GLUCGSE, ELECTROLYTE. AND FLUID
ABSORPTION FROM THE SMALL INTESTINE
0F GNOTOBIOTIC PIGS INFECTED WITH
ESCHERICHIA COLI‘

Thesis for the Degree of M‘ S.
MICHIGAN STATE UNIVERSITY
MARIANNE EILEEN SHIRK
e1971

 

I' N ES I ‘5 _
.\ ‘ ‘
1 "L 1., w
'70 - 7
, x: I
‘. é. ‘
111;! (I

 

   

LIBR A 3: z’vL‘

Michigan State
University

    
   
   

 
 
 
   

 
 

E: swims av E'?
. mm; a sun:
I mnmnmv mo.

LIBRARY BINDEF’S
srnmcront. mum]

    

ABSTRACT
GLUCOSE, ELECTROLYTE, AND FLUID ABSORPTION
FROM THE SMALL INTESTINE 0F GNOTOBIOTIC

PIGS INFECTED WITH ESCHERICHIA COLI

BY

Marianne Eileen Shirk

Eighteen Zeweek—old hysterotomyvderived gnotobiotic pigs from 2.
litters were used to study the effects of colibacillary diarrhea on the
absorption of.d-glutose, sodium, chloride, potassium, and fluid. The.
animals were paired and 1 pig from each pair was exposed per as to
3 x 106 Escherichia coli 0138:K81:NM.organisms. Twenty-four hours after
infection, isolated jejunal 100ps were prepared in each animal for in
vivo measurement of nutrient absorption.. These loops were continuously
perfused for a period of 4 hours with a solution of 26 mM glucose.in
lactated Ringer's solution.

Glucose, sodium, chloride, and fluid absorption was not impaired in
the E. coZi-infected loops. These monocontaminated loops consistently
absorbed more glucose.and produced less fluid than the germfrse control
loops. Mean glucose absorption (fiMJcm/4 hr) was 55.1 in the infected
animals-but only 28.4 in the-controls. Fluid production in the germfree
control loops averaged 0.46-ml/cm/4-hr.while the infected loops actually
absorbed 0.07 ml/cm/4 hr. Sodium.movement directly correlated with fluid

production (P<.001). Chloride movement generally followed a similar

Marianne Eileen.Shirk
pattern. The germfree intestinal loop secreted 68 qu/cm of sodium and.
41 qu/cm of chloride during the-4-hour period. The E. coZi-contaminated
loops absorbed 10 and 7 qu/cm/4 hr of sodium and chloride, respectively.
Differences in potassium secretion were not significant.

The most consistent hematologic finding in the pigs exposed to
E. coli was a marked leukocytosis with a regenerative shift to the left.
Some evidenceEOf dehydration was present in infected animals.

The histologic appearance of the duodenum, the intestine anterior
and posterior to the loop, and the ileum did not differ between control
and infected animals. Varying degrees of epithelial vacuolization,
submucosal edema, and cellular infiltration were noted. Similar changes
were also seen in the isolated 100ps.. The severity of the lesions did
not correlate with changes in absorption.

The findings in the-present investigation did not support the theory.
that malabsorption is-a prominent-feature in the pathogenesis of coli-

bacillary diarrhea.

GLUCOSE, ELECTROLYTE, AND FLUID ABSORPTION
FROM THE SMALL INTESTINE 0F GNOTOBIOTIC

PIGS INFECTED WITH ESCHERICHIA CCLI

By

Marianne Eileen Shirk

A THESIS

Submitted‘to
. MichiganIState university
in partial fulfillment of the requirements
for the degree of~

MASTERiOF SCIENCE

Department of Pathology

1971

’/

/

n],
'7<+"/

To my mother and father

Eleanor and Wesley Lickfeldt

ii

ACKNOWLEDGEMENTS

I wish to express my gratitude and appreciation to my major.pro—
fessor, Dr. Glenn L. waxler, for his expert guidance and encouragement
during the course of this research project.

I would also like to thank Dr. C. C. Morrill, Chairman of the
Department of Pathology, Michigan State University, for his support and
confidence in continuing my education. In addition, I would like to,
thank Drs. Allan Trapp and David Ellis for serving on my guidance
committee.

I am indebted to the following persons for their invaluable techni-
cal assistancez, Dre. D. 'T. Drees andB. R. Christie for aiding merin
the procurement of the gnotobietic pigs used in this research work; Mr.
James Southern, animal caretaker at the Veterinary Research Farm; Mrs.
Dottie Fenner, medical technologist; and Mrs. Mae Sunderlin and Mrs.

Frances Whipple, histologic technicians.

iii

TABLE OF CONTENTS

INTRODUCTION 0 O l O O O O A O O O O O O O O O O O .

LITERATURE REVIEW 0 O O O O O I O O . O O ‘ O O O O 0
me GnotOb iotic P18 0 I O O O O O O

Colibacillosis in Neonatal Pigs . . . .

Glucose and.Electrolyte Absorption in the

A Perfusion Technique for Absorption Studies. .‘

MATERIALS AND METHODS. . . . . . .-. . . . . .
General Plan. . . . . . .'. o.~ . . . .
Experimental Animals. . . . . .'. . . .
Bacteriologic Procedures. . . . . . . .
Hematology. . .'. . . . . . . . . . . .
Experimental Procedure. . . . . .'. . .
Chemical Determinations . . . . . . .'.
Calculations. . . . . . . . . . . . . .

RESULTS. . . . . . . . .‘. . . .~. . . . . . .
Procurement and Rearing Procedures. .'.
Bacteriologic Findings. . . . . .‘. .
Clinical and Surgical Observations. . .
Hematology. . . . .'. ... . . . . . . .
Analysis of Perfusion Fluid . . . . . .

Histopathology. . . . . . . . . .'. . .

DISCUSSION 0 O ‘0 O O O I. O O O .0 O O O ‘0 O O O '

“neral .Cpments O O O . O I . O O I. O O O 0

iv

Normal State.

Page

13

13

- 14

14
14.
15
17
17.
24
24
26
26
26
26
27
34
35
52

52

Hematology. . . . . . . . . . . . . . . . . . . . . . . . . 53
Intestinal Absorption . . . .'. . . ... . . .'. . . . . ~,- 54
Histopathology. . . . .'. . . . . . . .-. . . .-. . . . . ., 56
Speculations on the Pathogenesis of Colibacillary Diarrhea. 58
SUMMARY. . . . . . . . . .'. . . .-. . . . . . . .-. . . .-. . . .- 59
BIBLIOGRAPHY.. . . . . . . . . .-. . . .-. ._. .-. -.- .-. .g. . . 61

VITA o s o s o 'o s s o s o o o 's o o o 'o o o o Is a o o o o o o ‘0 o 67

Table-

10

LIST-OF TABLES

Results of hematocrit determinations (Z)- . . . . .~. . .

Length, placement, and distention of perfused loops . . .'

Comparison of body weight, sex, age, and length of
Bull intestine O O O O O O O O O O O O ' O O O O O O O O O

Preinfection hematologic findings . . .-. . . .'. . . . .

Mean values of preinfection hematologic findings. .

Postinfection hematologic findings. . . . . . .-. . . .‘.
Mean values of postinfection hematologic findings . . .‘.

Comparison of pH of intestinal contents in infected and
central p188 O O O O O O I .. O .0 O O O I O O O I. O I I ' O O

Fluid movement and d-glucose absorption in E. coli-
infected and control loops. . . . . . . . . . .'. . . .'.

Electrolyte movement in E. coZi-infected and control loops.

vi

Page.
28

29

30-
31~
32-
33~

34-

36

37

39

Figure

10

LIST OF FIGURES

Canulation of the perfused loop. The posterior seg-
ment (A) has-been canulated. The tubing to be placed

in the anterior portion of the loop is.shown (B). . . .

The .completed perfused loop. The anterior canula (A)

and the posterior canula (B) are in place. The 100p is

ready to be returned to the abdominal cavity. . . . . .

Laboratory setup. Following surgical procedures, the;
perfusion was.begun. Control pig (A), infected pig
(B), influx tube from reservoir (C), efflux tube from
reservoir (D), pump (E), reservoir (F), intravenous

tubing (G), water bath (H). .-. . . . . . . . .'. ._

Glucose absOrption and net fluid movement in E. coli-
infectedand control loops. . .'. .7. .-. . . .-. . .

Sodium iOn and net fluid movement in E. coZi-infected
and contr01 loops 0 ' O O O O O O O O O O I O '0 O C O 0

Fig 284 (infected). Duodenum 28 hours after exposure.
Only mild epithelial vacuolization (V) is present
accompanied by cellular infiltration into the lamina
propria (I). Note the abundance of goblet cells (G).
The villi are long and slender with little branching.

Pig 284 (infected). Field similar to that in Figure 6.

Note goblet cells (C) and vacuolate epithelium (V).
Neutrophils (N) are invading the lamina propria . . .

Pig 285 (control). Small intestine 5 cm anterior to

loop.- Extreme epithelial vacuolization (V) is evident

with dilatation of the lacteals (L). Noninflammatory
edema (E) is present in the submucosa. Note that the
epithelium is intact. . . . . . . . . . . . . . . . .

Pig 287 (control). Perfused 100p. Note the marked
epithelial vacuolization (V) and submucosal edema (E).
There is increased cellularity of the submucosa. The
epithelium is intact. The villi are long and slender.
No blunting or shortening is evident. . .-. . . . . .

Pig 286 (infected). Perfused loop. Vacuolate epithelium

is seen.” Note.the infiltration of the lamina propria
With neutrophils (N). o o o o o o :. o o o o o 'o o o 0

vii

Page

21

'21

'23

38

40'

41

42

- 43

45

46

Figure

11

12

13

Page

Fig.287 (control). Perfused loop. Marked fibrinous
serositis is present (8). Hyperemia (H) and cellular
infiltration are apparent (I). Muscularis mucosa is
designated (M). .'. . . . . °.° .‘. . . .‘. . . .'. . . . 47

Pig 197 (infected). .Perfused loop. Although epithelial
vacuolization is still extreme, only slight submucosal
edema is present (E). Note that the villi are intact . . 48

Pig 289.(control). Small intestine 5 cm anterior to

the ileocecal valve. Acute necrotic enteritis is

evident, with marked hemorrhage (H) and destruction of

the epithelium (D).. The submucosa is edematous (E),

with marked hyperemia (M). Note the presence of

lymphoid tissue (L) . . . . .-. . . . . . . .'. .1. .'. . 50

viii

INTRODUCTION

Neonatal colibacillary diarrhea or white scours is one of the most
perplexing problems facing the swine industry today. Large economic
losses have resulted from high mortality rates in baby.pigs during the-
first week of life. It_has been estimated that over 40% of all deaths
in neonatal pigs-are due to Escherichia coli infections (Kenworthy.
and Allen, 1966a).

Extensive research has been conducted on the pathogenesis of this.
disease, which is characterized by sudden onset, profuse diarrhea, rapid
dehydration and death. The mechanism.which causes the fluid accumula-
tion in the intestinal tract is still not clearly-understood. Does the
diarrhea result from increased intestinal secretion? Does malabsorption
become apparent as in transmissible gastroenteritis? Or are both of
these factors involved simultaneously?

An in viva system using the perfused intestinal 100p technique was
adapted to gnotobiotic pigs. The absorption of d-glucose, and the net-
movement of sodium, potassium, chloride, and fluid was determined in
E. coZi—infeeted and germfree animals, It was hoped that data obtained
from this study would support or contradict current theories on the

pathogenesis of colibacillary.diarrhea.

LITERATURE REVIEW

The Gnotobiotic Pig

 

Although the use of the gnotobiotic pig as an important research
animal has steadily increased since the early-19608, many characteristics
of this germfree animal have not yet been investigated (Alexander et'
al;, 1969).' Meyer at al. (1964) stated that the pig was a very good
experimental mode1.animal because of its similarities to man in the
areas of nutrition, hematology, anatomy.of the vascular system, skin,
eyes, and gastrointestinal tract. Alexander~et al. (1969), Bette et a1.
(1960), and Lecce and Keep (1962) reported that, because of its size
and ability for early immunologic competence, the gnotobiotic pig would
become an excellent.mode1 system for the study of immunity and disease.
Being colostrum-deprived, the germfree pig does not receive naturally
produced maternal antibodies, since swine have epithelial-chorial
placentas across which antibodies do not pass. In the study of.a par-
ticular disease, the germfree animal is free from interfering antibodies,
natural resistance, and intercurrent infections. With no immune sub?
stances present, the germfree animal is quite uniformly susceptible to

infectious agents (waxler, 1970).

Research Model for Colibacillosis. Kohler and Bohl (1966) pointed out
some advantages of using gnotobiotic pigs to study.colibacillosis. Gnoto-
biotic pigs at birth are free from demonstrable antibody and from other
infectious agents that may alter the course of the disease. Also,

2

3
there are no extraneous strains of E. coli present that could interfere
with productionoof the disease experimentally. Lastly, adequate control
animals are available. Kohler and Bohl (1966) infected gnotobiotic pigs~
with E. coli and found that the clinical signs appeared to be identical
to those seen in the naturally occurring disease caused by organisms of
the same serotype. Saunders et al. (1963) produced a clinical syndrome
comparable to field outbreaks of colibacillosis by exposing gnotobiotic
pigs to certain serotypes of E. coli. Kenworthy and Allen.(l966a) and
Miniats (1970) encountered difficulties when they attempted to reproduce
neonatal scours or edema disease in conventionally reared pigs. This
problem.has been attributed to a failure of E. coli organisms to establish
a significant population in the conventional host following artificial
infection. Escherichia coli will, however, readily colonize and pro-

liferate in the gnotobiotic gut.

Procurement. Waxler et al. (1966) and Waxler (1970) described 2 tech-
niques-hysterectomy and hysterotomy-dwhich have been successfully used
in obtaining germfree pigs. The hysterotomy technique was preferred for
several important reasons:

1) Individual attention can.be given to the newborn to stimulate'
breathing and increase survival rate.

2) Surgery is performed in a completely closed environment.

3) The sow can be saved for future litters if so desired.
Meyer et al. (1964) and Alexander et al. (1969), in contrast, favored the
hysterectomy.procedure because of the facility of time and the ease of
anesthetizing the sow. The pigs procured in this manner were free of

demonstrable fungi, bacteria, PPLO, ascarids, and viruses.

4

Colibacillosis in Neonatal Pig§_

 

Colibacillosis, White scours, or diarrhea neonatorum is an acute,
sometimes highly fatal; enteritis or gastroenteritis of suckling pigsv
(Dunne and Bennett, 1970). Kenworthy.and Allen (1966a) reported that a
1960 survey in England indicated 40% of the pigs that died at 4 months of
age or younger had enteritis associated with Escherichia-coli. In_addi-
tion, Kohler and Bohl (1966) referred to a survey in England in which
E. coli organisms were the pathogens most frequently linked to the.presence.
of disease in pigs less than 3 weeks of age. White scours in pigs and
calves has been compared to diarrhea in human-infants where E. chi has

been isolated as an infective agent (Saunders et aZ., 1960).

Bacterial Characteristics. Escherichia coli is a common inhabitor of

 

stables and pens and causes.diarrhea in neonatal farm animals (Barnum
et al., 1967).' Infection is dependent upon the pathogenicity of the,
bacterium-i.e., the ability to‘produce toxin--the resistance of the,
host, and the number of infecting organisms (Dunne and Bennett, 1970).
Many serotypes of E. coli have been identified by classification of
0, H, and K antigens. Somatic or O antigens are lipopolysaccharides.
One hundred forty—seven 0 antigens have been typed. There are 49 H'or
flagellar antigens and 89 K or sheath (envelope) antigens. All 0 anti?
gene have been isolated from colibacillary diarrhea with the exception
of 017, 018, 019, and 055. Over 812 of enteropathogenic strains of
E. coli are hemolytic. This is.not, however, important in the-patho-
genesis of this disease.(Kenworthy and Allen, 1966a).' Serotype 0138:K81
has been incriminated as a.pathogen.in both baby pig enteritis and edema
disease (Dunne and Bennett, 1970;.Gossling and Rhoades, 1966; Hagen and.

Haugen, 1963; Nielsen et aZ.,.l968). Hagen and Haugen (1963) studied

5
the enterotoxemic effects of several serotypes.of.E. coli including 0138
isolated from infected pigs. Cell-free extracts of the fecal material
from infected animals caused rapid death when injected intravenously into
pigs or mice.. Similar preparations from "healthy control animals" had no
effect when injected. Kohler (1968) concluded that the enterotoxic
activity of E. chi may be the cause of diarrhea in young pigs. Inocue
lations.withtcell-free filtrates in conventional piglets that had received
colostrum produced severe diarrhea in 1.5 to 4.5 hours. In addition,
comparable findings have been made in germfree pigs (Kohler and Cross,
1969).' Moon et al. (1966a) claimed that the ability of.a strain of
E. coZi’to cause distention in.a ligated loop system was an indication
of its enteropathogenicity and not its general pathogenicity. Nielsen
and Sautter (1968) reported that serotype 0138 was enteropathogenic as
indicated by the ligated 100p technique. Results have been variable
when researchers tried to reproduce distention in ligated loops with-
various enteropathogenic strains of.E. chi in young pigs (Truszczynski.
and Pilaszek, 1969). Smith and Halls (1967) were successful, however,
in correlating pathogenicity with loop dilation.

Kenworthy and Crabb (1963) reported that the intestinal tract of the
healthy piglet is sterile at birth. During colibacillosis, pathogens are
capable of almost completely replacing the-resident~intestinal.flora,
especially in the jejunum. Saunders et a1. (1963) stated that the number
of organisms had little effect on the_severity of infection. Host
resistance played a more important role. Six-day-old pigs were much-

more,resistant than lZ—houreold pigs.

Clinical Signs, Colibacillosis occurs in pigs 1 to 8 days of age. Some

 

may become 111 within 12 hours following birth and die of an acute

6

septicemia in 48 hours with or without clinical signs of diarrhea. More
commonly, however, pigs exhibit a yellowishdwhite watery diarrhea, dehydra-
tion, and rough hair coats. Tails may become necrotic and slough. Mor-
tality is highest during the first week of life when it.may reach 70% or
more (Dunne and Bennett, 1970). Saunders et‘aZ. (1960) found that rapid
death of most of a litter occurred within the first 48 hours of life.
Colibacillosis occurs most frequently without bacteremia or morphologic

evidence of enteritis (Moon et al., 1966b).

 

Hematologic Aspects. Alexander at al. (1969) reported that the gross
morphology of internal organs and quantitative measurements of formed
blood elements were very similar between germfree and conventional pigs.
Meyer et a1. (1963) stated that the leukocyte and erythrocyte.counts and
hemoglobin levels of germfree pigs were within normal limits of conven-
tional values. Serum proteins, however, were low. Britt (1967) found
no evidence of anemia, hemoconcentration, or serum electrolyte (sodium
and potassium) derangements in germfree pigs infected with E. coli-
Infected pigs responded with either leukocytosis or leukopenia. No
significant difference was noted between infected and control groups in

regard to hemoglobin and packed cell volume.

Histopathologic Observations. Cross and Kohler (1969) studied the auto-

 

lytic changes in the intestinal tract of germfree, E. coZi-monocontaminated,
and conventional pigs. The most striking differences were found in the
ileum.and colon. The colonic submucosa in the germfree gut was.thinnsr

and had shorter, broader villi than either the E. coZi-monocontaminated

or conventional gut. The conventional and monocontaminated intestines

had increased cellularity compared with the germfree animals. The greatest

numbers of neutrophils, lymphocytes, macrOphages and histiocytes were.

7
found in the conventional animals; least numbers were found in the germs
free pigs; and the E-‘COZidmonocontaminated gut was.intermediate in this
respect. Cross and Kohler (1969) described the contaminated gut as being
in.a state of "physiological inflammation." Dubos et a1. (1963) described
the germfree intestine in the mouse as rather prenatal and undifferentiated.
There was an absence or near absence of inflammatory cells in the lamina
propria. ,The crypts were shallow and the villi were delicate. Distended
goblet cells were evident. This condition of the intestine was rapidly
changed with the addition of one or more microbial species. Staley at
al. (1968) studied the microscopic appearance of the newborn jejunum of
the pig.' They reported that the simple columnar epithelium contained few
goblet cells but that vacuoles were very prominent in the upper 2/3 of
the cells lining the long, slender villi. Alexander et a1. (1969)
reported similar findings in germfree pigs. The epithelial cells of the_
jejunum.and ileum.were extremely.vacuolate, especially at the tips of the
villi, until 7 weeks of age. The presence of cellular elements-varied.
considerably among germfree pigs of the same age group. Plasma cells,
eosinophils, and reticuloendothelial cells were found in both the.lamina
prOpria and the submucosa. This cellularity, however, was much less than
in-conventionallytreared pigs of the same age.

Gilka (1969) described the histopathologic lesions of the jejunum.
in newborn piglets with E- 301i gastroenteritis. He found inflammatory
edema in the lamina propria with cellular infiltration, prhmarily neutro-
phils, in the villi. Thirteen.of 17 pigs had no glycogen in the hepatic
cells. No specific changes were seen in the kidney. Dunne and Bennett
(1970) reported that the intestinal reaction of newborn pigs with coli-
bacillosis was similar to the reaction to many low-grade irritants.

Grossly, the intestines were filled with-a yellowish-gray mucous material.

8
Microscopically, there was.a marked distention and vacuolization of the
intestinal epithelial cells with enlarged mucus-producing goblet cells.
Kohler (1967a) reported the infiltration of neutrophils in the upper small
intestine of both E. coZi-infected and control pigs. No epithelial des-
quamation or villous atrophy was-seen in either group. Moon et al.
(1966) occasionally found denudation of the epithelium, neutrophilic
infiltration of the lamina propria and distention of the lamina propria
with nonprotein fluid in the jejunum. Inconsistent changes were reported
in other body organs. Moon et al. (1970a) described similar lesions
but noted that these changes were inconsistent from animal to animal.
They found no lesions in 7 of 18 newborn pigs infected with E. coli.
Seven of 18 exhibited loss of the intestinal epithelium, congestion and
some hemorrhage with thrombosis in the submucosal vessels. In_4 of thev
18 animals studied, villous atrophy was present in the anterior intestinal
tract while the remainder of the gut appeared normal. Smith and Jones
(1963) reported no difference between baby pigs with E. coli diarrhea
and their normal healthy litter mates during the first week of life. No
evidence of inflammation was present. Kenworthy and Allen (1966b)
described the monocontaminated gut as having long, slender, uniform villi
with little cellular infiltration. They concluded that the morphology of
the small intestinal epithelium was related to a biochemical interaction
between bacteria and diet. Whether these bacteria were antagonistic or,
symbiotic was important. Kohler and Bohl (1966), in studying lesions of
colibacillosis in gnotobiotic pigs, saw no inflammatory.reaction in either
the small or large intestine in control and infected animals. Kohler and
Cross (1969) found no histopathologic lesions in gnotobiotic pigs with

diarrhea produced by cell-free filtrates of E. coli cultures.

9

Pathogenesis. Moon et al. (1970b) and Nielsen et a1. (1968) concluded

 

that the pathogenesis of cholera (Vibrio cholerae) in man and colibacillosis
in animals was similar. Because of these similarities, references to the
pathogenesis of colibacillosis in this thesis will be supplemented with
research material on Vibrio cholerae infection in man and in other experi-
mental animal model systems.

Nielsen et a2. (1968) divided the pathogenesis of colibacillary diar-
rhea into 4 important categories: (1) infection, (2) proliferation, (3)

production of enterotoxin and (4) production of lesions.-

Infection. Infection was considered to be dose-dependent: i.e.,
the greater the number of organisms to which the animal was exposed, the
greater the possibility of infection. Dunne and Bennett (1970) stated
that a septicemia develops in the newborn, especially if the infective
dose was given prior to the ingestion of colostrum. Kenworthy and Allen
(1966a) argued that the amount of fluid accumulation was independent of

the number of infecting organisms.‘

Proliferation. Proliferation of E. chi organisms is due to several

 

physiologic factors. First, baby.pigs are achlorhydric for a period
lasting from 1 to 14 days after birth. The proliferation of organisms
is therefore not retarded by the-low gastric pH normally present in
older pigs (Dunne and Bennett, 1970; Barnum et al.. 1967). Smith and
Jones (1963) found that, even if the gastric pH decreased soon after.
birth, E. coli organisms had the ability to proliferate in great.numbers
in the anterior small intestine where the population was normally small.
Secondly, the mucus-producing cells of the gut do not actively secrete
until after the first meal. The.bacteria are therefore able to cling to

the mucosa and increase in large numbers in the anterior small intestine

10
(Nielsen et aZ., 1968). In addition, the newborn pig's intestine is
sterile at birth. No established flora or nutrient competition is pres-
ent (Barnum et aZ., 1967). Kohler and Cross (1969) concluded that coli—
bacillary diarrhea could not be explained in terms of growth rate or

distribution between pathogenic and nonpathogenic strains.

Production of enterotoxin. Escherichia coli organisms are capable

 

of producing an enterotoxin that causes dilation in ligated gut loops.
The reaction is similar to that seen with Vibpio cholerae organisms.

This cell-free filtrate causes diarrhea without grossly altering the
mucosal epithelium (Nielsen et aZ., 1968). Moon et a1. (1970) produced
comparable experimental results with E. coli and V, cholerae enterotoxins
in the rabbit ileum. Kohler and Cross (1969) were able to reproduce
clinical signs of colibacillosis with cell-free filtrates in gnotobiotic
pigs. Burrows and Musteikis (1966) concluded that vibriosis was essen-
tially a toxemia. Its effect in the gut could not be blocked by anti-

biotics, nor was it a result of increased bacterial population.

Production of lesions. Many theories have been proposed to explain

 

the pathogenesis of the fluid accumulation seen in colibacillosis;'

l) Capillary permeability is altered causing the escape of large
osmotically active particles into the intestinal lumen. There is also
an increase in hydrostatic pressure (Barnum.et az,, 1969; De and Chatterje,
1953; Fordtran, 1967; Norris and Majno, 1968). Moon et al. (1966a)
reported, however, that there was no difference in the osmolarity of the.
luminal contents between control and E. cOZi-infected ligated-loops.
They concluded that fluid accumulation occurred through an intact membrane
before inflammation occurred.' Norris and Majno (1968) found capillary.

endothelium and basement membranes of the intestinal epithelial cells

ll
intact both under the light and electron microscope. In addition, no
intravenous Evans blue dye appeared either in the intestinal wall or in
the efflux fluid (Norris and Majno, 1968; Leitch et al., 1966). Moreover,
the protein content of control and infected loops was very low (Norris
and Majno, 1968; Nielsen and Sautter, 1968). It is impossible to explain
fluid production on the basis of hydrostatic pressure alone. The escape
of osmotically active particles does not occur because intraluminal fluid
during infection remains either isosmotic with plasma or hypotonic
(Nielsen etaZ., 1968).

2) Absorption of osmotically active particles is blocked-—i.e., by
inhibition of the sodium pump (Fordtran, 1967; Norris and Majno, 1968).
Nielsen and Sautter (1968) and Leitch et al. (1967) found no disturbance
in the sodium-absorptive mechanism. Norris et‘aZ. (1967) and Norris and
Majno (1967) had similar results using radioactive sodium. Moon.et al.
(1966a) found that the sodium content.per unit volume was the same in
E. coli and control ligated loops..

3) Bacteria produce osmotically active metabolites; however, entero-
toxin does not_increase osmotic pressure of luminal contents (Barnum et
aZ., 1967).

4) Changes in the absorptive surface (epithelial alteration) occurs
causing a decrease in efficiency to absorb particles--i.e., malabsorption
(Fordtran, 1967; Kenworthy and Allen, 1966a; Dunne and Bennett, 1970;
Barnum et aZ., 1967; Norris and Majno, 1968). Kohler (1967b) concluded
that colibacillary diarrhea was not.explainab1e_by morphologic changes.in
the digestive tract. Norris and Majno (1968) found the electrolyte con-
tent of cholera ligated loops to be very close to normal values of.
intestinal fluid. Torres-Pinedo et al. (1966) found impaired jejunal

absorption of glucose, sodium chloride and fluid during E. chi diarrhea

12

in human infants. Bywater (1970) found that glucose absorption was not
altered using E. chi enterotoxin in calves. Carpenter et al. (1968),
using Vibric cholerae enterotoxin in dogs, found glucose absorption
unaffected. Moreover, intraluminal glucose actually enhanced isotonic
fluid absorption from the ileum and jejunum. Fordtran (1967) reported
that glucose absorption during cholera decreased sodium and fluid losses.
Serebro et al. (1968) found equal or slightly increased glucose absorpw
tion in rabbits affected with vibriosis when compared to control animals.

5) <Direct stimulatory effect on the epithelial cell by the bacteria
or their products (Barnum.et-aZ, 1967; Norris and Majno, 1968); Nielsen
et al. (1968) concluded that there was an irritant stimulation to the-
mucosal cells to secrete fluid. This intestinal fluid was equal in come
position to normal intestinal secretions. This secretory effect could
be rapidly reversed by using Diamox (Norris and Majno, 1968; Leitch and.
Burrows, 1968). This-theory is rapidly gaining popularity as one of the.
most logical explanations of the pathogenesis.of colibacillary diarrhea

(Barnum et al., 1967).

Additional Clinicopathologic Alterations. Medway et a1. (1969) reported
that death’in neonatal diarrheain calveswas due to severe metabolic
acidosis from loss of bicarbonate ions.' Plasma potassium.and chloride
levels were not altered.‘ Plasma sodium did not.decrease until 4 days
after infection., Fisher (1965) and Barnum et al. (1967) concluded that
neonatal death from colibacillosis in pigs and calves was due to progres-
sive dehydration with decreased plasma sodium levels and increased.
potassium levels leading to cardiac failure. Moon et al. (1970) reported
similar findings in l-day-old baby.pigs. Benyajata (1966) reported no
alterations in serum electrolyte values in adult human experimental

cholera produced by choleragen.

13

Glucose and Electrolyte Absorption in the Normal State

 

According to Fordtran (1967) the driving force for sodium absorption
in.the duodenum and jejunum is the active transport of nonelectrolyte
substances like glucose and amino acids. Simultaneous fluid absorption
also occurs. These active transport systems operate either within the
cell membrane by actual penetration of solutes or by passage through
aqueous pores or channels. Due to differences in pore sizes-in various
areas of the intestine, the intraluminal fluid composition of solutes is
therefore related to the level of intestine being studied.

Crane (1960) stated that water absorption was necessary for glucose
absorption.- Newey (1967) concluded that sodium was essential for glucose
absorption. In addition, Newey supported the carrier mechanism hypothesis
in which the sodium-glucose carrier complex moves in response to sodium
ion concentration.. Further, maximum glucose absorption occurs in the
upper small intestine. Even in the presence of high concentrations of

glucose, the absorptive capacity of the gut does not become saturated.

A Perfusion Technique for Absorption Studies -

 

Jacobs and Luper (1957) first reported the use of an in viva pere
fusion system in rats to study intestinal absorption. The main advantages
of this system were an approximation of the normal living state.of the
animal and, secondly, the perfusion fluid could be sampled at any time
during the experiment. They found this to be a very reliable method of
measuring the absorption of nutrients. Serebro et aZ; (1968) adapted
this technique to measure the absorption of d-glucose and fluid produc-

tion in the jejunum of rabbits with toxin-induced cholera.

MATERIALS AND METHODS

General Plan

 

The pathogenesis of diarrhea produced by Escherichia coli is not
clearly understood. This research project was proposed to determine if
intestinal malabsorption and/or increased intestinal secretion by the‘

E. coli-infected gut may be responsible for the diarrhea. Germfree pigs
derived by hysterotomy were the.experimental animals. Using the isolated
loop technique described by Serebro et a1. (1968), net absorption of
glucose and the net absorption or production of fluid and electrolytes--
namely sodium, potassium,and chloride-dwere measured during a 4-hour
perfusion period. Animals were paired according to weight and sex. One
pig from each pair was_infected per 08 with E. coli serotype-Ol38:K81:NM.

A total of 18 jejunal loops was studied.

Experimental Animals

 

Procurement. Using the technique described by Waxler et a2. (1966), 18

 

germfree pigs were delivered by hysterotomy into sterile plastic isolators.
Yorkshire gilts in their 112th day of gestation were anesthetized with

20 to 25 ml of 2.5% procaine hydrochloride* injected into the epidural
space at the lumbosacral articulation. This was.followed by presurgical

*a,
tranquilization with promazine hydrochloride. The gilts were killed

 

*
Epidural,.Haver-Lockhart, Kansas City, Mo.

**
Sparine, Wyeth Laboratories, Cleveland, Ohio.

14,

15
upon completion of delivery. Litter 1 contained 8 animals. Ten live
pigs and l stillborn pig were obtained from Litter 2. The newborn.germe
free pigs were then transferred to the rearing room where they remained

for approximately 10 days until absorption studies were begun.

Rearing. In the rearing area, the newborn pigs were placed in isolators
so that they could be paired according to sex and weight as closely as
possible. No more than 4 pigs were placed in an isolator. These experi-
mental animals were maintained on a sterile semisynthetic diet.* In
Litter 1, pigs were fed 3 ounces 3 times daily (t.i.d.) for 3.5 days,

4 ounces t.i.d. for 3.5 days, and 5 ounces t.i.d. for the remainder of
the experiment. With Litter 2, however, dietary intake was increased in
an attempt to increase pig size to facilitate surgical technique. Three
ounces t.i.d. on Day 1 were followed by 4 ounces t.i.d. for the next 4.5

days.and the 5 ounces t.i.d. as before.

Bacteriologic Precedures

 

Preparation ofiInfective Agent. Escherichia coli serotype 0138:K81:NM

 

was used in this research project to produce colibacillosis in the:
previously germfree animals. This organism had been cultured, preserved
by lyophilization in glass ampules, and then stored at -20 C until needed
(Christie, 1967). Three days prior to infection of the first experimental
animal, an ampule was opened under sterile conditions and the contents
were inoculated into semisolid 3111“ medium (0.15 gm agar/100 ml). This

culture was maintained at 37 C. Blood agar plates39* were streaked at the

 

*

SPF-Lac, Borden Company, New York, N.Y.'
**

Bacto Brain Heart Infusion, Difco Laboratories, Detroit 1, Mich.

sea
Defibrinated Sheep Cells, Colorado Serum Co., Denver, Colorado.
Tryptose Blood Agar, Difco Laboratories, Detroit 1, Mich.

16

same time and incubated for 24 hours to insure the presence of a pure
culture osz. coZi. Eighteen hours prior to the intended exposure time,
E. coli organisms were transferred from the BHI medium to liquid thio-
glycollate* medium and incubated at 37 C. (For Litter 2, nutrient tryptose
agar slants** were used in place of the semisolid BHI medium.)

A series of standard density (fixed number of particles/ml) nephelometer
tubes was prepared according to the method of McFarland (Hepler, 1968).
In addition, a fixed amount of thioglycollate medium equal to the.concen-
tration in the culture tubes was placed into each nephelometer tube. In
this way, the density of the culture tubes could be compared to the standard
with a higher degree of accuracy. After estimation of the initial density
of the culture vial, serial dilutions were made with sterile physiological
saline solution to obtain a final concentration of approximately 3 x106
organisms/ml. The usual dilution procedure was as follows: 0.2 m1 of
the 18-hour thioglycollate culture was pipetted into 10 ml of sterile
saline. One milliliter of this solution was transferred into 9 m1 of

sterile saline, resulting in a final concentration of.approximate1y

3 x 106 organisms/ml (dilution factor of 1:500).

Method of Exposure. Approximately 24 hours before perfusion studies were
begun, 1 ml of inoculum.containing 3 x 106 organisms/m1 was passed into
the isolator reserved for infection. The pig intended for exposure was.
also transferred into this isolator. The screw-capped vial containing
.thfi culture was opened within the unit, mixed with the semisynthetic

diet and fed to theganimalg'

 

*
Fluid Thisglycollate Medium, Difco Laboratories, Detroit 1,
Mich. . ‘ . '

**
Tryptose Agar, Difco Laboratories, Detroit 1, Mich.

17

Determination of Sterility of Environment. When the pigs were 5 and 14

 

days of age, swabs from rectal and cage areas.were removed from the
isolators and both inoculated into BHI and on.blood agar plates. They
were then incubated at 37 C and examined at routine intervals for evidence

of growth.

Hematology‘
At 12 days of age, 3-m1 blood samples were obtained from each pig

in Litter l. The initial preinfection samples were collected in ethylene-
diaminetetraacetic acid (EDTA)-treated vials. Postinfection sampling was
done just prior to preparation of the isolated loops. The same procedure
was carried out on Litter 2 with the exception that the preinfection
samples were taken at-8 days of age.

Total leukocyte counts, differential leukocyte counts, microhemato-
crit packed cell volume, and hemoglobin determination by the cyanmethemo-

globin technique were conducted according to Benjamin (1961).

Experimental Procedure

 

Preparation of Animals. Approximately.24’houre following oral exposure.
to E. chi, the animal pair (1 control and l infected) was removed from
the rearing room and transferred to the experimental surgical area. Food
had been withheld 5 hours prior to scheduled surgery time. Each infected
animalwas observed for clinical signs of colibacillosis. The pigs were

weighed and the results recorded.

Preparation of Materials. Sterile technique was practiced during the

 

experimental surgery whenever possible. Instruments, gloves, perfusion.

tubing, and venous catheters were sterilized either by chemical means or

. ..

18
by autoclaving. The perfusion solution contained 26.mM* glucose (468
mg/100 ml) and 0.3 mM phenolsulfonphthalein (PSP)** in lactated Ringer's.
solution.*** Flasks (250 m1 capacity) containing 100 ml of the solution

were autoclaved and then frozen until required.

Surgical Technique; Pigs were anesthetized with sodium pentobarbital.+

The initial dosage was estimated at 0.5 ml/kg and given to effect via the
anterior vena cava.

When surgical anesthesia was reached, the jugular vein of each pig
was canulated with polyethylene tubing.H The catheter was then connected
to an intravenous fluid administration outfit.+++‘ Lactated Ringer's
solution was the fluid used to maintain hydration at the rate of 60 ml/hr
or 1 drop every 3 seconds. Additional anesthesia could be administered
via this route. Hematocrit readings were taken from the ear vein every
hour to monitor the state of hydration. All surgical manipulations,were,
performed on the germfree control animals first.‘

Following surgical preparation of the area, a 6-inch midline abdominal
incision was made beginning just posterior to the umbilicus. The ileo-
cecal junction was,located and brought up through the incision.' Using

the approximation that 1/2 foot equals 7 to 10 vascular arcades, a point

 

*
mM.- millimoles.

**
Phenol Red, Hartman-Leddon Co., Philadelphia, Pa.

***
Lactated Ringer's Injection, USP, Abbott Laboratories, North

Chicago, Ill. 60064.
1.Pentobarbital Sodium Injection, Haver-Lockhart, Kansas City, Mo.

++Intramedic, Clay Adams Division of Becton-Dickinson and Co.,
Parsippany, N.J. 07054.

+++Veno-pak, Abbott.Laboratories, North Chicago, 111. 60064..

l9
3 feet anterior to the ileocecal junction was identified and an intestinal
clamp.applied. A second point 20 cm anterior to the first was determined
and clamped off. The anterior and posterior ends of the loop were estabé
lished. The intestinal contents in the loop had been removed by gentle
massage prior to clamping.

The posterior segment of the loop was canulated first with 1/4 inch
diameter tygon tubing* (Figure l). The walls of that portion of the
tubing inside the loOps were perforated with several holes to insure.
adequate drainage of the loop. The anterior segment was canulated with
unperforated tubing (Figure 2). Caution was exercised to make sure that
the loop had been securely fastened to the tubing because of the slippery
surfaces of the intestine and of the canula. The intestine anterior and
posterior to the loop was ligated during preparation of the loop to pre-
vent.spillage.into the abdominal cavity. The loop was then replaced
into the abdomen. The incision was held together-with towel clamps
leaving only the tubing exposed. This method of closure facilitated
periodic.observations of the loop to make sure overdistention of the
loop had not occurred.' The total time required to complete the surgery

was 45 to 60 minutes.

Perfusion. When preparation of the jejunal isolated loop had been.com-,
**

pleted in each animal, the perfusion pump was.started.‘ The solution

was pumped at the rate of 2 ml/minute from the reservoir flask through

the-loop-and back into the reservoir flask. The perfusing fluid was

 

*
Tygon, U.S. Stoneware, Akron, Ohio.

**
Sigmamotor Model T-8, Middleport, N.Y.

20

Figure l. Canulation of the perfused loop. The posterior
segment (A) has been canulated. The tubing to be placed in the
anterior portion of the loop is shown (B).

Figure.2.‘ The completed perfused loop. .The anterior canula (A)
and the posterior canula (B) are.in.place. The‘loop is ready to be
returned to the abdominal cavity.

 

21~

 

 

 

 

 

 

 

Figure 2-

22
continually circulated in this manner for 4 hours. The reservoir flasks
were held at 40 C by the use of a water bath* (Figure 3).

During the 4-hour experimental period, body temperature was moniv
tored and kept_within the normal-range (100 to 102 F) with the aid of 2
heat lamps. In Litter 2, plasticecoated heating pads (placed under the
pigs) were used in place of the heat lamps.u

ThreeUmilliliter samples were taken from.each reservoir at 0, 1, 2,.
3, and 4 hours. These samples were placed into stappered vials and
frozen for analysis of glucose,_sodium, potassium, and chloride concen-.
trations. At the termination of the experiment any fluid remaining in

the loop was squeezed into the reservoir.

Necropsy and Histopathology. The pigs were euthanatized at.the end of

 

the 4-hour period with a lethal dose of sodium pentobarbital. The pH ofv
the duodenum, jejunum (anterior to the loOp),and ileum was measured with
pH papers.**

The length of the loop, distance of the distal end of the loOp from
the.ileocecal valve,and the length of the entire small intestine (pylorus-
to cecum) were measured. Tissues were taken for histopathologic examina-
tion from the-duodenum, 2 inches-anterior to the loop, midloop, 2 inches
posterior to the loop, the ileum (4 inches anterior to the ileocecal

valve), liver,and kidney. These samples were preserved in 10% buffered

formalin and later sectioned at 6 u and stained with hematoxylin and eosin.

 

* ' ,.
#DES 117720 Precision Scientific Co., Chicago, Ill.

** . .
pHydrion Papers, Micro Essential Laboratory, Brooklyn, N.Y.

23

 

 

 

 

 

Figure 3. Laboratory setup. Following surgical pro-
cedures, the perfusion was begun. Control pig (A), infected
pig (B), influx tube from reservoir (C), efflux tube from
reservoir (D), pump (E), reservoir (F), intravenous tubing
(C), water bath (H).

24

Chemical Determinations

 

Each 3-m1 sample was thawed and analyzed for glucose, sodium, potas-
sium, and chloride concentrations. Glucose content was measured by the
ortho-toluidine method (Feteris, 1965). Sodium and potassium ions were,
determined by flame photometeryf according to the method described by
Gradwohl (1956). Chloride ion concentration was estimated using the
method of Schales and Schales (1941). The volume of fluid remaining in.
the reservoir was measured in a graduated cylinder. Fifteen milliliters--
i.e., the amount taken out during sampling-dwas added to this.figure to

obtain the total volume of fluid remaining in the reservoir.

Calculations

 

In the present study, absorption is defined as the disappearance of
a particular nutrient from the perfusion fluid. Production is considered
an increase in the amount of,a particular nutrient in the perfusion fluid.

The amount of glucose absorbed by the isolated loop during the 4-
hour experiment was determined in the following manner:

1). Glucose absorbed (mg) - initial glucose in reservoir (mg) -
glucose (mg) removed in sampling - glucose (mg) in reservoir after the
4-hour perfusion.

2) The initial glucose in reservoir (mg) - glucose concentration
of sample I (mg/ml) x 100 ml.

3) Glucose (mg) removed in sampling - [concentration in Sample I
(mg/m1) x 3 m1 (amount in each sample) + Sample II x.3 ml...;.)].

4) Glucose (mg) left in the reservoir - concentration in Sample V

(mg/m1) x the final volume (ml) in the reservoir.‘

 

*
Coleman.Model 21 Flame Photometer, Coleman Instruments, Inc.,
42 Madison.Street, Haywood, Ill.

25

The final results were converted into the units mM/cm/4 hr:

glucose (mg) absorbed * length of loop (cm) % 180 mg glucose mM.

Absorption or production of sodium, potassium,and chloride ions was.
determined in a similar manner except that the final expression was calcu-
1ated in-qu/cm/4 hr. Fluid movement was estimated in this manner:

Initial volume (100 m1) - [final volume (ml) + 15 ml (removed in.
sampling)] - fluid absorbed. A negative number indicates production.
Final figures were expressed in ml/cm/4 hr - fluid movement (ml) % length
of loop (cm). Statistical analysis of the data between pairs of animals'

was carried.out by the_£_test method (Goulden, 1952).

RESULTS

Procurement and Rearinngrocedures
Survival rate was lOOZ-in each litter of germfree pigs delivered
by hysterotomy. These baby pigs adapted readily to the artificial
environment.‘ They quickly learned to drink from the feeding pans

within the stainless.steel cages.

Bacteriolggic Findings,

None of the swabs-taken from rectal and cage-areas within the iso-
lated environment produced bacterial growth either on blood agar or in.
BHI media. E. coli was recovered in pure culture from.each experimentally
infected animal 24 hours after exposure. AllJthe control pigs remained

germfree.until they were removed from the isolators for surgery.

Clinical and Surgical Observations
Profuse watery diarrhea with an inflamed perineal area was the most -
consistent clinical sign of colibacillosis in the monocontaminated animals.
24-hours after exposure. Appetites of the infected animals remained

normal.

Anesthesia. Sodium pentobarbital was a very satisfactory anesthetic agent.
It was seldom necessary to administer additional amounts during the 4-

hour experimental period.

26

27
Hydration. Although individual variation occurred, in general, intra-
venous fluid supplementation was adequate to maintain hydration during
the perfusion study using a flow rate of l drop/3 seconds. Hematocrit

readings are recorded in Table 1.

Measurement of the Intestine. At necropsy, considerable variation existed
in the placement of the perfused 100p--i.e., the distance of the posterior
segment of the loop from the ileocecal valve. This was noted‘especially}
in Pairs 5 and 9. The length of the loop as determined at necrOpsy was
consistent with the estimations made during surgery. Differences between
loop lengths in infected and control animals were not statistically sig-

nificant. One loop, Pig 196, was observed to be overdistended (Table 2).

Animal Data. The body weight, sex, and length of small intestine are

 

listed in Table 3. The average weight-was 1.72 kg. The mean intestinal
length from the pylorus to the ileocecal valve was 463.2 cm (15.25 ft).
Increase.in body weight.correlated roughly with an increase in intestinal

length.

Hematology

The hematologic findings from preinfection samples from Litters l
and 2 are shown in Table 4. The arithmetic means of these values are.
summarized in Table 5. The lZ-day-old pigs had consistently higher hemo-
globin.and packed cell volume values. In addition, the ratio of neutro-
phils to lymphocytes was greater in the younger pigs. No obvious dif-

ference was noted between the other parameters.

28

Table 1. Results of hematocrit determinations (Z)

 

 

 

 

 

 

Infected* Control'

Pair Time (hr) Time (hr)

Number 0 2 4 0 2 4
l 37 38.2’ 46 32 27 29.5
2 34 35.5 43 30 29' 28.2‘
3 39 43.5 43 30 30.5 35.2
4 . 42. 45 39 24.5 28 34
5 24 27 34 25 27 27
6 35 36 38 26.5. 31 30
7 37.5 36 37 28 32 36
8 38 35 35 21.5 35 35
9 30 33 32 27 27 27

Mean 35.2 36.6 38.6 27.2 29.6 31.3

 

29

Table 2. Length, placement, and distention of perfused loops

 

 

Length of Loop (cm) Placement of Loop+(cm) Distention

 

 

 

 

Pair Number. 1* C** I C I C
1 23.5 22.5 N.D.++ 81.3 - -
2 22.0 22.0 106.7 91.4 - -
3 20.0 21.0 68.6 85.1 ‘ — -
4 20.0 20.0 66.0 63.5 + -
5 27.9 24.1 76.2- 142.2 - -
6 30.54 38.1‘ 99.1. 108.0 - -
7 29.2~ 27.9 104.1 105.4 - -
8 35.6 31.7 100.3 119.4 - -
9- 25.4 25.4 91.4 121.9 - -
Mean 26.0 25.8 89.0 104.6

 

e
I - Infected.

**
C - Controls.

1'Distance of posterior end of loop from ileocecal valve.

++N.D. - Not Determined.

30

Table 3. Comparison of body weight, sex, age, and length of small

 

 

 

intestine
Length of
Small
Animal Age Body Weight Intestine
Number (days) Sex (kg) ' (cm)
190* 12 F 1.18 411.5
191 12 M 1.70 823.4
192 13 M 1.52 411.5~
193 13 F- 1.52 411.5
194 14 F 1.28 398.8
195 14 M 1.52 398.8
196 v 15 M 1.42 325.1
197 15 M 1.38 396.2
283 11 M l.90~ 452.1
284 11 M 1.75 457.2
285 12 M 2.08 487.7
286 12 F 1.55 449.6
287 13 F 2.08- 508.0
288 13 M 2.18 510.6
289 14 M 2.00 434.3
290 14 M 1.80 449.6
291 15 F 2.50- 561.3
292 15 F 1.60 483.0
Mean 1.72 463.2

 

*
Even numbers - infected animals.

31

 

 

 

 

 

 

.Haseoewn I m .eumuoaoa I z .eumooneaha I A .Afinaouusea wouseswmwmwmvmomm .wwn wwwmmwpwemno em.mwmmwmmeoe I m
H H o 0 N4 N em «H6.N N.NN 6.N m NmN
o H N o mm 0 on emo.OH N.NN N.m m wHaN
o o H N as m mm NNN.N N.NN N.N m oNN
m o o N we o em oem.m N.mN n.» m NmN
m H o N os m .Nm omN.o o.om N.N N NNN
H H o H Ne n Hm oom.HH o.NN N.» N NNN
H o H N He N Na NNN.6 c.6N N.» m 6NN
N o o H Nm H (we ooo.oH N.6N m.N N mNN
o o o H .ns o em omH.N N.NN ¢.m m «NN
N H o s «e N es NNN.N N.NN ,H.N N NNN
o o o H Nn m an QHn.e «N m.m NH NNH
s o o o .No o Hm an.e oe N.HH -NH 6mH
H o o H mm 4 NN ooN.N mm m.m .NH an
o o o H Nn N am onm.N oe N.oH NH emH
N o o H as n .nN «on.N He o.HH NH NNH
o o o H 66 0 mm omN.N mm .m.oH .NH NmH
m o o N Ne e we NoN.oH ANN o.HH .NH HaH
N o o N .oe o Nm oHo.N 6N N.m NH omH

ommz m m z 4 .m2 «New Naa\om3 Auv HHa oonaev finesse nonasz

mHHemsuuaoz Hanna >om .eHnoHNoamm ,ume NHN

Anv.unsoo euwuoxsoa Howusouomwwn oeuoeuuoo
.quHvawm onoHoumaen dowuoomsfioum .e edema

32-

Table 5. Mean values of preinfection hematologic findings

 

 

 

Determination Litter l Litter 2
Hemoglobin (gm/100 ml) 10.4 9.09
PCV (2) 37.4 27.5
Total corrected leukocyte count
(cells/mm3) 7,663 8,476
Neutrophils
Segmented (x) 37.6 49.9
Nonsegmented (X) 3.4 2.6
Lymphocytes (2) 57.9 44.9
Monocytes.(%) 1.1 1.7
Basophils (z) 0.0 0.4
Eosinophils (2) 0.0 0.5
NRBC (I100 leukocytes) 2.5 A. 1.5

 

The.hemograms from samples taken 24 hours following exposure of the
monocontaminated pigs to E. coZi are presented in Table 6. Mean values
from control and infected pigs were combined and compared in Table 7.

All infected animals, except Pig 284, had leukocytosis, relative and
absolute neutrophilia, relative lymphOpenia, and a regenerative left
shift. Moderate monocytosis also occurred. In-addition, the monocon-
taminated pigs had-consistently higher hemoglobin-and packed cell volume
values than the~germfree controls. Pig 284 had a relative neutrophilia
and lymphopenia with a left shift. However, total numbers of leukocytes-

were not increased.

33

.Haoo woods use woueoaoss I ommz

Haneoswmoo I m .Hasmoeon. I m .euho

 

 

 

 

 

 

Iomoa I: .euhooneaha I A .Hannouusea ceaseswoeaod I mz .eaasnouusoa mouseawem I new: .meflouuaooa

o o o N . _HN NH NN oNN.N NN N.N NN NH NNN
no N o N «N NH .NN ooN.NH NN N.HH _«N .«H oNN
o o o « HN NN oN oNN.NH NN N.HH «N NH NNN
o o o N «N NH NN NHN.NH NN N.HH NN NH NNN
o o o .N «N NH HN o«N.N «N N.N NN HH «NN
o o H o NN H NN oNN.N N.NN N.N -n- NH «HNN
H o .H H NN « NN NNo.« N.HN o.N -- .«H «NNN
H o o N «N N H« wNNN.« NN N.N -- NH «NNN
H o o N NN N NN NN«.N NN N.N -- NH .NNN
N o H H NN « NN NNN.N NN N.N -- HH «NNN
N o o N NH NH «N NoN.«H H« N.NH NN NH NNH
_H o o H N NN NN NNH.NH NN N.HH NN «H «NH
H o o N 0H NH oN H«N.NH NN N.NH NN NH NNH
H o o N NH NH NN NNN.NH NN H.NH NN NH NNH
N o o N NN 0 ON HNN.« N.NN N.N n- NH «NNH
o o o H NN « NN NNH.N NN N.N -- «H NNNH
N o o H NN N NN NNo.N NN N.N -- NH .NNH
N o o N NN N NN NNN.N NN N.NH -- “NH «HNH

oNNz N N. xx 3 H .Nz NNNNN NaaNUN3 HNV HHa NNHNaNV aoHuooNaH N.Nmev “unauz

eawnmouusoz .Hmuofi >om canoawoamm Houw< owd mam

ANV assoc sumooxsoq. Heaudouomman wouoouuoo meson
mNeHeaHN UHNoHoumsme eoHuoeuaHuNoN .N.oHHmN

34

Table 7. Mean values of postinfection hematologic findings.

 

 

 

Determination Control Infected
Hemoglobin (gm/100 ml), 8.72 11.1
PCV (Z)~ 26.5 34.4
Total corrected leukocyte count
(cells/m3) ’ ‘ ‘ 5,113 14,725
Neutrophils
Segmented (%) 38.3 60.7
Nonsegmented (Z) 3.65 16.9
Lymphocytes (2) 56.1 17.1
Monocytes (z) 1.95- 3.95-
Basophils (Z) 0.3 0.0
Eosinophils (2). 0.0 0.0
NRBC (/100 leukocytes) 2.2 1.8

 

Analysis of Perfusion Fluid .

 

pH Determinations. Determinations of pH were made at the termination of

 

the perfusion period from the duodenum, isolated loop, and terminal.
ileum (Table 8). None of the differences between pairs of.anima1s was

statistically significant when the £_test.was applied.

Glucose and Fluid Movement. The amount of glucose absorbed (uM/cm/4 hr)

 

and net fluid movement (m1/cm/4 hr) in E. cold-infected and germfree
loops are shown in Table 9. It is apparent that the infected animals
consistently absorbed more glucose (P<.01) and produced less fluid
(P<.01) than did the control pigs.' In Figure 4, net fluid movement is

plotted against glucose absorption.

35

Electrolyte Mbvement. The results of the net sodium, potassium, and
chloride ion exchange across the intestinal mucosa are listed in Table 10.
The.variation in absorption of sodium ions between germfree and mono-
contaminated loops was highly statistically significant (P<.001).'
Increased sodium ion secretion directly correlated with excessive fluid
production in control pigs (P<.001). Sodium was absorbed by most of the
monocontaminated loops (Figure 5). There was no statistical difference
in potassium and chloride movement when infected and control datanere

compared.

Histopatholggz.

 

Duodenum. Little difference was noted between-the macroscopic.structure
of the duodenum in the infected and control animals. Both the.germfree
and monocontaminated intestine had long; slender villi with varying
degrees of vacuolization of the epithelial cells and dilatation of the
subepithelial lymphatic-channels (Figure 6). Some hyperemia was seen.

in the submucosal vessels of both control and monocontaminated animals.
Inflammatory cells, including neutrophils, lymphocytes, plasma cells,

and reticuloendothelial cells, could be seen scattered through the tissue

of.both groups (Figure 7).

Small Intestine 5 cm Anterior to the Perfused Loop. At this level of

 

intestine, the epithelial cells appeared vacuolate. In-some instances
this distention was extreme.‘ Lacteals were dilated.' Noninflammatory~
submucosal edema was evident in almost every section examined (Figure 8).
Varying degrees of cellular infiltration with neutrophils were evident.
Reticuloendothelial cells, plasma cells, and lymphocytes could be identi-

fied. The severity of the submucosal edema correlated with the amount

36

Table 8. Comparison of pH of intestinal contents in infected and
control.pigs

 

 

 

 

 

 

 

Pair- Duodenum Perfused Loopp. Ileum
Number. 1* C** I C I
2 7 7 7.5. 7 8
3 6 7 7.5 8 8.5

4 7 8 8 7.5 6
5 6 7 7 7 5 7
6 6.5 7 7 6.5 7
7, 6.5 7. 6 7.5 7.5
8 6.5. 5.5 7 7.5 7
9 6.5 N.D.+ 7 7.5 7.5
Average \ 6.4 6.9 7.1 7.3- 7.4
*
Infected.
**
Control.
1.

N.D.-- Not Determined.

37

Table 9. Fluid movement and d-glucose absorption in E. coZi-infected
and control loops'

 

Net Fluid Movement
(m1/cm/4 hr)

 

Glucose absorbed
(uM/cm/4 hr)

 

 

Pair Number I* C I C
1 -0.19** —0.33. +48.3 +10.5
2 +0.11 -0.91 +48.l' +20.l
3 -0.10 -0.48 +18.2 +24.l
4 +0.38 -0.40 +20.7 +21.3
5 +0.03 +0.08 +56.9 +33.2
6 +0.07 -0.03 +66.7 +36.7
7 +0.27 -0.61 +77.2 +41.0
8 0.00 -0.88 +66.9 +37.8
9 +0.08 —0.55 +92.9 +30.5
Mean +0.07. -0.46 +55.l +28.4

 

*
I - infected, C --control.

**
+ indicates absorption, - indicates production.

38

X-Comu
O-hctcm

Gum W WOW"!-

 

 

 

 

-.8 -.6 -.4 -.2 0 0.2 9.4
Pen-rm Pt! me Pusan elm/3m. Poem

 

 

 

Figure 4. Glucoseabsorption and net fluid move-
ment in E. aoZi-infected and control loops.

39~

.soauoavoum mausoavea I “soaumuoenm,eoueoauau +
«e

.Houuaoo I u “masseuse 1 He

 

 

 

 

 

 

 

N.H«- _No.N + oN.N- oN.« - N.NN - o.oH+ can:
N.NN- NN.N n oH.N- .oN.N - N.NN - N.NN- N
N.NN- N.NN+ oN.Nn ‘oN.N - N.NHH- N.NH- N
N.NN- N.NN+ NH.N- oN.N - N.NNH- N.NN+ N
N.NN- N.NNn NN.N- .oN.N - N.NN - N.NN- N
N.NN- o.oN+ oo.N- NH.o u oo.N + N.Ne+ .N
N.NN- o.oN+ oN.N- N.NH- N.NN - N.N«+ «
N.NN- o.NH+ oN.N- oN.N.n . N.HN - oo.N +. N
N.NHn oN.N n .oN.N- oN.N - o.«HH- o.«H+ N

NN.N - N.HH- oH.«- oN.N - .«o.N«- «.N.NH+ H

o H o . H o .«H H.Nasz NHNN
Hue «NauNNNNV ANN «NaUNNNNV ANN «NaUNNNNV
.ssfiemsuom
.emooH Houuaou was weuoemaNINNQQ_.m sw.uaoae>oa euhaouuoeHm .OH.oHan

 

 

Figure 5..

MI Ion beam no: rum-m.

 

40

 

 

 

—.8 -.6 -.4 —.2 0 2.2 0.9
Penm- Itr Fww W um. hem

Sodium ion and net fluid movement in

E. coZi-infected and control loops.-

41

 

Figure 6. Pig 284 (infected). Duodenum 28 hours.
after exposure. Only mild epithelial vacuolization (V)
is present accompanied by cellular infiltration into the
lamina propria (1). Note the abundance of goblet cells
(G). The villi are long and slender with.little branch-
ing. Hematoxylin and eosin; x 150.

42

 

~--..o—r

 

 

Figure 7_, Pig 284 (infected). Field similar to
that in Figure 6. Note goblet cells (G) and vacuolate
epithelium (V). Neutrophils (N) are invading the lamina-
propria. Hematoxylin and.eosin; x 600.

‘y. eff:r‘
‘ ‘0
\- -

e_

- 5 ':.'+~v'”'r3
' 1... .. " '1‘:‘ 3:} ('1'? :.S
o, . 1‘ 3w,

'
err‘
a

 

 

 

Figure8. Pig 285 (control). Small intestine 5 cm
anterior to loop. ‘ Extreme epithelial vacuolization (V)
is evident with dilatation of the lacteals (L). Noninflam-
matory edema (E) is presentin the submucosa. ' Note that
the epithelium is intact. Hematoxylin and eosin; x .60.

44

of hyperemia and cellularity. The above changes were recorded in both
control and infected animals. The monocontaminated pigs tended to have

a greater number of gobletcells,~less epithelial vacuolization, and a
larger number of infiltrating cells than did the germfree intestine. The
contrast, however, was not striking. In~l infected animal (Pig 288),

an acute necrotic enteritis with denudation of the epithelium, extensive.

cellular invasion, and marked submucosal edema was observed.

Perfused Loop. This area of the intestine was the site of the absorption
study.; The following lesions were recorded in 10 animals (4 controls and
6infected). Six of these 10 animals were members of Pairs 6, 7, and
9—-1.e., Pigs 285 and 286, 287.and 288, and 291 and 292. Moderate to
extreme'vacuolization of the epithelium was noted. The epithelium was,.
however, intact in almost every section examined. Marked hyperemia of
the lamina propria and submucosa and.marked edema of the submucosa were
evident (Figure 9). Inflammatory cells were_increased in number when
compared to sections taken anterior to the loop. Neutrophils were pre—
dominant (Figure 10). Mild.to severe.fibrinous serositis was present in
each section (Figure 11).

The lesions recorded in the remaining 8 loops were mild. Of the 8
loops (5 controls.and 3 infected) 4 were from animal-Pairs 3 (194 and
195) and 5 (283 and 284). Little or no edema or cellular invasion.was
noted. The epithelium in both control and infected loops was markedly

vacuolate (Figure 12).

Small Intestine.5 cm Posterior to Perfused Logp. The microsc0pic appear-

 

ance of the intestine at this level was similar to that of the sections
5 cm anterior to the loop.‘ In the control animals, the epithelium*wasu

markedly-vacuolate but intact. The most consistent finding was moderate

45

\J' ..
3»?

\
' k.-
”e"-‘.' 1

V, ‘
‘\\‘\
"I

v. .
p» .'-';'
.- .3-

white
“‘3" .IIvf‘w”
..- v M‘ I-h-~ »
~ '“'n’]“
‘fl- 9' \
«n.

(,-

t
1),“;

x
\

“.QMa'p' ‘
If.“‘
.ntns

«fr
.35.),
r
l
| ,‘
I
' t

.‘ EB.
-” 0..
' l.‘ I'

‘0‘-

._:_} '
a"! "l -

 

5
x
2

 

 

Figure 9. Pig.287 (control). Perfused loop. Note,
the marked epithelial vacuolization (V) and submucosal
edema (E). There is.increased cellularity of the sub-
mucosa., The epithelium is intact. The villi are long
and_slender. No blunting or shortening is evident.
Hematoxylin and eosin; x 60.

46

 

 

Figure 10. Pig 286 (infected). Perfused loop.
Vacuolate epithelium is seen. Note the infiltration
of the lamina propria with neutrophils (N). Hematoxylin
and eosin; x 600.

47

 

 

 

Figure 11. Pig 287 (control). Perfused loop.
Marked fibrinous serositis is present (S). Hyperemia
(H) and cellular infiltration are apparent (I).- Muscu-

laris mucosa is designated (M). Hematoxylin and eosin;
x.60.

48

 

 

Note that the

Perfused loop.

Pig 197 (infected).
Although epithelial vacuolization is still extreme, only
Hematoxylin and eosin; x 60.

slight submucosal edema is present (E).

Figure 12.
villi are~intact.’

49
to marked submucosal edema. Cellular infiltration, predominantly neutro-
phils with lymphocytes and plasma-cells, was noted in most of the control
animals. In 5 of the controls, mild fibrinous serositis was present at
the mesenteric attachment.

The above histologic observations were also made in the infected
intestines. Even though the changes were somewhat more pronounced, the
difference again was not striking. One.of the infected segments (Pig
196) had undergone extensive change. There was acute necrotic enteritis
with neutrophilic exudation, hemorrhage,.and marked submucosal inflamma-

tory,edema.‘ The epithelium had been severely damaged.

Small Intestine 5 cm.Anterior to the Ileocecal Valve. Similar histologic
changes were noted between infected and control animals. Villi were long
and slender.‘ Epithelial vacuolization varied from moderate to extreme.
Varying degrees of submucosal edema and cellularity were present.‘ Two
sections (Pigs 289 and 196) had extensive lesions.. There was acute
necrotic enteritis with nearly complete destruction of the epithelium.
Hemorrhage and inflammatory.exudate were present in the intestinal lumen
(Figure 13). In Pig 194, discrete areas of necrosis were found in.the

mucosa e ‘

Liver. The_only consistent change seen.in the liver was the presence of
moderate to severe centrolobular vacuolization of the hepatic_parenchymal
cells. These vacuoles were negative for both fat and glycogen when-
stained with special stains. Extramedullary hematopoiesis was noted in
several tissue sections. The above observations were made in both the

germfree and E. coli-infected animals.

 

50

 

 

Figure 13. Pig.289 (control). Small intestine 5
cm anterior to.the ileocecal valve. Acute.necrotic.
enteritis is evident, with marked hemorrhage (H) and
destruction_of the epithelium (D). The submucosa is
edematous (E), with marked hyperemia (M). Note the
presence of lymphoid tissue (L). Hematoxylin and eosin;
x 60.

51

Kidney. No lesions were found in the renal cortex of either experimental
group. In 3 instances, hydropic changes in the collecting tubules were

notedin monocontaminated pigs only.

DISCUSSION

General Comments

 

Procurement and Rearing'of Gnotobiotic Pigs, The hysterotomy.technique
used in this research project was quite satisfactory. Survival rate was
100%. Isolation procedures and rearing methods provided and maintained

a germfree environment during the experimental period.

Experimental TeChniques. The perfusion method of studying intestinal
absorption provided a suitable means of measuring the in viva absorption.
or production of solutes and fluid. However, the actual placement of

the loop--i.e., the distance of the posterior end of the loop from the
ileocecal valve-dwas difficult to standardize from pig to pig. It was
hoped that,.using the measuring technique described by Serebro et'aZ.
(1968),.the loop would be placed in the jejunum approximately 3 feet
anterior to the ileocecal valve. The mean distance of the perfused loops.
from the ileocecal valve in the present research project was approxié
mately 3 feet. However, the range was considerable--2.l feet to 4.7
feet. The Serebro et‘al. (1968) method of measuring intestinal length.
was not consistently reliable. In addition, the mean intestinal length
was about 15 feet, placing the loop in the last 1/5 of the intestine.
According to Newey (1967) maximum.g1ucose absorption occurs in the upper
small intestine. Swallow et-aZ. (1968) found that the canine proximal

jejunum was 10 times as sensitive to cholera toxin as the distal ileum.

52

53
Bywater (1970) found no difference between the ileum and jejunum.in his
studies on toxin-induced colibacillosis in the calf. Nielsen at al.
(1968) stated that the posterior segment (ileum) of the small intestine
was resistant to the effects of enterotoxin because ileal transport proces-
ses were more efficient than those in-the more cephalad portion of the
intestine. Smith and Halls (1967) found the last 1 to 3 meters (3 to
9 feet) unreactive to E. coli organisms._ They-suggested that the last.
10 feet and the first 3 feet of the small intestine should not be used
for studying the effects of E. coli.

The use of heat lamps in Litter 1 to maintain body temperature wasg
inadequate. The skin directly beneath the lamps often became parched
while.body,temperature remained below 99 F. The plasticwcoated heating
pads placed under the pigs in Litter 2.were much more satisfactory in.

maintaining body temperature.

Hematology

According to the extensive literature review compiled by Calhoun.
and Smith (1970) the preinfectipn hematologic values obtained from the
germfree pigs were all within normal ranges for conventional pigs of the
same age. The fact that the lZ-day-old pigs had-consistently higher.
hemoglobin and packed cell volume values than the 8-day-old pigs was-
supported by the findings of Britt (1967). None of these animals had
evidence of anemia.

A marked leukocytosis with a regenerative left shift was the most
characteristic finding in the E. coli-infected animals. Some evidence
of hemoconcentration was found in the-monocontaminated animals when com-

pared to controls.

54
Intestinal Absorption

Although the intestinal fluid in.the infected animals-tended to~
be.more acid than.in the controls, no statistically significant.differ-
ences were found in the pH values of the duodenum, isolated loop, and
terminal ileum.between the monocontaminated and germfree animals.u Smith
and Jones (1963) found no difference in the pH values of intestinal chyme.
between healthy and E. cold-infected baby pigs raised conventionally.
However, Kohler (1968) and Kohler and Cross (1969) reported that fecal
diarrheal fluid produced in response to E. coli‘enterotoxin was signifi—
cantly-more alkaline than in-controls. Fecal pH was not measured in the
present.study.

It was originally proposed that PSP would be used as a nonabsorbable
marker. Net fluid movement could be calculated from the PSP concentra-
tion.(Serebro et al., 1968). Due to the extreme turbidity of the per-
fusion solution, consistent and reliable results could not be obtained.
Determdnation of PSP concentration, therefore, was not done. According
to Serebro et a2. (1968) actual measurement of fluid remaining in the~
reservoir with a graduated cylinder correlated within 1 to 2 ml of the
calculated fluid movement using PSP concentrations.‘

In the present research project, the absorption of d-glucose in
,E, golfihinfected animals was significantly greater (P<.01) than in the,
germfree controls. Serebro_et a1. (1968) found an equal and even slightly
increased absorption of d-glucose in rabbits with.toxin-induced cholera.
when compared to controls. Bywater (1970) reported no difference in
gflnucose absorption in Thiry-Vella loops between .. control calves andcalves
with toxin-induced colibacillosis. Carpenter at al. (1968) had similar

results using Thiry-Vella loops in dogs with toxin-induced cholera.

55
Torres-Pinedo et a1. (1966), however, found impaired glucose absorption
in human infants with colibacillary.diarrhea.

Solute movement-ei.e., glucose, sodium, and, to a lesser extent,
chloride correlated with fluid movement. That is, the control loops
consistently absorbed less glucose and produced more fluid, sodium, and
chloride than the E. deli-infected loOps. It had been postulated that_
the control loops would absorb more glucose and electrolytes and produce
less fluid than the infected loops.. Bywater (1970) found a marked secre-
tion of sodium, chloride, and fluid into the lumen of Thiry-Vella lOOps
subjected to E. coli enterotoxin. Carpenter et a1. (1968) and Swallow
et al. (1968) in experiments with-toxin-induced cholera in dogs found
that copious amounts of isotonic fluid were.produced in the infected
animals. However, they found that the addition of glucose to the per—
fusion fluid enhanced the absorption of fluid in both control and
infected loops. In other words, the presence of intraluminal glucose
markedly decreased the isotonic fluid production in the cholera loops.
Serebro et al. (1968) reported similar findings. Infected loops, however,
still contained more fluid than control loops (Carpenter_et al., 1968;
Swallow at aZ., 1968; Serebro et-aZ., 1968). Parsons (1967) concluded-
that the presence of intraluminal glucose stimulated water absorption.
Moreover, water absorption does not occur without simultaneous solute
absorption.‘ The fact that potassium movement-was inconsistent and net
significant was also found by Bywater (1970).'

Questions were then raised in interpreting the results. First, why
did control animals absorb less glucose than did the infected animals?
Secondly, why did the_E- OOZi-contaminated loops absorb fluid while the
control loops actually produced fluid and electrolytes?‘ The possibility

that the E; 001i organisms.themselves utilized glucose from the perfusion

56
fluid should be considered as one explanation as to why the infected pigs
absorbed more glucose than the germfree controls. Perhaps the E. coli-
infected intestine behaves in a manner that more closely approximates a.
natural physiological state and the absorptive mechanisms of the germfree
gut are rather prenatal, undifferentiated.and not as efficient (Dubos.
et aZu, 1963). Therefore, the presence of intraluminal glucose stimulated‘
and enhanced the absorption of fluid and.electrolytes to a greater degree
in the E. coZi-infected gut than in.the germfree gut. Levin (1967)
reported that increased blood flow will cause increased absorption.
Substances such as glucose can be stered in or metabolized by the intes-
tinal epithelial-cells. Moreover, Dowling (1967) stated that glucose
utilization and metabolism.increased in.intestinal hypertrophy and in
devitalized cells. Several other factors may have had an influence on
absorption results. Perfusion was carried out under artificial condi-
tions. Perhaps flow rate should have been slower to allow more absorpv
tion. The amount of perfusion fluid should have been greater in order
to maintain initial solute concentration and pH. That is, the volume of
the reservoir should.be increased from 100 to 250 or 400 ml. PSP should
be removed from the solution. PSP determinations are-not feasible and.
the substance may be irritating especially to the mucosa of the germfree
gut. Moreover, the loops should be placed more anterior to the ileo-
cecal valve--at least 6 feet. Also, additional studies should be cone
ducted in which glucose would be.omitted from the perfusion fluid in l

litter of pigs.

Histopathology_

 

The state of inflammation and edema observed in the perfused loops

did not-correlate with the amount-of fluid production or the amount of

57
glucose absorbed. For example, in animal Pair 3, the histologic.lesions
seen in both animals were mild with little or no edema or inflammation
present. The control pig produced 480 ul/cm/4 hr of fluid and absorbed
24.1-uM/cm/4 hr of glucose. In contrast, the monocontaminated pig pro-
duced 100 ul/cmdé hr of fluid but absorbed only 18.2 uM/cm/4 hr of glucose.
In animal_Pair 5, which~had~lesions‘very similar to Pair 3, different
results were obtained. The control loop absorbed 80 ul/cm/4 hr of fluid
while the infected loop absorbed 30 ul/cm/4-hr. However, only 33.2 uM/
cm/4 hr of glucose were absorbed by the control pig but 56.9 uM/Cm/é‘hrr
of glucose were absorbed by the monocontamdnated gut.

Severe histologic changes were seen in Pigs 285-and 286 (Pair 6).
Marked,submucosa1 edema was present with cellular infiltration and
hyperemia. Still, the infected pig absorbed 70 ul/cm of fluid and 66.7
uM/cm.of glucose while the control pig produced 30 ul/cmaof'fluid and
absorbed only 36.7 uM/cm of glucose during the 4-hour period. Regardless
of whether the histologic changes.were severe or mild, the infected
animals consistently absorbed more fluid, sodium, and glucose than did

the . control animals :

Infected. Control

Mean glucose absorption (MM/cm/4 hr) 55.1 28.4'
*

Mean fluid movement (ml/cm/é hr) 70.0 -46.0

Mean sodium absorption (qu/cm/4 hr) 10.0 -68.0

 

*
Negative number indicates production.
The submucosal edema observed in both germfree and infected intestines

may have been related to handling of the intestines during surgery.

58

Speculations on the Pathogenesis of Colibacillary Diarrhea

 

The results of this research project suggest that intestinal.malab-
sorption is.not a primary feature of colibacillary diarrhea in Zdweek-old
gnotobiotic pigs. Glucose, sodium, chloride, and fluid absorption in
E. deli-infected loops was.significantly greater than in the germfree
controls. Turk and Stephens (1966) found that the absorption rate of
zinc was significantly higher in chickens infected with coccidiosis-than
in control birds, especially during the first day after exposure. They
concluded that mild intestinal damage with slight-inflammation increased
absorption while severe_damage and hemorrhage decreased or stopped
absorption.- In addition, increased.cellular,permeabilittho~certain
substances occurred as a result of inflammation due to infection. It is
possible that in the-present study intestinal inflammation in the pres-
ence of E. coli organisms.stimulated the absorption of fluid, sodium.
and glucose in isolated perfused loops. The germfree gut, however,
with similar histologic changes did not.have comparable absorptive

capacity.

SUMMARY

Eighteen 2-week—old hysterotomy-derived gnotobiotic pigs from 2
litters were used to study the effects of colibacillary diarrhea on the
absorption of d-glucose, sodium, chloride,-potassium, and fluid. The
animals were paired and l pig from each pair was exposed per as to
3.x 106 Escherichia coli 0138:K81:NM organisms. Twenty-four.hours
after infection, isolated jejunal loops were prepared in each animal
for in'vivo measurement of nutrient absorption. These loops were con-
tinuously perfused for a period of 4 hours with a solution of 26 mM=
glucose in lactated Ringer's solution.

Glucose, sodium, chloride, and fluid absorption was not impaired in
the E. coZi-infected loops. The monocontaminated loops consistently.
absorbed more glucose and produced less_fluid than the germfree control
loops. Mean glucose absorption (“M/cm/é hr) was 55.1 in the infected-
animals but only 28.4 in the controls. Fluid production in.the germfree
control loops averaged 0.46 ml/cm/4 hr while the infected loops actually '
absorbed 0.07 pl/cm/4 -hr. Sodium movement directly correlated with fluid
production (P<.001). Chloride movement generally followed a similar
pattern. The germfree intestinal-loop.secreted 68 qu/cm_of sodium and
41 qu/cm.of chloride during the 4—hour period. The E. coZi-contaminated
loops absorbed 10 and 7 qu/cm/4 hr of sodium and chloride, respectively.

Differences in potassium secretion were not significant.

59

60

The most consistent hematologic findings in the pigs exposed to
E. coli were a marked leukocytosis and regenerative shift to the left.
Some evidence of dehydration was present in infected animals.

The histologic appearance of the duodenum, the intestine anterior
and posterior to the loop, and the ileum did not differ between control
and infected animals. Varying degrees of epithelial vacuolization,.
submucosal edema, and cellular infiltration were noted. Similar changes
were also seen in the isolated loops. The severity of the lesions did
not correlate with changes in absorption.

The findings in the present investigation did not support the theory
that malabsorption is'a prominent feature in the pathogenesis of coli-

bacillary-diarrhea.

BIBLIOGRAPHY

BIBLIOGRAPHY

Alexander, T. J. L., Miniats, O. P., Ingram, G., Thomson, R. G., and
' Thackeray, E.-L.; Gnotobiotic pigs: Procurement, microbial
flora, serum.proteins, and lymphatic tissues. Can. Vet. J., 10,
(1969): 98-105.‘

Barnum, D. A., -Glantz, P. J., and Moon, R. W. :- Colibacillosis. ‘Ciba -
Vet. Monogr. Ser., Ciba Found., Summit, N.J. (1967).

Benjamin,.M..M.: Outline.0f’Veterinary Clinical Pathology, 2nd ed.,
The Iowa State university Press, Ames, Iowa (1961).

Benyajati, C.: Experimental cholera in humans. Brit. Med. J., 1, (1966):

Bette, A. 0., Lamont, P. H., and Littlewort, M. C. G.:- The production
by hysterectomy of pathogenefree, colostrum-deprived pigs and
the foundation of a minimal disease herd. Vet. Rec., 72, (1960):
461-468.

Britt, A. L.: Pathsloyic Effects ofEscherichia colic 083:K:NM in Gnoto-
biotic Pigs. Ph.D. Thesis, Michigan State University (1967).

Burrows, W., and Musteikis, G. M.:‘ Cholera infection and toxin in the
rabbit ileal loop. J. Infect. Dis., 116, (1966): 183-190.~

Bywater,.R. J.: Some effects of Escherichia coli enterotoxin on net
fluid, glucose and electrolyte transfer in calf small intestine.
' J. Comp. Path., 80, (1970): 565-573.

CalhOun, L. M., and Smith, E. M.: Hematology and hematopoietic organs,
in Diseases of'Swine, 3rd ed., Howard W. Dunne, ed., The Iowa
State University Press, Ames, Iowa (1970).

Carpenter, C. C. J., Sack, R. B., Feely, J. G., and Steenberg, R. W.:.
Site and characteristics of electrolyte loss and effect of intra-
luminal glucose in experimental canine cholera. J. Clin. Invest.,
47, (1968): 1210-1220;

Christie, B..R.: Experimental Colibacillosis in GWotobibtic Pigs. M.S.
Thesis, Michigan State University (1967).

Crane, R. K.: Intestinal absorption of sugars. Physiol..Rev., 40,
’ (1960): 789-825.

61

62

’ Cross, R. F., and Kohler, E. M.: Autolytic changes in the digestive
system of germfree, Escherichia coli monocontaminated, and con-
ventional baby pigs. Can. J. Comp. Med. Vet., 33, (1969): 108-112.

De, S. M., and Chatterje, D. N.: An experimental study of the mechanisms
of action of Vibrio cholerae on the intestinal mucous membrane.
J.-Path. Bact., 66, (1953): 559-562.

Dowling, Rn R.:» Compensatory.changes in intestinal absorption. Br.
Med. Bull., 23, (1967): 275-278.

Dubos, R., Schaedler, R. W., and Costello, R.: Composition, alteration:
and effects of the intestinal flora. Fed. Proc., 22, (1963):
1322-1329.

Dunne, H. W., and Bennett, P. C.: Colibacillosis and edema disease, in.
Lfiseases of'Swine, 3rd ed., Howard W. Dunne, ed., The Iowa State
University Press, Ames, Iowa (1970).

Feteris, W. A.:- A serum glucose method without protein precipitation.
Am. J. Med. Tech., 31, (1965): 17-21.' 1

Fisher, E. W.:{ Death in neonatal calf diarrhoea. Brit. Vet. J., 121,
(1965): 132-138.

Fordtran, J. S.:* Speculations on the pathogenesis of diarrhea. Fed.
Proc., 26, (1967): 1405—1414.

Gilka, R.: A comparative study on.the pathological picture of trans-
' missible gastro-enteritis (TGE) and E. coli gastro-enteritis.in
newborn piglets with special reference to the gastroeintestinal

tract. Acta Vet. Acad. Sci.‘Hung-tomus., 18, (1968): 39-50.

Gossling, J., and Rhoades, H. E.: Serologic types of Escherichia coli
isolated from certain pigs with.enteric disorders. Cornell Vet.,
56, (1966):.345-353.

Goulden, C. H.: lMethods of'Statistical Analysis, 2nd ed., John Wiley '
Q Sons, Inc., New York (1952).

Gradwohl, R. B. R.:; Clinical Laboratory Methods ananiagnosis, Vol. 1,
C. V. Mosby Co., St. Louis, Mo. (1956).

Hagen, 0., and Haugen, A.: Bacteriological and toxicological investiga-n
' tions on colienterotoxemia in pigs. Nord. Vet. Med., 15, (1963):
497.

Hepler, 0..E.: Manual of'Clinical LaboratorynMethods, 4th ed., Charles
‘ C. Thomas, Springfield, Ill. (1968).

Jacobs, F. A., and Luper, M.: Intestinal-absorption by perfusion in
' situ. J.vApp1. Phy., 11, (1957): 136-138.

Kenworthy, R., and Allen, W. D.: The significance of Escherichia coli
to the young pig. J. Comp. Path. & Therap., 76, (1966a): 31-44.

63

Kenworthy, R., and Allen, W. D.: Influence of diet and bacteria on small

intestinal.morphology;with specia1.reference to early weaning and
Escherichia coli. J. Comp. Path. & Therap., 76, (1966b): 291-296.-

Kenworthy, R., and Crabb, W. E.: The intestinal flora of young pigs with
reference to early weaning and scours. J. Comp. Path. & Therap.,
73, (1963): 215-225.

Kohler, E. M.:‘ Enterotoxic activity of filtrates of Escherichia coli in
young pigs. Am. J. Vet. Res., 29, (1968): 2263-2274.

Kohler, E. M.: Studies of Escherichia coli in gnotobiotic pigs. IV.
Comparison of enteropathogenic.and nonenteropathogenic strains.
Can. J. Comp. Med. Vet. Sci., 31, (1967a): 277-282.

Kohler,.E. M.: Studies of Escherichia coli in gnotobiotic pigs. V.
Evaluation of the effects of oral and parenteral administration
of immune‘serum. Can. J. Comp. Med. Vet. Sci., 31, (1967b): 283-289.

Kohler, E. M., and Bohl, E. R.:: Studies of Escherichia coli in gnoto-~
biotic pigs. I. Experimental reproduction of colibacillosis.
Can. J. Comp. Med., 30, (1966): 199-203.

Kohler, E. M., and Cross, R. F.: Studies of Escherichia coli in gnotoe
biotic,pigs.l VI. Effects of feeding bacteria-free filtrates of
broth cultures. Can. J. Comp. Med. Vet. Sci., 33, (1969): 173-177.

Lecce, J. G., and Reep, B. R.: Escherichia coli associated.with colostrum-
' free neonatal pigs raised in isolation.‘ J. Exp. Med., 115, (1962):
491-501.‘

Leitch, G. J., and Burrows, W.:‘ Experimental cholera in the rabbit'
ligated intestine: Ion and water accumulation in.the duodenum,
ileum, and colon. J. Infect. Dis., 118, (1968): 349-359.

Leitch, G. J., Burrows, W., and Stolle, L. C.:' Experimental cholera in
the rabbit intestinal loop: Fluid accumulation and sodium pump
inhibition.. J. Infect. Dis., 117, (1967): 197-202.

Leitch, G. J., Iwert, M. E., and Burrows, W.: Experimental cholera in
the rabbit-ligated ileal loop: Toxin-induced water and ion move—
ment. ‘ J. Infect, Dis. , 116, (1966): 303-312.

Leman, A.-D., ed.: swine Health: common Diseases Affecting Baby Pigs--
‘ Research Review. College of Vet. Med., University of Illinois,
Urbana-Champaign, (Oct. 1970): 18-33. ‘

Levin, R. J.:, Techniques, terminology, and parameters in intestinal
absorption. Br. Med. Bull., 23, (1967): 209‘212.

Linderman, B., and Solomon, A. R., as cited by Smythe, D. H., and Whitman,
R.: Membrane transport in relation to intestinal absorption.
Br. Med. Bull., 23, (1967): 231*235.

64

Medway, W., Prier, J. E., and Wilkinson, J. 8.: A Textbook of'Veterinary
Clinical Pathology, The Williams and Wilkins Co., Baltimore, Md.'
(1969).'

Meyer, R. -C., Bohl, E. H., Henthorne, R. D., Tharp, V. -L., andtBaldwin,
D. E.: The procurement and rearing of gnotobiotic swine. Lab.
Animal Care, 13, (1963): 655.

Meyer, R. G., Bohl, E. H., and Kohler, E. M.: Procurement and maintenance
of germefree swine for microbiological investigations. Appl. Micro-
biol., 12, (1964): 295-300.’

Miniats, O. F.: Response of gnotobiotic pigs to oral infection with an
enterOpathogenic and with an edema disease strain of Escherichia
coli. Can. Vet. J., 11, (1970): 52-56.

Moon, H. W., Nielsen, N. 0., and Kramer, T. T.:, Experimental enteric
colibacillosis of the newborn pig: Histopathology of the small_
intestine and changes in plasma electrolytes. Am. J. Vet. Res.,
31, (1970a): 103-112.

Moon, H. W., Sorensen, D. K., and Sautter, J. H.: Escherichia coli
infection of the ligated intestinal loop of the newborn pig.
Am. J. Vet. Res., 27, (1966a):'l317-1325.

Mbon, H. W.,-Sorensen, D. K., Sautter, J; H., and Higbee, J. M;:. Associ-
ation-of Escherichia coli with diarrheal disease of the.newborn.
Am. J. Vet. Res., 27, (1966b): 1007-1011.

Moon, H. W., Whipp, S. C., Enstrom, G. W., and Baetz, A. L.: Response
of the rabbit ileal loop to cell free products from.Eecherichia.
coli enteropathogenic for swine. J. Infect. Dis., 121, (1970b):-
182-187.

Newey, H.: Absorption of carbohydrates. Br. Med. Bull., 23, (1967):
236-240.

Nielsen, N. 0., Moon, H. W., and Roe, W. E.: Enteric colibacillosis in
swine. J.A.V.M.A., 153, (1968): 1590-1606.

Nielsen, N. 0., and Sautter, J. R.: Infection of ligated intestinal.
loops with hemolytic Escherichia-coli in the pig. Can. Vet. J.,.
9, (1968): 90-97.

Norris, H. T., Schultz, T., Stanley, G., Curran, P.-F., and Finkelstein,
Ru A.: Active sodium transport across rabbit ileum in experi-
mental _cholera induced 'by choleragen. J. Infect. Dis. , 117-,
(1967): 193-196. .

‘9

Norris, H. T., and Majno,.G.:. Observations on the pathogenesis of
experimental human cholera in the rabbit.' Am. J. Path., 50,
(1967): 32a.

65

Norris, H.-T., and Majno, G.: .On the role of the ileal epithelium in.
the pathogenesis of experimental cholera. Am. J.'Path., 53,
(1968): 263-279.

Parsons, D. S.:, Salt and water absorption by the intestinal tract.
Br. Med. Bull., 23, (1967): 252-257.

Saunders, C.-N., Stevens, A. J., Spence, J. B., and Sojka, W. J.:
Escherichia coli infection in piglets.~ Res. Vet. Sci., 1, (1960):.
28-35 0

Saunders, C. N., Stevens, A. J., Spence, J. B., and.Betts, A. 0.:
Escherichia coli infection: Reproduction of the disease—in‘
'pathogen-free' piglets.. Res. Vet. SCi., 4, (1963): 347-357.

Schales, 0.,.and Schales, S. S.:' A simple method for determination of
chloride in biological fluids. J. Biol. Chem., 140, (1941):
879-885.

Schedl, H. P., and Clifton, J. A.:r Small intestinal absorption of
steroids. Gastroenterol., 41, (1961): 491-494.

Serebro, H. A., Bayless, T. M., Hendrix, T. R., Iber, F. L., and McGonagle,
T.: Absorption of d—glucose by the rabbit jejunum during cholera
toxin-induced diarrhea. Nature, 217, (1968): 1272-1273.

Smith, H. W.,.and Halls, S.:A Observations by the ligated intestinal
segment and oral inoculation methods on Escherichia coli infection
in pigs, calves, lambs, and rabbits. J. Path. Bact., 93, (1967):

Smith, H. W., and Jones, J. E. T.: Observations on the_alimentary tract
and its bacterial flora in healthy and diseased pigs. J. Path.
Bact., 86, (1963): 387-412.

Staley, T. E., Jones, E. W., and Marshall, A. E.: The jejunal absorptive
cell of the newborn pig: An electron microscopic study. Anat.
Rec., 161, (1968): 497-515.

Swallow, J. H., Code, C. F., and Freter, R.: Effect of cholera toxin
on water and ion fluxes in the canine bowel. Gastroenterol.,
54, (1968): 35-40.

Torres-Pinedo, R., Rivera, C. L., Fernandez, S., and Maldonado, M.:
Studies on infant diarrhea. II. Absorption of glucose and net
fluxes of water and sodium chloride in a segment of the jejunum.
J. Clin. Invest., 45, (1966): 1916-1922.‘

Truszcznski,.M., and Pilaszek, J.: ,Effects of infection of enterotoxin,
endotoxin or live culture of-Escherichia coli into small intestine
of pigs. Res. Vet. SCi., 10, (1969): 469-476.

Turk,-D. E.,.and Stephens, J. F.: Effect of intestinal damage produced
by Eimeria necatrir infection in chickens upon absorption of
orally administered zinc-65. J. Nutr., 88, (1966): 261-266.

66

Waxler, G. L.: Gnotobiotic pigs, in Diseases of'Swine, 3rd ed., Howard-
W. Dunne, ed., The Iowa State University Press, Ames, Iowa (1970).

Waxler, G4 L., Schmidt, D. A., and Whitehair, C. K.: .Technique for
rearing gnotobiotic pigs. Am. J. Vet. Res., 27, (1966): 300-307.

VITA

The author was born August 11, 1944, in Detroit, Michigan. She
was graduated from Plymouth High School, Plymouth, Michigan, in June,
1962. After attending the University of Michigan for 2 years, the author
transferred to Michigan State University in 1964. She received her-
Doctor of Veterinary Medicine degree in June of 1968. On August 31,
1968, the author was married to Philip Lee Hageman, who is currently
in the United States Army. The author began her-degree program in
September of 1968 while serving as a graduate assistant in the Department
of;Pathology.‘ From 1969 to 1971, the author has been employed in a.
mixed practice with her father, Dr. W. E. Lickfeldt, of Fenton, Michigan,
while completing the requirements for a Master of Science degree in

pathology.

67

1293 03103 9286

ll
l
l
l
l
I
l
l
I
l
l
l
l
l
l
l
l
|
|
l
l
l
Ill
l
l
l
l
l
l
l
Ill
l
l
l
l
l
l
||l
l
l
l
l
l
l

3