EXPERIMENTAL ENTERIC COLIBACILLOSIS IN GNOTOBIOTIC SWINE UTILIZING- THE LIGATED LOOP TECHNIQUE Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY JAMES P. DAVIDSON 1974 y lam—Inna av "0“ G SQNS' BUUK BINDERV INC. I BRA {Y BINDERS mm; r-» u ABSTRACT EXPERIMENTAL ENTERIC COLIBACILLOSIS IN GNOTOBIOTIC SWINE UTILIZING THE LIGATED LOOP TECHNIQUE By James P. Davidson The ligated loop technique was evaluated in 23 gnotobiotic pigs, 3 In) 4 weeks of age, in an attempt to define the nature of false positive loops. Five serotypes, encompassing 6 strains of Escherichia coli, were utilized in this experiment. The 6 strains were divided into 3 subgroups: 1) strains enteropathogenic for pigs regardless of age, 2) strains enteropathogenic for pigs under 2 weeks of age and nonenteropathogenic for pigs over 6 weeks of age, and 3) nonentero- pathogenic strains. Each-experimental loop was inoculated with a single strain, noninjected interloops separated each inoculated loop, and all 6 strains were tested in each of 15 experimental animals. Strain enteropathogenicity in gnotobiotic pigs, based upon sig- nificant visual loop distention, compared favorably with results obtained utilizing the same technique and strains of E. coli in conventional animals with the exception that the gnotobiotic jejunal loop may be somewhat less sensitive to certain strains. In addition, postoperative mortality rates among gnotobiotic pigs were higher than reported in conventional pigs. Light microscopic intestinal lesions in these gnotobiotic pigs were similar to those reported in the conventional pig. Mucosal James P. Davidson erosions and ulcerations were often seen in maximally distended loops. Two false positive loops occurred during the experiment. Both were in sacs inoculated with strain 115. False positive loops were not seen in other sacs, whether inoculated or not. Results of these experiments support the hypothesis that false positive loops in conven- tional pigs arise from spontaneous infection. EXPERIMENTAL ENTERIC COLIBACILLOSIS IN GNOTOBIOTIC SWINE UTILIZING THE LIGATED LOOP TECHNIQUE By James P. Davidson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Pathology 1974 To my wife Ann and our son Marc ii ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to his major professor, Dr. Glenn L. Waxler, for his expert and meaningful guidance, encouragement and enduring patience throughout the course of this investigation. My sincere thanks go to my academic committee: Drs. S. D. Sleight and V. L. Sanger of the Department of Pathology, for their friendly help and advice in understanding the complexities of experimental pathology as they related to this research problem; Dr. D. J. Ellis of the Department of Large Animal Surgery and Medicine, for his valuable counsel and suggestions within the realm of clinical and practical aspects of the disease; and Dr. C. C. Morrill, former Chairman of the Department of Pathology, for providing the facilities and supplies for this research. I am also indebted to the following persons for their expert technical assistance: Mr. James Southern, animal caretaker at the Veterinary Research Farm; Mrs. Dottie Fenner, Mr. Charles Bares and Mr. Warren Taylor, medical technologists; and Mrs. Nina Miller, Mrs. Mae Sunderlin and Mrs. Frances Whipple, histopathologic technicians. This research was supported, in part, by a grant from the Diamond Shamrock Corporation, Cleveland, Ohio. iii INTRODUCTION. TABLE OF CONTENTS LITERATURE REVIEW 0 C O O O O O O O O O C O O Colibacillosis in Neonatal Pigs. . . . The Ligated Loop Technique . . Clinical signs. . . . . . . . . Postmortem findings . . . . . . Histopathologic observations. . The germfree intestine . Colibacillosis . . . . . Diagnosis . . . . . . . . . . . Bacteriological considerations. Pathogenesis. . . . . . . . . . Immunological considerations. . The Gnotobiotic Pig. . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . Experimental Animals . . . . . . . . . Procurement . . . . . . . . . . Rearing O I I O O O O O O O O 0 Determination of sterility. . . Surgical manipulation and inoculation Control animals. . . . . Bacteriological Procedures . . . . . . Preparation of the infective agent. Culturing of inoculating syringes . Necropsy and Laboratory Procedures . . Histopathologic technique . . . System for Statistical Evaluation of Results RESULTS . . . Presurgical Observations . . . . . . . iv Page wOmVU'l-b-I-‘ww I—‘ l7 19 19 19 19 22 22 24 24 24 26 27 29 29 32 32 Surgical Observations. . . . . Clinical and Gross Findings. . Bacteriological Monitoring . . Histopathologic Findings . . . The small intestine . . The liver, kidney and mesenteric lymp Statistical evaluation of results DISCUSSION. 0 O O O C O O O O O O O O Postoperative Clinical Features. Gross Findings--The Peritoneal Reactivity of the Serotypes. . SUMMARY . . . . . . . . . . . . . . . BIBLIOGRAPHY. . . . . . . . . . . . . VITA. O O O O O O O I O I 0 O O I O O 11‘ Cavity. nodes. Page 32 32 4O 40 41 50 52 56 56 57 57 69 71 77 LIST OF TABLES Table Page 1 Age, sex and number of E2 coli-inoculated and gnotobiotic control pigs . . . . . . . . . . . . . . . . . 29. 2 Serotype, strain number and enteropathogenicity of the experimental organisms . . . . . . . . . . . . . . . . 25 3 Criteria utilized in statistical evaluation of strain enteropathogenicity . . . . . . . . . . . . . . . . 30 4 Reactions of serotypes of E. coli in experimental jejunal loops. . . . . . . . . . . . . . . . . . . . . . . 34 5 Summary of reactions of serotypes of E. coli tested in experimental jejunal loops. . . . . . . . . . . . . . . 38 6 Individual strain scores, strain means and standard error of the strain means. . . . . . . . . . . . . . . . . 53 7 Comparison of experimental and published data for the 6 strains of E. coli tested in porcine intestinal loops. 0 O O O O O O I I O O O O O O O O O O O O O O O O O 59 8 Loop distention patterns for animals harboring false positive loops . . . . . . . . . . . . . . . . . . . . . . 66 vi Figure LIST OF FIGURES Pig 9, interloop 3 (noninoculated). Jejunum, 24 hours postinoculation. Notice the presence and abundance of goblet cells (C) interspersed among the absorptive epithelial cells (A) of the mucosa. . . Pig 9, loop 4 (Strain 987). Jejunum, 24 hours postinoculation. Field similar to Figure 1. Notice the absence of goblet cells in the mucosa. . . . . . . Pig 11, loop 4 (Strain 123). Jejunum, 24 hours postinoculation. Notice the essentially normal configuration of the section, with the villi (V) being long, fingerlike projections of the mucosa. A mild peritonitis is seen on the serosal surface (S) Pig 14 (control), loop 4, inoculated with sterile broth. Jejunum, 24 hours postinoculation. Villi (V) are slender, fingerlike projections of the mucosa, extending into the lumen (L). Notice the moderate peritonitis (P) on serosal surface . . . . . . . . . . Pig 22, loop 3 (Strain 115). Jejunum, 24 hours postinoculation. Notice the extensive mucosal ulceration and erosion (M), shortened villi (V), and peritonitis (P) on the serosal surface . . . . . . Pig ll, loop 6 (Strain 987). Jejunum, 24 hours postinoculation. Notice the extensive ulceration and erosion of the mucosa (M) with remnants of villi (V) still present, severe edema (E) in the submucosa, and the stretching of the tunica muscularis (T) into a thin layer. Severe areas of peritonitis are seen on the serosal surface (S) . . . . . . . . . . . . . . Plot of mean and standard error of the mean for each strain. Taken from Table 6 . . . . . . . . . . . vii Page 43 44 45 46 48 51 55 INTRODUCTION Despite major medical advances during the last half century, disease characterized by diarrhea remains as a leading cause of death throughout many of our domestic species and among the peoples of the developing nations. The etiologic implication of Escherichia coli in certain of these syndromes has led to greater scientific efforts to understand the ubiquitous role of this organism, a normal constituent of the intestinal flora of vertebrates. Reliable procedures are now available to differentiate the enteropathogenic from the nonentero- pathogenic strains of E. coli; among these is the ligated loop technique. The utilization of the gnotobiotic pig in the study of colibacil— 10513 has provided important information toward the understanding of this complex enteric syndrome by allowing the study of a single patho- genic strain in the absence of antibody and commensal intestinal flora. In our experiments the ligated loop technique was utilized in the gnotobiotic animal in an attempt to define: 1. The comparative aspects of the ligated loop technique between the gnotobiotic and conventional pig. 2. The etiologic nature of false positive reactions encountered in the ligated loop technique. LITERATURE REVIEW Enteric disease associated with the enteropathogenic forms of Escherichia coli affects a broad spectrum of animal life including man. The original correlation associating the occurrence of disease, a diarrheal syndrome in calves, with the presence of E. coli came in 1891 with the work of Jensen (Sojka, 1965). Since that time the organism has been identified and implicated in neonatal diarrhea of lambs, goats, pigs, poultry and human beings (Barnum et aZ., 1967). Comparisons have been drawn between enteric colibacillosis in pigs and calves and the disease syndrome found in human infants (Saunders et al., 1960). Colibacillosis in Neonatal Pigs The term colibacillosis applies to a group of diseases caused by Escherichia coli. The enteric form of the disease occurs in three distinct age groups of pigs: neonatal (to 4 days of age), pre—weaning at 3 weeks, and at weaning (Stevens, 1963a,b). Enteric colibacillosis in suckling pigs is most commonly manifested as an acute enteritis or gastroenteritis, clinically, and the syndrome has therefore been referred to as baby pig scours, coliform enteritis, diarrhea neonatorum or neonatal colibacillary diarrhea (Leman, 1970). Stevens (1963a,b) estimated that E. coli infections were associ- ated with approximately 75% of all baby pig diarrheas. An English survey in 1960 indicated that of those pigs dying at 4 months of age 3 or younger, 40% had enteritis associated with E. coli (Kenworthy and Allen, 1966a). The morbidity of the disease is highly variable. Not all litters during a farrowing season will become infected, nor will all the pigs within an infected litter show signs of the disease (Barnum et al., 1967). Once an enzootic has begun in an establishment, the severity increases as more sows farrow (Barnum, 1971). Clinical signs. Enteric colibacillosis is most often diagnosed in pigs 1 to 8 days of age. Onset of the disease may be recognized as early as 12 hours postnatally in a few pigs in which a bacteremia and septi- cemia have occurred. Although diarrhea is an inconsistent sign in pigs so affected, death usually occurs within 48 hours. More typically, signs in infected pigs will begin with a whitish-yellow, watery diarrhea with ensuing dehydration, emaciation and a roughened hair coat. The tail and perineum commonly become pasted with fluid or semifluid feces, and may also become inflamed. In some cases the tip of the tail becomes necrotic and eventually sloughs. Mortality rates are usually higher in younger pigs. Pigs acquiring the disease within the first 3 days postnatally have a mortality rate close to 70% (Dunne and Bennett, 1970); however, the mortality rate is variable and seldom reaches 100% (Leman, 1970). Postmortem finding . Gross lesions of enteric colibacillosis have been inconsistent in pigs infected with known enteropathogenic strains. Kohler and Bohl (1966), using a group of experimentally infected gnotobiotic pigs, demonstrated that only about 50% of the necropsied animals had gross abnormalities. The lesions consisted of petechial hemorrhages on the adrenals, mesenteric lymph nodes and in the mesen- tery, with no evidence of inflammation or congestion in either the 4 small or large intestine. Intestinal contents were watery and yellow to pale green. Barnum et al. (1967) confirmed the absence of inflamma- tory lesions and observed evidence of dehydration, a full stomach and intestines filled with fluid and undigested food. Histopathologic observations The germfree intestine. Morphologically, certain organ systems of the germfree animal deviate significantly from those found in con- ventional animals. In the intestinal tract Dubos (1966) observed that the germfree animal had shorter crypts, longer villi and a paucity of lymphoid and inflammatory tissue in the epithelium and lamina propria. Staley et al. (1968) reported that the jejunal epithelium of the newborn germfree pig was composed of simple columnar epithelium which was highly vacuolate on the distal two-thirds of long, slender villi. Cross and Kohler (1969) compared the autolytic changes in the digestive systems of germfree, monocontaminated and conventional baby pigs. They observed an increasing degree of cellularity in the lamina propria and submucosa which was directly proportional to the diversity of the microbial population of the gut. Neutrophils, lymphocytes and macro- phages were increased in number and, for this reason, they applied the term "physiologic inflammation" to the microscopic appearance of a non— germfree intestine. The most striking microscopic changes seen in the intestine were found in the terminal ileum and colon. They reported the villi of germfree pigs to be longer and thinner when compared to E. coli monocontaminated and conventional pigs. No significant micro— scopic alterations were found by Alexander et al. (1969) in the duodenal sections from germfree pigs, except that eosinophils were numerous in the lamina propria and Peyer's patches of the jejunum and ileum. 5 In a study of gnotobiotic and germfree pigs, Kenworthy and Allen (1966b) found a direct relationship between the morphological character- istics of the intestine and the degree and variety of bacterial con- tamination. The villi of a gut contaminated with a single nonpathogenic strain of bacteria were long, slender and uniform, with little cellular infiltration, and closely resembled the villi of the germfree gut. Dual contamination of the gut with a nonenteropathogenic and an entero- pathogenic strain of E. coli altered the structure of the gut by producing villi that lacked uniformity and were more conical with flattened tips. In histological sections taken from intestines of pigs reared in the conventinal environment, the villi were shortened and many were found in a leaf-like configuration. Cellular infiltration into the lamina propria was extensive. Alexander et al. (1969) also examined microscopic tissue sections taken from the gnotobiotic pig. They reported numerous neutrophils and eosinophils in the medullary regions of the mesenteric lymph nodes and a sparse population of lymphocytes and reticuloendothelial cells in the same area. Colibacillosis. There are conflicting reports concerning the histologic appearance of intestinal sections taken from either gnoto- biotic or conventional pigs afflicted with colibacillosis. Histopatho- logic changes were inconsistently seen in conventionally reared animals afflicted with a diarrheal syndrome attributed to E. coli (Moon et aZ., 1970; Barnum et aZ., 1967). Occasionally villous atrophy or desquamated epithelium was found in the anterior aspect of the intestine and occurred only in pigs in which diarrhea had been observed (Moon et aZ., 1970). Barnum et al. (1967) stated that catarrhal enteritis or villous 6 atrophy may be more frequent in pigs with a history of prolonged diarrhea. Dunne and Bennett (1970) considered the microscopic changes of the gut taken from newborn piglets with colibacillosis to be similar to the reaction associated with low-grade irritants. Mucosal altera- tions consisted of enlarged goblet cells and distended, vacuolated absorptive cells. Gilka (1968) reported that colibacillosis produced an acute serous enteritis in the conventional newborn piglet. The edema was localized in the lamina propria and was accompanied by a neutrophilic infiltration into the villi. A mild neutrophilic infil- tration was observed by Kohler (1967) in the villous lamina propria of pigs infected with a single strain of enteropathogenic E. coli. Smith and Jones (1963) found no inflammatory changes in newborn piglets with colibacillosis and concluded that there was no histopatho- logic difference between these animals and their healthy littermates during the first week of life. Intestinal sections collected from neonatal colostrum-deprived pigs afflicted with colibacillosis were examined ultramicroscopically by Staley and co—workers (1969). In subclinical cases, fine structural alterations of the absorptive cell were observed primarily in the distal jejunum and ileum and were associated with the presence of the organism within the cell. The bacterium gained entrance into the absorptive cell by phagocytosis; this feature of the infection was thought to relate to the absence of lesions observable with the light microscope. The histopathologic changes seen in the gnotobiotic pig. also vary considerably. Kenworthy (1970) found an acute inflammatory response of the intestine following the inoculation of E. cOZi-Ol41. 7 Absorptive epithelium was shrunken, vesiculated and vacuolated. The distal aspect of these cells was often surrounded by neutrophils and was occasionally separated from the basement membrane. An acute exu- dative response was observed in the lamina propria both in animals whose death was attributable to the disease,as well as surviving animals in which clinical signs could no longer be detected. Necrotic and degenerative changes in the tunica media of the arteries and arterioles of the submucosa immediately below the involved areas of the absorptive epithelium suggested that mucosal lesions could be ascribed to ischemia. Only minimal lesions were seen in gnotobiotic pigs infected with an enteropathogenic strain of E. coli (Drees and waxler, 1970). These changes, consisting of edema of the lamina propria, dilation of the central lacteal, and an infrequent polymorphonuclear cellular infiltra- tion, were most prominent in the terminal jejunum and ileum. No inflammatory lesions were found in either the small or large intestine of gnotobiotic pigs infected with enteropathogenic E. coli (Kohler and Bohl, 1966). In an additional study by Kohler and Cross (1969), utilizing a cell-free filtrate of E. coli cultures, no histo- pathologic lesions were observed. Diaggosis. Because no single sign, lesion or laboratory test may be used alone to diagnose colibacillosis, it is necessary to evaluate all available epidemiological, clinical and laboratory data to arrive at a final diagnosis (Sojka, 1971). Many workers (Moon et aZ., 1970; Grun et al., 1967; Muylle and Oyaert, 1965) have studied the alterations in the blood stream of pigs afflicted with colibacillosis. However, these findings have not been as widely used diagnostically as those involving the characterization 8 of the organism. A presumptive diagnosis is frequently based upon the isolation of E. coli, in pure culture, from gastric and duodenal contents taken from animals with diarrhea prior to death (Leman, 1970). Enteric colibacillosis of preweanling pigs must be differentiated from transmissible gastroenteritis (TGE), enterotoxemias due to Clos- tridium perfringens type C, hog cholera, salmonellosis, strongyloidosis and noninfectious digestive disturbances (Barnum et al., 1967; Leman, 1970). Bacteriological considerations. Escherichia coli is a normal inhabitant of the intestine of vertebrates (Leman, 1970) and a common inhabitor of the environment in which animals are found. 0n farms, it may be commonly isolated from stables and pens and is known to cause diarrhea in neonates (Barnum et aZ., 1967). Because of the ubiquitous nature of the organism, and its seeming inconsistency to produce diarrhea, certain means of detecting entero- pathogenic strains of E. coli were developed to aid in the accurate diagnosis of the disease. The definition of enteropathogenicity among strains of E. coli evolved as techniques were developed to establish their identification. Moon and Whipp (1971) reserved the term "entero— pathogenic Escherichia coZi" (EEC) for those strains that could colonize the jejunum in numbers equal to their population in the colon. In addition, an abundant net movement of water and electrolytes across intact intestinal epithelium into the lumen must occur while the EEC are confined to the intestine. Gyles and Barnum (1967) stated that the two requirements for an organism's enteropathogenicity were epidemio- logical association and the ability to produce a positive gut loop reaction. 9 Hemolytic strains are commonly enteropathogenic; however, Sojka (1965) reported that this characteristic cannot be used as the sole criterion of pathogenicity. Hemolytic strains of E. coli may be non- enteropathogenic and not all enteropathogenic strains of the organism are concomitantly hemolytic. Escherichia coli may be further divided into serotypes based upon the demonstration of 3 of its many antigenic components. The somatic, "O", antigens are lipopolysaccharide and comprise 146 subgroups. The "K" antigens, associated with polysaccharides of the envelope or capsule, are of 3 types: L, B and A, and 91 have been recognized. However, recent evidence by ¢rskov et a2. (1971) questioned the validity of the subclassification of L, B and A, and the very existence of the "K" antigen itself. These workers suggested that the "K" antigen may eventually be included in the larger somatic, "0", antigen grouping. Forty-nine flagellar or "H" antigens have been identified and are thought to be protein (Sojka, 1971). A limited number of serotypes of E. coli have been demonstrated to be enteropathogenic in the ligated loop test. The inoculum may contain organisms or be in the form of a cell-free supernatant from such cultures (Sojka, 1971; Barnum et al., 1967; Kohler and Bohl, 1966; Kohler, 1968). Because the serotyping technique has been widely employed, entero- pathogenic strains of E. coli have been identified throughout the world wherever swine are produced (Kohler, 1972). Pathogenesis. Although many phases of the pathogenesis of colibacil- losis have been defined in recent years, our understanding of this complex disease remains incomplete. 10 Currently, the pathogenesis of colibacillosis is commonly accepted to involve "...a susceptible pig infected with an enteropathogenic strain of E. coli which has the capacity to proliferate in the proximal small intestine and to produce and release enterotoxins in adequate amounts to cause an alteration in the normal fluid and electrolyte trans- port functions of the intestine, with resultant diarrhea, dehydration and death." (Kohler, 1972) Environmental factors, reflecting conditions within the farrowing house, play an important part in the incidence of colibacillosis among susceptible piglets (Kohler, 1972). Arbuckle (1968) reported that the intestines of young pigs appear to support the proliferation of entero- pathogenic strains of E. coli more readily than does the digestive tract of adult swine, and this may account for the abundance of entero- pathogenic strains of the organism in farrowing houses utilized on a continuous basis. The importance of an adequate infective dose of the organism.was demonstrated by Kramer and Nderito (1967). Neonatal animals acquire intestinal E. coli as a result of inges- tion. This portal of entry is also utilized by the enteropathogenic strains as the primary route of infection (Leman, 1970). Reports indicate that entry may be gained via the blood stream as well. Christie (1967) demonstrated the production of diarrhea in gnotobiotic pigs by subcutaneous inoculation of an enteropathogenic strain of E. coli into the umbilical stump. Onset of diarrhea was delayed until the organisms had established themselves in the intestinal tract, 24 hours later. In the study of naturally occurring cases of colibacillosis in neonatal pigs, the constant site of infection has been the intestinal tract. Invasion and proliferation of enteropathogenic strains of E. coli in various organs outside of the gastrointestinal tract have been 11 observed in the terminal stages of the disease in colostrum-fed pigs (Stevens, 1963a). The ability of enteropathogenic strains to establish infection within the host is related to the susceptibility, or resistance, of the pig. Numerous host factors influence this susceptibility. Dietary stress is highly significant. The gestational diet of the dam not only influences the size, vigor and immediate postnatal health of her off— spring, but also affects the quantity and quality of the milk with which the newborn will be fed (Kohler, 1972). Smith and Jones (1963) reported that the high gastric pH of the newborn created an environ- ment conducive to the overwhelming propagation of ingested entero- pathogenic as well as benign strains of E. coli in the proximal small intestine. The relationship between the age of the host at the time of exposure and the ability of the organism to establish infection has also been stressed by many authors (Moon and Whipp, 1970; waxler et aZ., 1971). Colostrum consumption is probably the single most important factor in determining or altering the susceptibility of the newborn pig to enteropathogenic strains of E. coli. Although the antibodies provided in the colostrum are eventually absorbed from the intestine, and aid in the prevention of polyserositis and endotoxemic shock (Barnum et aZ., 1967), their primary protective function rests upon the direct intraluminal action upon the organism. Therefore, the quantity and adequacy of appropriate antibody within the intestinal lumen during the first day of neonatal life affords greater protection in the pre- vention of colibacillosis than does the antibody present in the serum (Kohler, 1967, 1969; Miniats, 1970). Although the aforementioned conditions are primal to the onset of colibacillosis, other factors contribute to the overall 12 pathogenesis of the disease. The influence of gastrointestinal stasis has yet to be fully explored. Nielsen and Sautter (1968) presented indirect evidence that lack of intestinal motility, and confinement of the organism in the ligated loop technique, were of importance in the expression of enteropathogenicity. White et al. (1969), in a radio- graphic study of specific pathogen-free (SPF) and conventional pigs, demonstrated that gastric stasis preceded diarrhea in suckling pigs. Enterotoxins and their relationship to the onset of colibacillosis have received extensive attention in recent years. Cell-free filtrates from enteropathogenic cultures of E. coli caused distention of ligated intestinal loops in a manner similar to that observed when viable organisms of the same strain were utilized (Smith and Halls, 1967a). These workers also reported that extracellular concentrations of the enterotoxins significantly exceeded the intracellular levels, and that the highest levels of the toxin were detected after 6 hours of initial incubation. Endotoxins from the same enteropathogenic strains produced no distention of ligated intestinal loops (Smith and Halls, 1967b). Initially the enterotoxin was characterized as heat stable (Smith and Halls, 1967b). However, Smith and Gyles (1970) suggested that, in addition, a heat labile enterotoxin was also elaborated by certain enteropathogenic E. coli. The production of both enterotoxins was demonstrated to be controlled by a plasmid (Smith and Gyles, 1969). This extranuclear genetic information could be transmitted to non- enteropathogenic E. coli, transforming them into enterotoxin producing (enteropathogenic) organisms (Smith and Halls, 1968; Smith and Gyles, 1969). The means by which enteropathogenic E. coli initiate diarrhea are not known. Kohler (1972) has stated that the heat-labile enterotoxin 13 of E. coli and the enterotoxin elaborated by Vibrio cholerae share a similar, if not identical, mode of action upon intestinal cells in the production of diarrhea. However, he suggested that extensive research remains to be done to confirm these comparisons. The enterotoxin of V. cholerae acts on the luminal surface of gut mucosal cells, affecting water and electrolyte transport (Pierce et al., 1971). The importance of increased concentrations of 3'5' adenosine monophosphate (cAMP) mediated through the action of V. cholerae enterotoxin has been stressed in several recent studies (Shafer et aZ., 1970; Kimberg et aZ., 1971; Sharp and Hynie, 1971). Increased concentrations of cAMP were demon- strated by Pierce et al. (1971) to result in changes in ion transport in gut mucosal cells, most notably the inhibition of active sodium absorption and stimulation of active chloride secretion. These ionic alterations resulted in net fluid accumulation within the intestinal lumen. The ensuing diarrhea results in dehydration, hemoconcentration and eventual acidosis. Passive congestion of various viscera has also been reported. The sequelae to colibacillosis is recovery or death with recovery being dependent upon the ability of the individual to respond positively and overcome the pathophysiologic stresses that have resulted from the disease (Kohler, 1972). Immunological considerations. Since the primal role of colostrum has been recognized in the prevention of colibacillosis, various studies have characterized the Specific antibody(ies) contained within the mammary secretion during the initial postpartum period. Wilson (1972) reported that protective antibodies responsible for the prevention of colibacillosis come from colostrum and milk. The l4 colostrum secretion of the immediate 24-hour postpartum period provides an initial surge of the immunoglobulins IgM and IgA, the latter being synthesized in the mammary gland itself (Porter, 1969a). Following this first supply of antibody, IgG and IgA are subsequently secreted and continue to be present in the milk for a longer period of time than the first group of immunoglobulins (Curtis and Bourne, 1971; Porter, 1969b). These antibodies remain within the lumen of the alimentary tract and act at the site of bacterial proliferation. Ultimately, they are digested or eliminated (Wilson, 1972). Wilson and Svendsen (1971) demonstrated that milk obtained from sows vaccinated with live, formalin- treated cultures of enteropathogenic E. coli afforded significant pro- tection against colibacillosis when compared with a similar milk diet obtained from unvaccinated sows. Oral administration of IgG isolated from milk obtained from vaccinated sows also provided protection against colibacillosis through the action of multiplication—inhibition and anti-enterotoxin factors (Wilson, 1972). Three hundred sows, kept under field conditions, were utilized in a well controlled immunization study conducted by Wilson. Results of this preliminary study indicated that immunization, and subsequent resistance of their offspring to colibacillosis following suckling, could be readily obtained through the utilization of homologous, autologous or heterologous vaccines of live, formalin-treated cultures of E. coli. The Ligated Loop Technique A convenient animal model for the study of human cholera was described by De and Chatterje (1953) in which sacs, created in the small intestine of rabbits, became distended with fluid when inoculated with vibrio organisms isolated from human patients. The loop technique was later adapted to study the effects of E. coli isolated from cases 15 of acute and chronic enteritis in man (De et al., 1956). Taylor at al. (1961) stated that the enteropathogenicity of E. coli could be determined by a positive loop reaction (distention), and that entero- pathogenicity was not always related to general pathogenicity of the organism. Smith and Halls (1967a) reported that rabbit intestinal loops inconsistently distended with fluid when known enteropathogenic E. coli, isolated from other species, were inoculated intraluminally. They reported, however, that positive loop reactions were almost always observed in the host test animal for which the organism was entero- pathogenic. The intestine of conventionally raised swine had been previously utilized, first by Namioka and Murata (1962) and subsequently by Moon et a1. (1966) and Nielsen and Sautter (1968) to study the enteropathogenicity of E. coli isolated from pigs. As evidenced in a recent review article by Sojka (1971), the ligated loop technique has gained general acceptance as an effective research tool. Positive loop reactions are confined within the limits of their sac, with adjacent, noninoculated loops remaining negative. Therefore, an advantage is realized in which several serotypes or strains may be simultaneously tested within the same animal. Problems of colonization relating to gastric acidity and peristalsis, which tends to "wash out" the intestine, are eliminated. When compared to the oral route of exposure, the ligated loop technique has the advantage of a signifi— cantly higher rate of reproducibility in response to enteropathogens (Moon and Whipp, 1971). However, certain limiting factors of the technique need to be considered when conducting the test. Moon et al. (1966) reported that loops made in the jejunal section of the small intestine are of greater enterosorptive sensitivity than those of the ileum. Enteropathogenic 16 E. coli do not cause positive responses 100% of the time, and a few pigs appear to be completely refractory (Mbon and Whipp, 1972). Age of the host alters the responsiveness of intestinal loops in much the same manner that it alters the observed susceptibility of pigs to natural enzootics (Moon and Whipp, 1970). The pathologic alterations observed in the ligated loop inoculated with enteropathogenic E. coli have done little to abate the confusion surrounding this disease. Gross lesions have been attributed to both the induced trauma of the procedure and the degree of enteropathogenicity of the organisms evaluated by this procedure. Moon et al. (1966) reported a serofibrinous peritonitis, with a fibrinous mesh encasing the ligated loops. Varying degrees of loop distention were observed which ranged from a negative response to maximal distention with an occasional ruptured loop. The flaccidity of the intestinal wall of the ligated loop was directly proportional to the degree of fluid accumulation and subsequent disten- tion. The fluid was characterized as being cloudy and of yellow, white or red color, and containing particulate matter of food and mucus. Nielsen and Sautter (1968) reported finding occasional areas of edema in the mesentery associated with the ligated loops. They also described gastric lesions, limited to the esophageal-cardial region, which were directly related to the ligation procedure. Nielsen and Sautter (1968) reported that there were no observable differences in the microscopic appearance of positive and negative loops, except in a group of 4 cesarian-derived pigs which had received an oral exposure to the enteropathogenic E. coli 0138:K81:Hl4 approxi- mately 1 month prior to exposure to the same organism via the ligated loop technique. The characteristic lesions in this special group l7 consisted of edema and a subintimal hyaline fibrinoid deposition in the small arteries and veins throughout the small intestine. The consistent lesion seen by Moon et a1. (1966) was a fibrinopurulent serositis on all intestinal loops. Other lesions appeared to be directly related to the degree of loop distention. Interloops and negative experimental loops were normal. In moderately distended loops, alterations ranged from simple architectural distortion to more significant changes. These changes included occasional subepithelial bullae, desquamation of apical villous epithelium and congestion. In addition, cellular infiltrations of polymorphonuclear and mononuclear cells in the lamina propria and submucosa were seen. Occasionally short villi, covered with basophilic cuboidal epithelium, were observed. Loops which were either maximally distended or had ruptured had the most severe changes. The intestinal wall from such loops consisted of a thin muscularis, flattened crypts and denuded, congested and hemorrhagic remnants of villi. Certain limitations of the ligated loop technique have been pre- viously mentioned. However, the most troublesome by far is the appearance of false positive loops, first described in pigs by Moon et al. (1966). Such loops, usually the noninoculated interloops, become distended with fluid and are indistinguishable from those experimental loops harboring known enteropathogens. The culturing of contents from such distended, noninoculated loops has not demonstrated the presence of any of the serotypes being tested in the experimental animal. The Gnotobiotic Pig Certain physiological and anatomical similarities between the pig and man have made the animal an ideal model for the comparative study of gastrointestinal disease (Meyer et aZ., 1964). Because no immune 18 substances are transferred in utero, and colostrum is withheld from the germfree pig, interfering antibody is absent (Kim et al., 1966). These animals are therefore more uniformly susceptible to infectious agents than are their conventionally reared counterparts (Waxler et aZ., 1971). Utilization of the gnotobiotic pig in studies of colibacillosis affords additional advantages over the conventional pig kept under field conditions. Since the occurrence of E. coli is widely spread throughout the environment in which pigs are maintained, the gnoto- biotic state eliminates all but those desired serotypes and strains. Without the presence of normal intestinal flora, the disease may progress under minimal restrictions and interference. All of these advantages result in the elimination of many of the experimental variables present under field conditions (Kohler and Bohl, 1966). The data obtained from E. cOZi—infected gnotobiotic pigs do not significantly differ from those recorded in enzootics of the disease (Saunders et al., 1963; Kohler and Bohl, 1966). MATERIALS AND METHODS Experimental Animals Procurement. Twenty-three germfree Yorkshire pigs, from 4 litters, were delivered by hysterotomy into sterile, plastic isolators, using the technique described by Waxler et al. (1966). Epidural anesthesia was administered to the gilts on the 112th day of their gestation. Twenty to twenty-five milliliters of 2.5% procaine hydrochloride* was injected into the epidural space at the lumbosacral articulation. Presurgical tranquilization with promazine hydrochloride** was required in 3 of the 4 gilts. Each gilt was euthanatized following the surgical delivery of all piglets. Approximately 4 hours after delivery, the germfree pigs were transferred into the rearing isolators where they remained for the next 3 weeks. Table 1 lists information concerning both experimental and control animals and includes age, sex and number of animals contained in each group. Rearing. In the rearing room the newborn pigs were randomly placed into individual cages within sterile, plastic isolators, so that each isolator contained no more than 4 pigs. The experimental animals in Litters l, 2 and 4 were maintained on the same sterile, semisynthetic * Epidural, Haver-Lockhart Co., Kansas City, Missouri. ** Sparine, Wyeth Laboratories, Cleveland, Ohio. 19 20 cm Hm mm mm mm mm mm mm aHeo oumeA Hm assumeummq< as nuoun oHHHOum e oumeA «awH oNHuosumoe< «amH goons maaumum e mummfia «a muaumsumoe< mH nuoun oHHumum w oumeq m oNHuosumoa< m FRET-ch21 om om mm mm mm mm Hm Hm Hm mm mm mm mm Hm Hm Hm Hm mm mm mm cm om mm N a a a a a a « c>e #:2121eaz:han.n.z: v-I v—I hak421212121k«21 .H~¢nn~or\ .vounuoau 0mm an .OmmH .aOHumoHaaaaoo Hadomuoo .aooz .3 .m .mom can BOO maHmuum maHvaHoaH .mwou .m oHaowonumoou0ua0.H mmnHo.O you now: an 59 Aomaav mean: was :ooz In: Asmaanmnv Hv AOOOHV Hmuunmm use domaouom .aooz O AHmoHH van Hmonnonv O O O mmH nonmav amass cam coo: o Asmasfimflv o Aooaav umuusmm was cmmamuom .:002 In- Hammad was Haasnmnv em «a 0.0 nHH AOmOHV ooan van coo: III AHoannofiv Om AOOOHV nonunnm van domaauom .aooz III AHmoHH can Hnannonv OOH mH.O m.mH mmO Aommav genes can coo: «.0 Aamasnmnv w AoomHv nouuaam can somamnom .aooz in: HammHH was Haasfianv mm mum.o o.Nq Hme Acnoav mass: was coo: o.~ aflmasmmnv Hm AOOOHV HouuSMm can someonom .eooz III nHmmHH pan Hnesflomv an ANOOHV Banyan Ode monO O.m I m.H AHmeoOosvv OOH N0.0 0.0¢ mom aosaav annex can moo: .me.H aflmasnmfiv nu n.~ n.0w ems muonunm xoOGH amoz AGOHumuoHv xapaH ANV eHmuum Nllmomaoomau one: mooaoomou OOOH oooH 0>HuHmom Hmesfloh o>HuHmom some coanHnem sump HnuaoeHuooxm moooH HmeHumouaH oaHuuoo aH woumou wNoo .m we maHauum O one now some moanHnso use HausaEHuooNo mo aonHuaoaoo .n oHnaH 60 colibacillosis. Moon and Whipp (1970) utilized this strain again and found that the enteropathogenicity of the organism had been lost. Strain 115 was supplied to us in its nonenteropathogenic state, and our experiences with this strain tended to support these findings up until the terminal experiment. Inoculation of this strain into Pig 22 produced not only a maximally-distended loop but one which ruptured during the 24-hour period following inoculation with the organism. The complete inoculation of this loop involved a second penetration of the needle through the intestinal wall. Although this event did not occur with any other loops during the course of this study, it appears highly unlikely that such an alteration in technique would be responsible for a significant distention. Moon and Whipp (1970) reported the occurrence of a single significant distention in the 52 trials with this strain. Although less likely to occur, our report of 1 significant distention in 15 trial loops is within the limits of probability based upon the experiences of Moon and Whipp (1970) with this strain. This inconsistency does, however, apparently address itself to the problem of falsely positive distended loops reported as a complication of the ligated loop technique when conven- tional animals were utilized. The problem of false positive loops will be dealt with at some length later in the discussion. Histopathologically, very little significance can be placed upon the occurrence of ulcerative and erosive lesions commonly associated with 4+ or maximally distended loops. Fresh et al. (1964) demonstrated that mucosal lesions, similar to those described above, were directly related to continuously excessive intraluminal hydrostatic pressures present in the ligated loop and had no primary relationship to the presence of enteropathogens or their toxins. Inflammatory changes, 61 also reported in conjunction with the mucosal destruction, relate to the normal host defense mechanisms associated with such tissue necrosis. Little variation was observed among serotypes concerning the quanti- tative extent of mucosal surface involvement when ulceration or erosion did occur. The degree of mucosal involvement generally ranged between 33 to 40% of the total absorptive area in each section examined. Those serotypes producing a consistently higher percentage of enterosorption did not correspondingly result in a more extensive degree of histopatho- logic mucosal change. Therefore, the presence of bacterial toxins appears to be insignificant in exaggerating the mucosal response to increased intraluminal pressures. The remaining changes, consisting of absorptive cell shortening, clubbing and/or the elimination of villus structure, and the thinning of the tunica muscularis all relate to increased intraluminal pressures occurring in association with loop distention and the subsequent stretching of the intestinal layers. It should be stressed, however, that although the presence of microscopic intestinal lesions associated with loop distention are not significant in and of themselves, their occurrence does relate to the enteropatho- genicity of the organism and the subsequent ability of the serotype or strain to cause enterosorption. The only histopathologic lesion directly related to the enteropathogenicity of the organism appeared to be the increased discharge of mucus by goblet cells. This lesion was reported by Moon et a1. (1971). Examination of histopathologic sections from liver and kidney failed to demonstrate significant lesions. Mesenteric lymph nodes were consistent in appearance with that described by Alexander et al. (1969), with the exception of the subcapsular sinusoidal areas which were heavily populated with the same neutrophil and eosinophil mixture reported for the medullary tissue. 62 The numerical rating system provided an excellent means by which the enteropathogenicity of the various serotypes might be compared. The control groups served to delineate those effects which were attributable to the ligation technique alone. Peritonitis and a mild "physiologic inflammation" of the intestinal mucosa were found to relate to the surgical technique and the presence of microorganisms in the intestinal lumen, respectively. With the exception of Strain 115, the location of the plotted points on the graph relating to total entero- pathogenic effects of single serotypes fell within a range predictable and consistent with the information originally supplied with the accompanying cultures by Dr. Moon. From Figure 7 it can be seen that 3 essential ranges of response are present. Strain 123 was unable to produce a total score of more than 2.5 and was supplied as a nonentero— pathogenic strain. The group of organisms represented by Strains 263 and 987 had been previously identified as being consistently loop- positive enteropathogens in pigs irrespective of age at time of inocu- lation. The mean scores of these 2 strains were substantially above the mean points plotted for either of the 2 other classifications of organisms, clearly ranking Strains 263 and 987 to be of greater entero— pathogenic capabilities than the other 4 strains tested. The age-resistance relationship reported by Moon and Whipp (1970) plays a significant role in the expression of the enteropathogenicity of Strains 637 and 431. These 2 strains are known to be consistently loop positive in conventional pigs under 2 weeks of age and relatively nonenteropathogenic in pigs over 6 weeks of age (Moon and Whipp, 1971). The plots in Figure 7 for these 2 strains are consistent with the above fact in that the mean scores of 2.7 and 4.5, respectively, obtained for 4-week-old pigs fall between the scores of the other 2 major 63 groups. The rating system also failed to demonstrate any significant differences between the rough and smooth variants of Strain 431 which appeared at midcourse in the experiment. The mean score for 431-S was 9.3, while that for 431-R was 9.5, and both values compared favor- ably with scores for the strain both prior to the split and following the return to a uniform colony appearance. The highly aberrant behavior of 2 test loops inoculated with the nonenteropathogen, Strain 115, is of great interest. The resulting scores from these 2 loops were well within the range reported for the highly enteropathogenic Strains 263 and 987 and resulted in an exag— gerated shifting of the mean score for the Strain 115 upward to the extent that it fell within the age-resistance class range of entero— pathogens. If those 2 remarkable scores were removed from the tally, the resultant mean score for serotype 115 would be 1.5 and safely within the range of nonenteropathogenicity. These 2 unexpected loop responses will be treated in a subsequent section of the discussion dealing with "false" reactions associated with the ligated loop technique. In general, the rating system clearly substantiated the direct relationship existing between the enteropathogenicity of a strain of E. coli and the gross and microscopic alterations observed in the gut following the ligated loop procedure. The total score was highly reflective of the enteropathogenicity and correlated with the aero- typic information originally supplied by Moon (1970) at the beginning of the experiment. The changes seen in the mitotic activity of cells within the crypts of Lieberkfihn were closely associated with the degree of loop distention observed in the experimental animals. Since the amount of 64 fluid accumulation directly relates to enterotoxin elaboration and thereby the enteropathogenicity of the organism, the decrease in cell division occurring in the crypts of Lieberkfihn may be in response to either the presence of the toxin, the mechanical effects of distention or a combination of both. The effects of dual microbial contamination within single sacs were inadvertently evaluated in Animals 15 and 16. These animals had been accidentally contaminated with an apparently nonpathogenic Micro- coccus sp. early in the neonatal period and were harboring the organism in their digestive tracts at the time of the experiment. No signifi- cant differences in loop response to the E. coli were found between the experimentally infected animals contaminated with the Micrococcus sp. and the larger body of experimentally infected pigs which had remained germfree prior to the inoculation of the test strains. Edema in the lamina propria and the submucosa was randomly dis- tributed throughout the loops of the experimental animals and in those control animals in which the ligation procedure was performed. There- fore, the occurrence of such alterations appears to relate directly to the surgery and resulting postoperative changes. Twice during the course of these experiments falsely positive loops were observed. Both occurrences were confined to Strain 115, with 1 example being detected both grossly and microscopically in Sac 3 of Animal 22 and the second being solely detected by the presence of extensive microscopic lesions in Sec 1 of Animal l. Aberrant distended loops (false positives) have been reported by numerous authors while utilizing the ligated loop technique in rabbits and pigs (Taylor et al., 1958; De and Chose, 1959; Moon et al., 1966; Moon and Whipp, 1971). 65 Mechanisms by which these false positive loops may arise have also been suggested. De and Chose (1959) attributed their occurrence to rapid injection or large volumes of the media placed within the ligated loop. Both of these points are inapplicable to our experimental system, and our findings tend to support the similar experience of Moon et al. (1966). Spontaneous infection of the sacs was thought to result in distention of interloops or sacs inoculated with known non- enteropathogens by Taylor et a1. (1958) and Moon et al. (1966). Moon and Whipp (1971) noted that false positive loops were limited to the caudal sacs of the ileal segment in rabbits, and they implicated the surgical technique and/or the normal secretory activity of this intestinal segment as the underlying cause of aberrant loop distention. Spontaneous infection was ruled out by these workers in that false positive loops also occurred in sacs not inoculated with E. coli. as well as in loops treated with specific antibiotics. In our experi- mental design, the placement of the loops within the ileal segment of the small intestine was avoided by positioning the initial, most caudal, ligature approximately 1 meter cranial to the ileocecal junction. Confirmation of this distance was made following euthanasia and aided in establishing the spatial relationship of the first sac to the ileocecal junction. In Animal 1, the first loop was located 124 cm cranial to the ileocecal valve. The distance in Animal 22 was 105 cm for the first ligature and, therefore, Sac 3 would have been no closer than 145 cm to the beginning of the spiral colon. Additional evidence failing to support the surgical/physiological theory of Moon and Whipp (1971) is that, of the over 200 loops made throughout the trials, only 1 of these sacs became falsely distended. This loop was inoculated with the Strain 115 that also resulted in moderate to severe 66 histopathologic alterations in an additional loop as well. False positive loops were not observed in any other sacs, whether from experimental or control animals. Enterosorption, which occurred in a few interloops from experimental animals and was attributed to leakage from adjacent significantly distended loops, was not considered to fit the definition of a false positive reaction. Therefore, the dual implication of Strain 115 as a potential enteropathogen appears to be inconsistent with the hypothesis of Moon and Whipp (1971) concerning their conclusions from the data obtained in rabbits and pigs. The leakage of toxins between loops must also be considered. The distention pattern for each of the 2 animals assists in delineating this effect, and is listed in Table 8. No interloop distentions occurred between the 115 inoculated loop and the next significantly distended loop, going in either direction. This information would apparently eliminate the possible effect of distention due to entero- toxin leakage. Table 8. Loop distention patterns for animals harboring false positive loops Sac location Animal Ileum —- l **I—1 2 I-2 3 I-3 4 I-4 5 I-5 6 -—- Duodenum 1 -* - + - + - + - - — - 22 + - - - +* - + - + - - * Sac receiving Strain 115. ** I = interloop. 67 Therefore, the most probable explanation for such false positive loop appearances relates to the subject of spontaneous infection. Spontaneous infection, in this context, as it applies to conventional animals, is interpreted to mean the activation of potential entero- pathogens present in the "normal" flora of the intestinal tracts of animals utilized for the ligated loop technique. These enteropathogens have remained in a quiescent or nonenteropathogenic state for unknown reasons. The conditions under which the expression of such potential enteropathogenicity is realized must somehow be satisfied, on occasion, by the ligation of the intestine into isolated loops. Such a procedure would tend to negate the effects of peristalsis, the flushing of the tract in general, and most probably other ill-defined physiologic alterations. This appears to be supported by the work of Moon and Whipp (1970) in which the Class III (nonenteropathogenic) strains of E. coli resulted in 6 false positive distentions out of 302 loops tested, or approximately 2% incidence. When examined separately by strain, the false positive rates for Strains 123 and 115 were 1.9 and 2.4%, respectively. Frequency of false positive reactions in inter- loops was not available either from their work or from the experiments of others. Our experiments demonstrated that Strain 115 caused false positive reactions at a rate of 13.3%. If, however, reactivity of the serotypes were judged strictly on gross distentions alone, as occurred in most of the cited work with the exception of Moon et al. (1966), the rate for this strain would be reduced to 6.7%. Although consider- ably higher, this latter figure compares more favorably with the report of Moon and Whipp (1970). Strain 115 has an interesting history concerning its enteropatho— genicity. When initially utilized by Moon et a1. (1966), the organism 68 caused a positive loop reaction rate of 96%. By 1970 Moon and Whipp reported that the rate had fallen to 2.4%, and they characterized the organism as belonging to the Class III, nonenteropathogenic, group. Our experiences support this reclassification of Strain 115. However, it appears that, especially in the germfree pig system, the occasional expression of an enteropathogenic trait in a nonenteropatho- gen may occur. Therefore Strain 115, and probably additional non- enteropathogens, might be considered to be rarely but conditionally enterOpathogenic. These strains might then account for the occurrence of a false positive event in ligated loops. The incidence of false positive loops was lower in our work when compared to reports in conventional pigs. Thus an advantage is realized when utilizing the gnotobiote in this testing procedure. However, it may be generally concluded that the germfree intestine appears to be less sensitive to the technique. SUMMARY The ligated loop technique was evaluated in 23 gnotobiotic pigs, 3 ix) 4 weeks of age, in an attempt to define the nature of false positive loops. Five serotypes, encompassing 6 strains of Escherichia coli, were utilized in this experiment. The 6 strains were divided into 3 subgroups: 1) strains enteropathogenic for pigs regardless of age, 2) strains enteropathogenic for pigs under 2 weeks of age, and nonenteropathogenic for pigs over 6 weeks of age, and 3) nonenteropathogenic strains. Each experimental loop was inocu- lated with a single strain, noninjected interloops separated each inoculated loop, and all 6 strains were tested in each of 15 experi- mental animals. Strain enteropathogenicity in gnotobiotic pigs, based upon significant visual loop distention, generally compared favorably with results obtained utilizing the same technique and strains of E. coli in conventional animals with the exception that the gnotobiotic jejunal loop may be somewhat less sensitive to certain strains. In addition, postoperative mortality rates among gnotobiotic pigs were higher than reported in conventional pigs. Light microscopic intestinal lesions in these gnotobiotic pigs were similar to those reported in the conventional pig. Mucosal erosions and ulcerations were often seen in maximally distended loops. Two false positive loops occurred during the experiment. Both were in sacs inoculated with Strain 115. False positive loops were 69 70 not seen in other sacs, whether inoculated or not. Results of these experiments support the hypothesis that false positive loops in con- ventional pigs arise from spontaneous infection. BIBLIOGRAPHY BIBLIOGRAPHY Alexander, T. J. L., Miniats, O. P., Ingram, D. G., Thompson, R. G., and Thackeray, E. L.: Gnotobiotic pigs: Procurement, microbial flora, serum porteins and lymphatic tissues. Can. Vet. J., 10, (1969): 98-105. Arbuckle, J. B. R.: The distribution of certain Escherichia coli strains in pigs and their environment. Brit. Vet. J., 124, (1968): 152-159. Barnum, D. A.: The control of neonatal colibacillosis of swine. Ann. N. Y. Acad. Sci., 176, (1971): 385-400. Barnum, D. A., Glantz, P. J., and Moon, H. W.: Colibacillosis. Ciba Veterinary Monograph Series. Ciba Foundation, Summit, N.J., 1967. Burrows, W., and Musteikis, G. M.: Cholera infection and toxin in the rabbit ileal loop. J. Inf. Dis., 116, (1966): 183-190. Christie, B.R.: Experimental Colibacillosis in Gnotobiotic Pigs. M.S. Thesis, Michigan State University, East Lansing, Michigan, 1967. Cross, R. F., and Kohler, E. M.: Autolytic changes in the digestive system of germfree, Escherichia coli monocontaminated and conventional baby pigs. Can. J. Comp. Med. Vet., 33, (1969): 108-112. Curtis, J., and Bourne, F. J.: Immunoglobulin quantitation in sow serum, colostrum and milk and the serum of young pigs. Biochim. et Biophys. Acta, 236, (1971): 319-332. De, S. N., Bhattacharza, K., and Sarkar, J. K.: A study of the patho- genicity of strainstacterium coZi from acute and chronic enteritis. J. Path. Bact., 71, (1956): 201-209. De, S. N., and Chatterje, D. N.: An experimental study of the mechanism of action of Vibrio cholerae on the intestinal mucous membrane. J. Path. Bact., 66, (1953): 559-562. De, S. N., and Chose, M. L.: False reaction in ligated loop of rabbit intestine. Indian J. Path. Bact., 2, (1959): 121-128. Drees, D. T., and Waxler, G. L.: Enteric colibacillosis in gnoto- biotic swine: A fluorescence microscopic study. Am. J. Vet. R880’ 31, (1970): 1147-1157. 71 72 Dubos, R.: The microbiota of the gastrointestinal tract. Gastro- enterology, 51, (1966): 868-874. Dunne, H. W., and Bennett, P. C.: Colibacillosis and edema disease, in Diseases of’Swine, 3rd ed., H. W. Dunne, ed. The Iowa State University Press, Ames, Iowa, (1970). Fresh, J. W., Versage, P. M., and Reyes, V.: Intestinal morphology in human and experimental cholera. Archives of Pathology, 77, (1964): 529—537. Gilka, F.: A comparative study on the pathological picture of trans- missible gastroenteritis (TGE) and E. coli gastritis in newborn piglets with special reference to the gastrointestinal tract. Acta Vet. Acad. Sci. Hung. Tomus, 18, (1968): 39-50. Cited by Leman, A. D., ed.: swine Health: Common Diseases Affecting Baby Pigs-A Research Review. College of Veterinary Medicine, University of Illinois, Urbana-Champaign, Illinois, (1970): 24-25. Grun, E., Gurtler, H., and Abadjieff, W.: Studies on pigs. XXVIII. Clinical, hematological and biochemical studies after injection of a freeze thawed extract of beta hemolytic E. coli. Arch. Exp. Vet. Med., 21, (1967): 1517-1530. Cited by Leman, A. D., ed.: Swine Health: Common Diseases Affecting Baby Pigs-A Research Review. College of Veterinary Medicine, University of Illinois, Urbana-Champaign, Illinois, (1970): 25. Gyles, C. L., and Barnum, D. A.: E. coli in ligated segments of pig intestine. J. Path. Bact., 94, (1967): 189-194. Kenworthy, R.: Effect of Escherichia coli on germ-free and gnoto- biotic pigs. 1. Light and electron microscopy of the small intestine. J. Comp. Path., 80, (1970): 53-63. Kenworthy, R.,and Allen, W. D.: The significance of Escherichia coli to the young pig. J. Comp. Path. Therap., 76, (1966a): 31-44. Kenworthy, R., and Allen, W. D.: Influence of diet and bacteria on small intestinal morphology, with special reference to early weaning and Escherichia coli. J. Comp. Path., 76, (1966b): 291-296. Kim, Y. B., Bradley, S. G., and Watson, D. W.: Ontogeny of the immune response. 1. Development of immunoglobulins in germfree and conventional colostrum-deprived piglets. J. Immunol., 97, (1966): 52-63. Kimberg, D. V., Field, M., Johnson, J., Henderson, A., and Gershaw, I.: Stimulation of intestinal mucosal adenyl cyclase by cholera enterotoxin and prostaglandins. J. Clin. Invest., 50, (1971): 1218-1230. 73 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., 31, (1967): 283-289. Kohler, E. M.: Enterotoxic activity of the filtrates of Escherichia coli in young pigs. Am. J. Vet. Res., 29, (1968): 2263-2274. Kohler, E. M.: Studies on the effects of immune serum on the course of E. coli infections in gnotobiotic pigs, in GermaFree Biology. Plenum Press, New York, N.Y., (1969): 219-225. Kohler, E. M.: Pathogenesis of neonatal enteric colibacillosis of pigs. J.A.V.M.A., 1960, (1972): 574-582. Kohler, E. M., and Bohl, E. H.: Studies of Escherichia coli in gnoto- biotic pigs. I. Experimental reproduction of colibacillosis. Can. J. Comp. Med. Vet., 30, (1966): 1294—1297. Kohler, E. M., and Cross, R. F.: Studies of Escherichia coli in gnotobiotic pigs. IV. Effects of feeding bacteria free filtrates of broth cultures. Can. J. Comp. Med. Vet., 33, (1969): 173-177. Kramer, T. T., and Nderito, P. C.: Experimental Escherichia coli diarrhea in hysterectomy-derived, one-day-old fasting pigs. Am. J. Vet. Res., 28, (1967): 959-964. Leman, A. D., ed.: Swine Health: Cbmmon Diseases Affecting Baby Pigs-A Research Review. College of Veterinary Medicine, Uni- versity of Illinois, Urbana-Champaign, Illinois, (1970): 18-33. Meyer, R. C., Bohl, E. H., and Kohler, E. M.: Procurement and main- tenance of germ-free swine for microbiological investigations. Appl. Microbiol., 12, (1964): 295-300. Miniats, O. P., Mitchell, L., and Barnum, D. A.: Response of gnoto- biotic pigs to Escherichia coli. Can. J. Comp. Med., 34, (1970): 269-276. Moon, H. W.: Personal communication, (1970). 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, (1970): 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, (1966): 1317-1325. Moon, H. W., and Whipp, S. C.: Development of resistance with age by swine intestine to effects of enteropathogenic E. coli. J. Inf. Dis., 122, (1970): 220-223. 74 Moon, H. W., Whipp, S. C., and Baetz, A. L.: Comparative effects of enterotoxins from Escherichia coli and Vibrio coli on rabbit and swine small intestine. Lab. Invest., 25, (1971): 133-140. Moon, H. W., and Whipp, S. C.: Systems for testing the enteropatho- genicity of Escherichia coli. Ann. N. Y. Acad. Sci., 176, (1971): 197-211. Muylle, E. and Oyaert, W.: Effect of coli endotoxin on some blood constituents of the pig. Zentralbl. Veterinaermed., 13, (1965): 405-409. Cited by Leman, A. D., ed.: Swine Health: Common Diseases Affecting Baby Pigs—A Research Review. College of Veterinary Medicine, University of Illinois, Urbana-Champaign, Illinois, (1970): 25. Namioka, S., and Murata, M.: Studies on the pathogenicity of Escherichia coli. II. The effect of a substance obtained from pseudo— eosinophils of rabbits on the organism. Cornell Vet., 52, (1962): 289-296. Nielsen, N. 0., and Sautter, J. H.: Infection of ligated intestinal loops with hemolytic Escherichia coli in the pig. Can. Vet. J., 9, (1968): 90-970 Orskov, F., ¢rskov, I., Jann, B., and Jann, K.: added at proof stage. Acta Path. Microbiol. Scand., 79B, (1971): 142-152. Cited by Sojka, W. J.: Enteric diseases in the newborn piglets, calves and lambs due to Escherichia coli infection. Vet. Bulletin, 41, (1971): 509-522. Pierce, N. F., Greenough, W. B., and Carpenter, C. C. J.: Vibrio cholerae enterotoxin and its mode of action. Bact. Rev., 35, (1971): 1-13. Porter, P.: Porcine colostral IgA and IgM antibodies to Escherichia coli and their intestinal absorption by the neonatal pig. Immunology, 17, (1969a): 617-626. Porter, P.: Transfer of immunoglobulins IgG, IgA and IgM to lacteal secretions in the parturient sow and their absorption by the neonatal piglet. Biochim. et Biophys. Acta, 181, (1969b): 381-392. 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. Saunders, C. N., Stevens, A. J., Spence, J. B., and Sojka, W. J.: Escherichia coli infections in piglets. Res. Vet. Sci., 1, (1960): 28-35. Shafer, D. E., Lust, W. D., Sercor, B., and Goldberg, N. 0.: Elevation of adenosine 3'5' cyclic monophosphate concentration in intestinal mucosa after treatment with cholera toxin. Proc. Nat. Acad. Sci., 67, (1970): 850-855. 75 Sharp, C. W. G., and Hynie, S.: Stimulation of intestinal adenyl cyclase by cholera toxin. Nature, 229, (1971): 266-269. Smith, H. W., and Gyles, C. L.: Escherichia coli enterotoxin. Vet. Rec., 85, (1969): 694. Smith, H. W., and Gyles, C. L.: The relationship between two appar- ently different enterotoxins produced by enteropathogenic strains of Escherichia coli of porcine origin. J. Med. Microbiol., 3, (1970): 387-401. Smith, H. W., and Halls, 8.: Observations by the ligated intestinal segment and oral inoculation methods on Escherichia coli infections in pigs, calves, lambs and rabbits. J. Path. Bact., 93, (1967a): 499-529. Smith, H. W., and Halls, 8.: Studies on Escherichia coli enterotoxin. J. Path. Bact., 93, (1967b): 531-543. Smith, H. W., and Halls, 8.: The transmissible nature of the genetic factor in Escherichia coli that controls enterotoxin production. J. Gen. Microbiol., 52, (1968): 319-334. 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. Sojka, W. J.: Escherichia coli in domestic animals and poultry. Review Series No. 7, Commonwealth Bureau of Animal Health, London, (1965): 104-156. Sojka, W. J.: Enteric diseases in newborn piglets, calves and lambs due to Escherichia coli infection. Vet. Bull., 41, (1971): 509-522. Staley, T. E., Jones, E. W., and Corley, L. D.: Attachment and pene- tration of Escherichia coli into epithelium of the ileum in newborn pigs. Am. J. Path., 56, (1969): 371-392. Staley, T. E., Jones, E. W., and Marshall, A. E.: The jejunal absorp- tive cell of the newborn pig. An electron microscopic study. Anat. Rec., 161, (1968): 497-515. Stevens, A. J.: Coliform infections in the young pig and a practical approach to the control of enteritis. Vet. Res., 75, (1963a): 1241-1245. Stevens, A. J.: Enteritis in pigs-A working hypothesis. Brit. Vet. J., 119, (1963b): 520-525. Taylor, J., Maltby, M. P., and Payne, J. M.: Factors influencing the response of ligated rabbit-gut segments to injected Escherichia coli. J. Path. Bact., 76, (1958): 491-499. 76 Taylor, J., Walkins, M. P., and Payne, J. M.: Relation of rabbit-gut reaction to enteropathogenic Escherichia coli. Brit. J. Exptl. Waxler, G. L., Christie, B. R., and Drees, D. T.: Use of the gnoto- biotic pig in the study of Escherichia coli infection. Ann. N. Y. Acad. Sci., 176, (1971): 141-149. Waxler, G. L., Schmidt, D. A., and Whitehair, C. K.: Technique for rearing gnotobiotic pigs. Am. J. Vet. Res., 27, (1966): 300-307. White, F., Wenham, G., Shorman, G. A. M., Jones, A. S., Rattray, E. A., and McDonald, J.: Stomach function in relation to a scour syndrome in the piglet. Brit. J. Nutr., 23, (1969): 847-862. Wilson, M. R.: Role of immunity in the control of neonatal colibacil- losis in pigs. J.A.V.M.A., 160, (1972): 585-590. Wilson, M. R., and Svendsen, J.: Immunity to Escherichia coli in pigs. The role of milk in protective immunity to E. coli enteritis. Can. J. Comp. Med., 35, (1971): 239-243. VITA The author was born on September 2, 1942, in Ashland, Ohio. His primary and secondary education was completed in Michigan, graduating from Farmington High School in 1960. The author entered Michigan State University in September of 1960 and was graduated in 1966 with the degree of Doctor of Veterinary Medicine. Shortly thereafter he was married to Anna Lottie Sollenberger, and entered the U.S. Army, being stationed in the Military District of washington. In August of 1968 he entered private practice in Southfield, Michigan, in association with Dr. D. L. Young. In August of 1969 the author returned to Michigan State University and entered the Master's degree program of the Department of Pathology, accepting a teaching assistantship. In September 1970 he entered the employment of Dr. Barnett Rosenberg, Department of Biophysics, Michigan State University, to begin research into the potential antiviral properties of various platinum coordination complexes. The author is a member of the A.V.M.A., M.V.M.A., Phi Zeta, and Sigma Xi. 77 MICHIGAN STAT IIIIII 3 1293 R IIIIIIIII 0307 IIIIIIIIIIIIIIIIES 9079 0