7:1513-23

7.Black RE, Merson MH, Huq I, Alim ARMA, Yunus M. Incidence and
severity of rotarzirus and Escherichia coli diarrhea in rural
Bangladesh. Lancet 198l;l:141-3

8.Black RE, Brown KH, Becker S, Alim ARMA, Huq I. Longitudinal studies
of infectious diseases and physical growth of children in rural
Bangladesh. II. Incidence of diarrhea and association with known
pathogens. Am J Epidemiol 1982;115:315-24

9.Black RE, Brown KH, Becker S. Effects of diarrhea associated with
specific enteropathogens on the growth of children in Bangladesh.
Pediatrics 1984;73:799-805

10.Broda P. Plasmids. San Francisco: W.H. Freeman and Company, 1979:1-22
11.Brunner F, Margadant A, Peduzzi R, Piffaretti JC. The plasmid pattern
as an epidemiological tool for Salmonella typhimurigm epidemics: compare
with the lysotype. J Infect Dis 1983;148:7-11

12.Brunton J, Clare D, Ehrman N, Meier MA. Molecular epidemiology of
antibiotic resistance plasmids of flemophilus species and Neisseria
gggggrhggge. Rev Infect Dis l986;8:525-37

48

49

l3.Buenemann H, Mueller W. Base specific fraction of double stranded
DNA: affinity chromatography on a novel type of absorbent. Nucl Acids
Res l978;5:1059-74

l4.Burgess NRH, McDermott SN, Whiting I. Aerobic bacteria occurring in
the hind gut of cockroach, Blatta orientalis. J Hyg 1973;71:1-7

15.Campos A, Sibley R. Kim Y, et al. Hemolytic uremic syndrome. Kidney
1981;14:23-30

16.Centers for Disease Control. Isolation of Ea goli 0157:H7 from
sporadic cases of hemorrhagic colitis-United States. MMWR 1982;31:580-1

l7.Centers for Disease Control. Outbreak of hemorrhagic colitis-Ottawa,
Canada. MMWR 1983;32:133-4

l8.Centers for Disease Control. Hemolytic-uremic syndrome associated
with Escherichia coli 0157:H7 enteric infections-United States, 1984.
MMWR 1985;34:20-1

l9.Chamovitz BN, Hartstein AL, Alexander SR, et al. Campylobacter jejuni
associated hemolytic uremic syndrome in a mother and daughter.
Pediatrics 1983;71:253-6

20.Clewell DB, Helinski DR. Supercoiled circular DNA-protein complex in
Escherichia goli: purification and induced conversion to an open
circular form. Proc Natl Acad Sci USA 1969;62:1159-66

21.Clewell DB, Helsinki DR. Properties of a supercoiled deoxyribonucleic
acid-protein relaxation complex and strand specificity of relaxation
event. Biochemistry l970;9:4428-40

22.Clowes RC. Molecular structure of bacterial plasmids. Bact Rev
1972;36:361-404

23.Cornelis P, Digneffe C, Willemot K, Colson C. Purification of
Escherichia coli amplifiable plasmids by high-salt sepharose
chromatography. Plasmid l981;5:221-3

24.Cravioto A, Cross RJ, Scotland SM, Rowe B. An adhesive factor found
in strains of Escherichia coli belonging to the traditional infantile
enteropathogenic serotypes. Curr Microbiol 1979;3:95-8

 

25.Doyle MP, Schoeni JL. Survival and growth characteristics of
Eacherichia gal; associated with hemorrhagic colitis. Appl Environ
Microbiol 1985;49:310-5

26.DuPont HL, Formal SB, Hirnick RB, et al. Pathogenesis of Escherichia
gall diarrhea. N Engl J Med 1971;285:1-9

27.DuPont HL, Olarte J, Evans DG, Pickering LK, Galindo E, Evans DJ.

50

Comparative susceptibility of Latin American and United States students
to enteric pathogens. N Engl J Med 1976;295:1520-1

28.Farmer JJ, Davis BR. H7 antiserum-sorbital fermentation medium: a
single tube screening medium for detecting Escherichia coli 0157:H7
associated with hemorrhagic colitis. J Clin Microbiol 1985;22:620-5

 

29.Francis DH, Collins JE, Duimstra JR. Infection of gnotobiotic pigs
with Escherichia coli 0157:H7 strain associated with an outbreak of
hemorrhagic colitis. Infect Immun 1986;51:953—6

 

30.Gillies RR. Bacteriocin typing of Enterobacteriaceae in: Bergan T,
Norries JR, eds. Methods of Microbiology; vol.ll. London: Academic
Press, 1978:78-86

3l.Gorbach SL, Banwell JG, Chatterjee BD, Jacobs B, Sack RB. Acute
undifferentiated human diarrhea in the tropics. I. Alterations in
intestinal microflora. J Clin Invest 1971;50:881-9

32.Gransden WR, Damm MAS, Anderson JD, Carter JE, Lior H. Further
evidence associating hemolytic uremic syndrome with infection by
Verotoxin-producing Escherichia coli 0157:H7. J Infect Dis 1986;154:522-
4

33.Gross RJ, Scotland SM, Rowe B. Enterotoxin testing of Escherichia
gal; cause epidemic infantile enteritis in the U.K. Lancet l976;l:629-31

34.Hackett PB, Fuchs JA, Messing JW. An introduction to recombinant DNA
techniques. Menlo Park: The Benjamin/Cumming Publishing Company, Inc,
1984:109

35.Hale TL, Sansonetti PJ, Schad PA, Austin S, Formal SB.
Characterization of virulence plasmid and plasmid-associated outer
membrane proteins in Shigella flexneri, Shigella sonnei, and Escherichia
coli. Infect Immun 1983;40:340-50

36.Hardy K. Bacterial plasmids. 2nd ed. Washington D.C.: American
society for Microbiology, 198613-22

37.Harris AA, Kaplan RL, Goodman LJ et al. Results from a four-month
stool survey for E_L coli-use of a new screening method. Clin Res
1984;32:761A

38.Harris JR, Wachsmuth IK, Davis BR, Cohen ML. High-molecular-weight
plasmid correlates with Eachagichia gall enteroinvasiveness. Infect
Immun 1982;37:1295-8

39.Helling RB, Goodman HM, Boyer HW. Analysis of R.EcoRI fragments of
DNA from lambdoid bacteriophages and other viruses by agarose
electrophoresis. J Virol 1974;14:1235-44

40.Hockin JC, Lior H. Hemorrhagic colitis due to Eagherichia coli

51
0157:H7. A rare disease? Can Med Assoc J 1986;134:25-6

41.John JF, Twitty JA. Plasmids as epidemiologic markers in nosocomial
Gram-negative bacilli: experience at a university and review of the
literature. Rev Infect Dis 1986;8z693-704

42.Johnson WM, Lior H, Bezanson GS. Cytotoxic Escherichia coli 0157:H7
associated with hemorrhagic colitis in Canada. Lancet l983;1:76

43.Kado CI, Liu ST. Rapid procedure for detection and isolation of large
and small plasmids. J Bacteriol 1981; 145:1365-73

44.Karch H, Heesemann J, Laufs R. O’Brien AD, Tacket CO, Levine MM. A
plasmid of enterohemorrhagic Escherichia coli 0157:H7 is required for

expression of a new fimbrial antigen and for adhesion to epithelial
cells. Infect Immun 1987;55:455-61

45.Karmali MA, Steele BT, Petric M, Lim C. Sporadic cases hemolytic-
uremic symdrome associated with fecal cytotoxin-producing Escherichia
gall in stools. Lancet l983;1:6l9-20

46.Karmali MA, Petric M, Lim C, Fleming PC, Steele BT. Escherichia coli
cytotoxin, hemolytic-uremic syndrome, and hemorrhagic colitis. Lancet
l983;2:1299-300

47.Karmali MA, Petric M, Lim C, Fleming PC, Arbus GS, Lior H. The
association between idiopathic hemolytic uremic syndrome and infection
by verotoxin-producing Escherichia coli. J Infect Dis 1985;151:775-82

48.Kauffmann F. The serology of coli group. J Immunol 1947;57:71-100

49.Klein RD, Selsing E, Wells RD. A rapid microscale technique for
isolation of recombinant plasmid DNA suitable for restriction enzyme
analysis. Plasmid l980;3:88-9l

50.Koster F, Levin J, Walker L, et al. Hemolytic uremic syndrome after
shigellosis. N Engl J Med 1978;298:927-33

51.Krishnann C, Fritzgerald VA, Dakin SJ, Behme RJ. Laboratory
investigation of outbreak of hemorrhagic colitis caused by Escherichia
gall 0157:H7. J Clin Microbiol 1987;25:1043-47

52.Lang D, Mitani M. Simplified guantitative electron microscopy of
biopolymers. Biopolymers 1970;9z373-9

53.Levine MM, Ristaino P, Sack RB, Kaper JB, ¢rskov F, ¢rskov I.
Colonization factor antigen I and II and type I somatic pili in
enterotoxigenic Escherichia coli: relation to enterotoxin type. Infect
Immun 1983;39:889-97

54.Levine MM, Edelman R. Enteropathogenic Escherichia coli of classic
serotypes associated with infant diarrhea: epidemiology and

52
pathogenesis. Epidemiol Rev l984;6:31-51

55.Levine MM, Nataro JP, Karch H, et al. The diarrheal response of
humans to some classic serotypes of enteropathogenic Escherichia coli is
dependent on a plasmid encoding an enteroadhesiveness factor. J Infect
Dis 1985;152:550-9

56.Levine MM. Escherichia coli that cause diarrhea: enterotoxigenic,
enteropathogenic, enteroinvasive, enterohemorrhagic, and enteroadherent.
J Infect Dis 1987;155:377-389

57.Levine MM, Xu JG, Kaper JB, et al. A DNA probe to identify
enterohemorrhagic Escherichia coli of 0157:H7 and other serotype that
cause hemorrhagic colitis and hemolytic uremic syndrome. J Infect Dis

1987;156:175-82

58.Maniatis T, Fritsch EF, Sambrook J. Plasmids in :Molecular cloning-a
laboratory manual. New York: Cold Spring Harbor Laboratories, 1982:3-4

59.Maniatis T, Fritsch EF, Sambrook J. Large scale preparation of
plasmid DNA in: Molecular cloning-a laboratory manual. New York: Cold
Spring Harbor Laboratories, 1982:86-96

60.Maniatis T, Fritsch EF, Sambrook J. Restriction Enzyme in: Molecular
cloning-a laboratory manual. New York: Cold Spring Harbor Laboratories,
1982:98-106

61.Maniatis T, Fritsch EF, Sambrook J. Agarose gel electrophoresis in:
Molecular cloning-a laboratory manual. New York: Cold Spring Harbor
Laboratories, 1982:150-172

62.Maniatis T, Fritsch EF, Sambrook J. Biochemical techniques in:
Molecular cloning-a laboratory manual. New York: Cold Spring Harbor,
1982:436-478

63.March SB, Ratnam S. Sorbital-MacConkey medium for detection of
Eaghegichia cal; 0157:H7 associated with hemorrhagic colitis. J Clin
Microbiol 1986;23:869-72

64.Martin T. Popick DG. Hemorrhagic colitis associated with Escherichia
goli 0157:H7. Can Med Assoc J 1984;131:1329

65.Mathewson JJ, Johnson PC, DuPont HL, et al. A newly recognized cause
of travelers' diarrhea: enteroadherent Escherichia coli. J Infec Dis
1985;151:471-5

66.Mathewson JJ, Johnson PC, DuPont HL, Satterwhite TK, Winsor DK.
Pathogenicity of enteroadherent Esaherichia goli in adult volunteers. J
Infec Dis 1986;154:524-7

67.Maxam AM, Gilbert W. Sequencing end labelled DNA with base-specific
chemical cleavage. Meth Enzymol 1980;65:499-560

53

68.Merson MH, Morris GK, Sack DA, et al. Travellers'diarrhea in Mexico.
A prospective study of physicians and family members attending a
congress. N Engl J Med 1976;294:1299-1305

69.Milch H. Phage typing of Escherichia coli in : Bergan T, Norries JR,
eds. Methods of Microbiology; vol.ll. LondonzAcademic press:1978:88-115

70.Morries JA, Sojka WJ. Eschierichia coli as a pathogen in animals in
Sussman M ed. The virulence of Escherichia coli. Orlando: Academic
Press, 1985:47-77

 

71.Newcast1e Asymptomatic Bacteriuria Research Group. Asymptomatic
bacteriuria in school-children in Newcastle upon Tyne. Arch Dis Child
1975;50:90-102

72.0'Brien AD, Lively TA, Chang TW, Gorbach SL. Purification of Shigella
aysenteriae 1 (Shiga)-1ike toxin from Escherichia coli 0157:H7 strain
associated with hemorrhagic colitis. Lancet 1983;2z573

73.0'8rien AD, Newland JW, Miller SF, Holmes RK, Smith HW, Formal SB.
Shiga-like toxin converting phages from Escherichia coli 0157:H7 strains
that cause hemorrhagic colitis or infantile diarrhea. Science
1984;226:694-6

 

74.0larte J, Ramos-Alvarez M. Epidemic diarrhea in premature infants.
Etiological significance of a newly recognized type of Escherichia coli
(Ol42:K86[B]:H6). Am J Dis Child 1965;109:436-8

 

75.0'Regan S, Robitaille P, Mongeau JG, et al. Hemolytic uremic syndrome
associated with ECHO 22 infection. Clin Pediatr 1980;19:125-7

76.¢rskov F. Genus I: Eacherichia in Kreig NR, ed. Bergey's manual of
systematic bacteriology. Balimore: Williams & Wilkins, 1984;l:420-3

77.¢rskov I, Orskov F. Serology of Escherichia goli fimbriae. Prog
Allergy 1983;33:80-105

78.Pai CH, Gordon R, Sims VH, Bryan LE. Sporadic cases of hemorrhagic
colitis associated with Escheziahia goli 0157:H7. Clinical,
epidemiologic, and bacteriologic features. Ann Intern Med 1984;101:738-
42

79.Perbal B. A pratical guide to molecular cloning. New York: John Wiley
& Sons, 1984:151-2

80.PHLS Communicable Disease Surveillance Center. Hemolytic uremic
syndrome. Communicable Disease Report l983;36:1

81.Prober CG, Tune B, Holder L. Xerainia paeuaotuberculosis septicemia.
Am J Dis Child 1979;133:623-4

54

82.Rappaport F, Hening E. Media for isolation and differentiation of
pathogenic Escherichia coii (serotype 0111 and 055). J Clin Microbiol
l952;5:311-2

83.Ray CG, Tucker VL, Harries DJ et al. Eateroviruses associated with
the hemolytic uremic syndrome. Pediatrics 1970;46:378-88

84.Ratnam S, March SB. Sporadic occurrence of hemorrhagic colitis
associated with Escherichia coli 0157:H7 in Newfoundland. Can Med Assoc
J l986;134:43-5,49

85.Reis MHL, Matos DP, Pestana de Castro AF, Toledo MRF, Trabulsi LR.
Relationship among enterotoxigenic phenotypes, serotypes, and sources of
strains in enterotoxigenic Escherichia coli. Infect Immun 1980;28:24—7

 

86.Remis RS, MacDonald KL, Riley LW, et al. Sporadic cases of
hemorrhagic colitis associated with Eschegichia aaii 0157:H7. Ann
Intern Med 1984;101:604-6

87.Riley LW, Remis RS, Helgerson SD, et al. Hemorrhagic colitis
associated with a rare Escherichia coli serotype. N Engl J Med
1982;308:681-5

 

88.Ryan CA, Tauxe RV, Hosek GW, et al. Escherichia aaii 0157:H7 diarrhea
in a nursing home: clinical, epidemiological, and pathological findings.
J Infect Dis 1986;154:631-8

89.Sack DA, Kaminsky DC, Sack RB, et al. Prophylactic doxycycline for
travellers' diarrhea. Results of a double-blind study of Peace Corps
volunteers in Kenya. N Engl J Med 1978;298:758-63

90.Sack RB. Enterotoxigenic Escherichia coli: identification and
characterization. J Infect Dis 1980;142:279-86

 

91.8chaberg DR, Zervos M. Plasmid analysis in the study of the
epidemiology of nosocomial Gram-positive cocci. Rev Infect Dis

1986;8z705-12

92.Sharp PA, Sugden B, Sambrook J. Detection of two restriction
endonuclease activities in Hemophilas parainfluenzae using analytical
agarose-ethidium bromide electrophoresis. Biochemistry 1973;12:3055-63

93.Shoyab M, Sen A. The isolation of extrachromosomal DNA by
hydroxyapatite chromatography. Meth Enzymol 1979;68:199-206

94.Smith HR, Day NP, Scotland SM, Gross RJ, Rowe B. Phage-determined
production of vero cytotoxin in strains of Escheriahia coli serogroup
0157. Lancet 1984;2:1242-3

 

95.Smith HW. Hemolysins of Eschegichia coli. J Pathol Bacteriol
1963;85:197-211

 

55

96.Spika JS, Parsons JE, Nordenberg D, Wells JG, Gunn RA, Blake PA.
Hemolytic uremic syndrome and diarrhea associated with Escherichia coli
0157:H7 in a day care center. J Pediatr 1986;109:287-91

97.Stewart PJ, Desmormeaux W, Chene J. Hemorrhagic colitis in a home for
the aged: Ontario. Can Dis Weekly Report 1983;9z29-32

98.Stougaard P, Molin S. Vertical dye-buoyant density gradients for
rapid analysis and preparation of plasmid DNA. Anal Biochem
1981;118:191-3

99.Strockbine NA, Marques LRM, Newland JW, Smith HW, Holmes RK, O'Brien
AD. Two toxin-converting phages from Escherichia coli 0157:H7 strain 933
encode antigenically distinct toxins with similar biologic activities.
Infect Immun 1986;53:135-40

 

100.Sussman M.E‘L coli in human and animal disease. In: Sussman M, ed.
The virulence of Escherichia coli. Orlando: Academic Press, 1985:7-45

 

 

101.Takahashi S, Nagano Y. Rapid procedure for isolation of plasmid DNA
and application to epidemiological analysis. J Clin Microbiol
1984;20:608-13

102.Tzipori S, Wachsmuth IK, Chapman C, et al. The pathogenesis of
hemorrhagic colitis caused by Escherichia coli 0157:H7 in gnotobiotic
piglet. J Infect Dis 1986;154:712-6

 

103.Wachsmuth K. Molcular epidemiology of bacterial infections: examples
of methodology and investigation of outbreaks. Rev Infect Dis
1986;8z682-92

104.Wells JG, Davis BR, Wachsmuth IK, et a1. Laboratory investigation of
hemorrhagic colitis outbreaks associated with a rare Escherichia coli
serotype. J Clin Microbiol 1983;18:512-20

 

105.Wells RD, Hardles SC, Horn GT, et al. RPC-5 column chromatography
for the isolation of DNA fragments. Meth Enzymol 1980;65:327-47

106.Zeugin JA, Hartley JL. Ethanol precipitation of DNA. Focus
1985;7(4):1-2

 

"I11111111111111“