H “ MW“ 312 293011 WHEN Hm“ W HHR 0376 ¢ ‘ “’ AGICZ MAY 9250-1999 THE IDENTIFICATION AND PARTIAL CHARACTERIZATION OF ENTEROTOXIN-LIKE ACTIVITY IN SALMONELLA SPECIES ISOLATED FROM HORSES BY Laura Ann Ruggeri A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1978 J \C/“-‘"‘ " ABSTRACT THE IDENTIFICATION AND PARTIAL CHARACTERIZATION OF ENTEROTOXIN-LIKE ACTIVITY IN SALMONELLA SPECIES ISOLATED FROM HORSES BY Laura Ann Ruggeri The suckling mouse assay was used to determine the presence or absence of enterotoxin-like activity in strains of Salmonella isolated from horses. A fluid accumulating factor was demonstrated in 9 out of 17 crude culture super- nates prepared from 18 h roller tube cultures of Salmonella grown in brain heart infusion (BHI) broth. Of these, 6 were Salmonella typhimurium, 2 were Salmonella anatum, and one was Salmonella cholerasuis. Enterotoxin-like activity was not stable to freezing (-70 C) and thawing, but was stable to boiling for 30 min. Maximum levels of enterotoxin-like activity occured in 8 to 14 h roller cultures of Salmonella and in 6 h shaker cultures of Salmonella in filtered BHI. Greater than 4.0 x 106 washed cells/mouse were required to stimulate fluid secretion in the suckling mice. Enterotoxin-like activity was elicited by crude culture supernates of Salmonella containing an average of 1.9 x 106 viable cells/0.1 ml. This difference in bacterial number Laura Ann Ruggeri and the fact that enterotoxin-like activity was retained by a Millipore filter (0.45um) suggests that enterotoxin was primarily cell bound and was released into the growth medium in low concentration. To Bob, encouragement, patience and understanding you were always there To Joyce and Frank Ruggeri To Donna ii ACKNOWLEDGMENTS I wish to thank Dr. Robert Moon, my research advisor, for his guidance throughout my graduate career. Special thanks go to him for all the time he spent reviewing this thesis and for his overwhelming concern. I also wish to thank Dr. Betty Werner, Bob Leunk, Rick Friedman, Marilyn Thelen and Hassan Tavakoli for their friendship during my graduate studies. Thank you all. iii TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O O O O O I O O O O 0 Vi LIST OF FIGURES O O O O O O O O O O O O O O O O O O O Vii INTRODUCTION 0 O O O O O O O O O O O O O O I O O O O O 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . 3 Enterotoxins of Escherichia coli . . . . . . . . . 3 Animal and Cell Models for Detection of Enterotoxins . . . . . . . . . . . . . . . . . . 5 Salmonellosis in Equidae . . . . . . . . . . . . . 9 Salmonellosis in Humans . . . . . . . . . . . . . 12 Pathogenic Mechanisms of Salmonella . . . . . . . . l4 MATERIALS AND METHODS . . . . . . . . . . . . . . . . 20 Cultures . . . . . . . . . . . . . . . . . . . . . 20 Media and Maintenance of Cultures . . . . . . . . . 20 Preparation of Enterotoxin Samples . . . . . . . . 21 Effect of Heat on Enterotoxic Activity . . . . . . 22 Chemicals . . . . . . . . . . . . . . . . . . . . . 22 Suckling Mouse Assay . . . . . . . . . . . . . . . 23 Statistics . . . . . . . . . . . . . . . . . . . . 24 RESULTS 0 O O O O O O I O O O O O O O O O O O O O O O 25 Production of Enterotoxin-Like Activity . . . . . . 25 Temperature Stability of Enterotoxin-Like Activity . . . . . . . . . . . . . . . . . . . . 25 Production of Enterotoxin-Like Activity Under Various Conditions of Bacterial Growth . . . . . 28 Loss of Enterotoxin- Like Activity Upon Filter Sterilization of Samples . . . . . . . . . 33 Activity of Lipopolysaccharide (LPS) in the Suckling Mouse Assay . . . . . . . . . . . . . 37 Activity of Live Salmonella in the Suckling Mouse Assay . . . . . . . . . . . . . . . . . . 37 iv Page DISCUSSION . . . . . . . . . . . . . . . . . . . . . 4O LITERATURE CITED 0 O O O O O O O C O C O O O O C O O 47 LIST OF TABLES Salmonella Species and Source of Isolation . . . Production of Enterotoxin-like Activity in BHI Roller Tube Cultures of Salmonella . . . . Loss of Enterotoxin-like Activity After a 6 Week Period of Storage at -70 C . . . . . . . . . . Retention of Enterotoxin-like Activity After Heat Treatment . . . . . . . . . . . . . . . . Loss of Enterotoxin-like Activity After Filter Sterilization of Supernatant Fluids . . . . . Inability of Salmonella LPS to Stimulate Enterotoxin-like Activity in Suckling Mice . . Reactivity of Suckling Mice to Washed Cells of Salmonella . . . . . . . . . . . . . . . . . . vi Page 26 27 29 30 36 37 39 LIST OF FIGURES Figure Page 1. Growth Curve of S. typhimurium E9288 in Roller Culture and Shaker Culture . . . . . . . . . 31 Titration of S. typhimurium E9288 Enterotoxin- like Activity in RoIler Culture and Shaker culture 0 O O O O O O O O O C O O I O O O O 2. 34 vii INTRODUCTION Acute diarrheal disease in both man and animals may be caused by a wide range of organisms. Several genera of enteric pathogens release enterotoxins into the lumen of the small intestine. The toxins, which act locally to stimu- late fluid release and net loss of body water and electro- lytes are best characterized in Vibrio cholerae. To date enterotoxins have also been demonstrated in Escherichia coli (37, 71, 86), Klebsiella pneumoniae (47, 50), Enterobacter cloacae (48), Vibrio parahaemolyticus (74), Clostridium perfringes (89), Yersinia enterocolitica (64), and Salmonella typhimurium (75, 80). Despite the fact that Salmonella accounts for a sig- nificant amount of enteric disease, comparatively little is known about the pathogenic mechanism(s) of these organisms. Salmonella differ from most other enterobacteria in their ability to penetrate the intestinal wall and survive within cells of the reticuloendothelial system (92, 93), but invasion of mucosa alone is not sufficient to evoke fluid loss in Salmonella enteritis (28). In 1974 (73) an entero- toxin associated with Salmonella was described in isolates from human patients with gastroenteritis. Subsequently other investigators (51, 75, 76, 79, 80) have produced, characterized, and partially purified enterotoxins from Salmonella enteriditis and S. typhimurium. The primary objective of this study is to determine whether clinical isolates of Salmonella from horses possess enterotoxin-like activity. Salmonella isolated from asymptomatic equine carriers and from clinically active horses suffering from both acute and chronic diarrhea were evaluated. Enterotoxin activity was tested for using the suckling mouse assay (13). When a number of positive isolates were identified a second objective was to char- acterize selected physiological parameters of production and stability of these enterotoxin—like substances. LI TERATURE REVI EW Bacterial enterotoxins cause diarrhea in humans and neonatal animals (calves, pigs, and lambs). The role of enterotoxins in equine diarrheal disease has not been documented. Human cholera is considered the pathophysiologic prototype of these diarrheal enterotoxic diseases. The enterotoxin responsible for fluid loss in cholera has been purified, its mechanism of action determined, and its anti- genic prOperties studied (27, 53, 65). In reality, enteric disease produced by enterobacteria is responsible for more world-wide morbidity and mortality than cholera itself (8). Enterotoxins of Escherichia coli Enterotoxigenic Escherichia coli (ETEC) is the most common cause of travelers diarrhea and is a major problem among newborn infants and neonatal animals. E. coli causes disease by colonizing the small intestine and producing enterotoxins. Adherence to epithelial cell surfaces is mediated by specific heat-labile, plasmid controlled surface antigens. These antigens have been characterized as the K88 antigen of swine-specific ETEC (41), the K99 antigen of bovine- and sheep-specific ETEC (62), and the colonization factor antigen (CFA) of ETEC isolated from man (21). The toxins (37, 71, 83), first isolated from strains entero- pathoqenic for pigs, have been designated as heat-labile (LT) toxins (39) and heat—stable (ST) toxins (86). Heat- labile toxin is a nondialyzable protein molecule destroyed by heating to 60 C for 30 minutes (39). Molecular weights of 20,000 (40), 100,000 (17), 200,000 (78) and greater than 300,000 daltons (52) have been reported. E, coli LT is immunogenic (39) and immunologically cross reactive with LT and cholera toxin (CT) in immunodiffusion studies (38). Antisera against either toxin will neutralize the various biological actions of both toxins (19, 82). E. coli LT stimulates adenyl cyclase activity and elevates intracellu- lar concentrations of cyclic adenosine 3', 5'-monophosphate (cAMP) (22, 42), but on a weight basis is 100 fold less toxic than cholera toxin (71). By contrast, ST is a smaller dia- lyzable molecule and is stable to boiling (86). Alderete and Robertson (1) recently purified ST, a polypeptide of 47 residues. The toxin has a molecular weight of 4,400, by SDS gel electrophoresis and gel filtration, and a molecular weight of 5,133 as determined by amino acid analysis. Bio- logical activity is not lost after treatment with pronase, trypsin, proteinase k, deoxyribonuclease, ribonuclease, phOSpholipase C, or acid. ST is a poor antigen (20, 83) and is weakly neutralized by anticholeragenoid (49). The synthesis of diarrhea—inducing enterotoxins by E. coli is plasmid mediated (87). Two classes of ENT plasmids have been recognized, one determining production of both LT and ST (37), the other determining only ST (40). Animal and Cell Models for Detection of Enterotoxins Enterotoxins can be readily detected in cell-free supernatant fluid following growth of bacteria in broth media. The large number of models available for entero- toxin assays were originally used to study V. cholerae enterotoxin and have been adapted for E. coli, Salmonella and other enteric pathogens. None are truly in 31339 assays and all are biologically based systems. E. coli LT is commonly measured by the rabbit ileal 100p assay (20, 84, 85), the Y-l adrenal tumor cell culture assay (15), the Chinese hamster ovary (CHO) cell culture assay (35), and the skin vascular permeability factor assay (23). Produc- tion of ST is measured in the rabbit ileal loop assay (20) and in the infant mouse assay (13). The Salmonella entero- toxins, which are the major tOpic of this thesis, have been studied in the rabbit ileal loop assay (79, 80), the infant mouse assay (51), the CH0 cell assay (76), and the skin per- meability factor assay (75). The most widely used intestinal assay is the adult rabbit ileal loop described by De and Chatterje in 1953 (12) for y. cholerae. In this assay the test material (either live bacteria or culture filtrates) is injected into alter- nating segments of the small intestine. Cholera toxin elicits a delayed, but steadily increasing, secretory response at low as well as high doses. The volume of fluid per length of intestinal segment is recorded (ml/cm) 18 hours after injection (4). Investigators have employed various modifications of the ligated loop assay for other enterics. Evans et a1. (20) showed that individual responses to E. coli ST and LT differed in a characteristic manner. Onset of net fluid accumulation in response to ST appeared between 4 and 6 hours, whereas maximum ratios elicited by LT occured between 16 and 18 hours after injection. As with other diarrheagenic organisms, the Salmonella have been studied extensively using the rabbit model. Sakazaki (73) first demonstrated enterotoxin-like activity in Salmonella. An improved ileal loop assay was developed by Sedlock and Deibel (79). In this model a greater response rate to enterotoxin in sterile filtrates was obtained when the intestinal lumen was washed with a balanced salt solution containing a mucolytic agent (N-acetyl-L-cysteine) prior to exposure to toxin or when the toxin was given in con- junction with a non—toxin producing, invasive Salmonella. It was suggested that intestinal mucin passively acts as a barrier preventing toxin access to epithelial cells. The invasive Salmonella was thought to reduce the presence of intestinal mucin and hence eliminate a component which may act to inhibit enterotoxin action. The infant mouse assay was originally described by Dean (13) in 1972 for E: coli ST. Four hours after intragastric inoculation of l to 4 day old suckling mice the ratio of the weight of the intestinal tract and its contents to the remaining body weight (IW/BW) was determined. Ratios between 0.070 and 0.090 were scored as questionably positive while those greater than 0.090 were scored as strongly posi— tive. This assay proved more convenient and reliable and the results paralleled data from the rabbit ileal 100p model. The reproducibility and optimal growth and test conditions for the assay of ST were examined by Giannella (30). Sixteen to 24 hour roller tube cultures of E, 92;; grown in casamino acid yeast extract broth prove to be the Optimal conditions for maximum production of ST. Samples were tested in l to 3 day old mice for 3 hours. A 95% confidence interval for positive IW/BW ratios was constructed with values greater than 0.083 being scored as positive, and those between 0.075 and 0.082 considered indeterminate. The day-to-day variation of the suckling mouse assay ranged from 10.5 to 15.7% depending on the bacterial strain. More recently Moon et a1. (57) showed that the infant mouse assay for ST could be made more sensitive, reliable, and simpler than the standard assay by using older mice, higher temperatures of incubation, and diarrhea as the index of response. Mice older than 4 days and/or held at 37 C were exposed to ST and found to develop diarrhea. In the standard assay, 1 to 4 day mice held at 25 C, responded to ST by accumulating intestinal fluids. The alteration of the response from fluid accumulation to diarrhea with increase in age and temperature is one reason why some mice older than 4 days or held at 37 C gave negative responses in the standard assay. Recently cholera toxin has been shown to stimulate a positive reaction in the suckling mouse assay (91). Koupal and Deibel (51) obtained positive results with §Elf monella after making slight modifications in the standard procedure. Rabbit skin tests can be used to detect enterotoxins in E. coli (23, 84), Z. cholerae (10) and Salmonella (75). In 1965, Craig (10) showed that intracutaneous injection of culture filtrates of y, cholerae caused erythema, induration and increased capillary permeability of small blood vessels in rabbit skin 18 to 24 hours after injection. Visualiza— tion of the reaction was enhanced by intravenous injection of Pontamine sky blue dye at the 18th hour. Salmonella pro- duces a permeability factor that is indistinguishable from the reactions of CT and LT. Peterson suggests that chroma- tography of culture filtrates on Sephadex G-100 is required for optimal induration activity (75). Many strains of Salmonella also produce a rapid vascular permeability response with a critical bluing time of one hour. No indu- ration is observed with the rapid permeability factor (PF) reaction. The early factor of Salmonella can be demon- strated only by administering the bluing dose within a critical time after intradermal injection of culture fil- trates. If this is not done, false positives occur or the entire reaction is blanched. Rapid PF activity is stable for at least 4 hours at 100 C, whereas the delayed PF of Salmonella, CT, and LT are heat-labile. Various tissue culture lines have been successfully used in the assay of enterotoxin. In all systems, a sterile filtrate is added to the tissue culture medium. A loga~ rithmically growing monolayer of cells is bathed with the suspension and the cultures are incubated. Donta et al. have shown that picogram quantities of purified cholera enterotoxin are capable of inducing morphologic changes and steroidogenesis in monolayer cultures of Y-l mouse adrenal cells (14). Similar cellular alterations are induced by LT (15). Chinese hamster ovary (CHO) cells are also sensitive to LT and CT. These cells accumulate CAMP and elongate. The effect is apparently mediated through cellular adenyl cyclase activation (35). Changes in CHO cells indistin- guishable from those produced by CT and LT are elicited by chromatographed culture filtrates of E, typhimurium (76). Salmonellosis in Equidae Salmonellosis represents a global problem for man and animals. Members of the genus have host-parasite relationships over a wide host range, being most prevalent in turkeys, followed by chickens, cattle and swine (58). Salmonella are also recognized as a serious and continuing IO problem in horses, the world-wide prevalence ranging from 0.37 to 27% (60). Salmonellosis in Equidae is a specific diarrhea-inducing disease with significant economic impli- cations, particularly in large equine clinic complexes. At least forty of the recognized Salmonella sero- types are known to infect horses (60), the most prevalent being Salmonella typhimurium, Salmonella anatum, Salmonella enteriditis, and Salmonella newport (7, 60, 68). None of these are normal residents of the intestinal tract of a healthy horse (72). Salmonella-induced colitis is regarded as an important enteric infection of young foals but is. often overlooked as a major cause of acute, rapidly fatal diarrhea in the adult horse (3, 7, 60, 61). Debilitated, aged and stressed horses, as well as foals, are highly vul- nerable to acute salmonellosis (59). A foal suffering from salmonellosis follows a characteristic disease pattern (7, 60, 61, 68). Initially the animal goes off feed, is depressed, weak, and its body temperature may reach 40 C or higher. Profuse, watery diarrhea developes leading to severe dehydration, elec— trolyte depletion, and acidosis. Blood urea nitrogen and packed cell volume values increase concurrent with leuko- penia; either the neutrophil count is less than 3,600/mm3 or a rapid decline in neutrOphil number is seen (16). Unless detection and therapy are immediate, the animal's chance of survival decreases rapidly. A foal may develOp 11 a peracute septicemic infection characterized by persistent fever, soft feces with excessive mucous, visceral absesses, and death in 24 to 72 hours. The mortality rate ranges between 17 and 50%. In the acute form the course of infec- tion lasts from 1 to 3 weeks. The fever and diarrhea are intermittent, and frequently the foal recovers. The chronic form lasts 3 weeks to several months. The outcome is usually fatal if the foal is ill greater than 3 weeks (60). Occasionally Salmonella escape the gut and infect joints and bones (most commonly the stifle and hock joints) (7, 68). In contrast to the foal, the normal adult horse is quite resistant to salmonellosis (68). Physical, environ— mental and surgical stress are important factors in the pathogenesis of adult horse illness (60). A horse with a subclinical infection may become acutely ill following transportation (9), surgery with anaesthesia (58), over- training (58), hot, humid weather (7, 58, 60, 66), new environment, changed feeding regimen, or pregnancy. Owen (63) has also presented clinical evidence that a Salmonella carrier can develop post-stress diarrhea which may be exacerbated by tetracycline therapy that depresses or alters the normal intestine bacterial antagonists of the Salmonella. Often a similar clinical picture occurs in which Salmonella cannot be isolated from the stool. A negative rectal swab is of little value in this instance. Whether the sample is from a living or dead animal, repeated and selective measures must be undertaken to confirm a negative result. 12 Owen (63) suggests that Colitis X may be part of the Salmo- ggllg_syndrome. Colitis X (36, 69, 90) has not been defined on the basis of an etiologic entity, but it does have reCOg- nizable characteristics similar to acute salmonellosis, i.e., history of stress, severe diarrhea, and signs of toxemia (leukopenia, fever, and death in 3 to 24 hours). Colitis X does not appear to spread laterally, whereas horizontal transmission of Salmonella is possible either by direct con- tact with an infected horse or by indirect contact with a contaminated environment (16). Various therapeutic regimens for salmonellosis are employed (59, 61, 68). Initially antibiotics are admini- stered parenterally in dosages within or higher than the recommended limits. Sensitivity is usually limited to nitrofuran products, chloramphenicol, and neomycin sulfate (61). However, specific antimicrobial therapy in peracute, many acute, and most chronic cases is generally ineffectual (60). Correcting dehydration, electrolyte imbalances, diarrhea, and shock with a balanced electrolyte solution such as lactated Ringers with NaHCO3 and glucose (97) is of more value. Medication to control the diarrhea and promote appetite is of equal importance. Salmonellosis in Humans The pathogenicity of Salmonella in humans was first discovered by Gartner (29) in 1888 in an epidemic of l3 gastroenteritis. Since then, the incidence of the disease has been on the increase (18). Modern methods of mass pro- cessing and distribution of many naturally infected foods promote the dissemination of the microbe. Furthermore, recent advances in surgery and medicine are attended by an increased risk of infection by Salmonella (6). There are four main groups of clinical manifestations in salmonellosis that may occur individually, simultaneously or consecutively in the course of infection. They are gastroenteritis, bacteremia with or without extraintestinal localization, typhoid-like patterns, and the carrier state (81). In man, gastroenteritis is the most common clinical picture (77). Symptoms may range from mild diarrhea to a fulminant form with rapid dehydration. Usually the disease is self-limited and lasts from one to four days. As early as 1951, McCullough (55, 56) investigated the pathogenicity of various food-born Salmonella for man. Human volunteers were fed various strains of Salmonella. All proved to be pathogenic with great variation in infec- tivity. The incubation period ranged from 8 to 72 hours and the cases varied in severity from mild brief enteritis to serious prostrating illness. Increased susceptibility to Salmonella enteritis after gastric surgery has been reported (2, 95). These patients are also subject to a more severe cholera-like illness (31). The disease, however, can be differentiated 14 from cholera because the diarrhea is of longer duration and has a final electrolyte concentration less than that of cholera (32). Pathogenic Mechanisms of Salmonella The existence of a Salmonella toxin responsible for the diarrhea has been postulated for years and various investigations describe a number of possible virulence factors which may contribute to the diarrhea associated with gastroenteritis. One prOposed mechanism is related to the ability of Salmonella to invade the intestinal epithelium during the course of infection. Due to the difficulty in working with human volun- teers and patients, laboratory animal models have been developed to study the pathogenesis of enteric disease. Since Rhesus monkeys develop symptoms similar to human Salmonella food infection (11), the morphological altera- tions of the intestinal tract have been examined in the monkey. In 1966, Kent et al. (44) discovered most severe and extensive lesions in the colon, with mild early inflam- matory reaction in the ileum followed by ileal lesions. The fact that E, typhimurium was isolated from all lesions suggested that bacterial invasion of the intestinal wall was necessary for Salmonella enteritis. But unlike cholera, Salmonella diarrhea in the monkey involved both the small and large intestine. 15 Takeuchi and Sprinz (92, 93) showed that guinea pigs preconditioned by starvation and opium (43) develOped acute enteritis following intragastric administration of E. typhimurium. By electron microsc0pe studies they were able to demonstrate bacteria penetrating epithelial cells of the brush border into the lamina propria, where they pro- liferated within macrophages. Once colonization and invasion of the intestinal lining occurs, symptoms of enteritis were manifested. In the rat Salmonella enterocolitic model (54), Powell et a1. (67) showed that the secretion of water and electrolytes by the ileum was the major determinant of diarrhea in the rat. Rout et al. (70) determined that alterations in fluid tranSport were correlated with changes in intestinal mor- phology and with intestinal concentrations of Salmonella in monkeys. Two possible mechanisms could account for the net transport of fluid and electrolytes into the bowel lumen; (l) passive transudation of fluid or (2) active electrolyte secretion. That certain invasive strains of Salmonella did alter active transport across ileum was determined utilizing the adult rabbit intestinal loop model (28). Taylor and Wilkins (94) first demonstrated that live, invasive cultures of Salmonella would cause dilation and fluid accumulation in ligated rabbit intestinal 100ps. The relationship of mucosal invasion to the mucosal inflammatory 16 response and fluid production by the rabbit ileum was investigated by Giannella et al. (33). Mucosal invasion is essential to the pathogenesis of salmonellosis but extensive mucosal inflammation did not seem to be a necessary prere- quisite or directly responsible for fluid secretion. Upon examining ion transport across isolated ileal mucosa (28), invasive Salmonella were found to inhibit active sodium absorption and stimulate active chloride secretion. Since the addition of CAMP (25) or cholera toxin (26) to normal ileal mucosa enhances these same secretory processes, increased levels of adenyl cyclase were looked for. Gian- nella (34) showed that E. typhimurium infection resulted in activation of mucosal adenyl cyclase and that the accumula- tion of cAMP concommittant with ileal fluid secretion sug- gests that the adenyl cyclase-cAMP system mediates the secretion. Although a bacterial property or factor, in addition to invasion and inflammation of the gastrointes- tinal mucosa seems to be responsible for fluid exorption, positive loops could not be induced by culture filtrates (33), suggesting enterotoxin was not responsible for fluid loss. Salmonella infection of the intestine and the attendant acute inflammatory process may result in the local synthesis of prostaglandins which activate mucosal adenyl cyclase activity (45). Elsewhere in the gastro- intestinal tract, acute inflammation results in synthesis l7 and release of prostaglandins (96), which can then stimulate intestinal cyclase activity (45). Indomethacin (88) (a potent inhibitor of prostaglandin synthesis) abolishes Salmonella-mediated adenyl cyclase activation and fluid secretion (34), but only partially inhibits cholera toxin- mediated secretion without altering the activation of adenyl cyclase (46). This difference suggests the involvement of prostaglandins in the Salmonella-mediated but not cholera toxin-mediated activation of adenyl cyclase. The ability of Salmonella to produce enterotoxic factors in culture filtrates was first demonstrated by Sakazaki et a1. (73) in the rabbit intestinal loop model. Eleven out of 13 Salmonella filtrates gave positive loop reactions. Histological examinations of sections of the intestinal mucosa from positive loops showed pathological changes similar to those produced by toxigenic E, ggil. Evidence for a protein enterotoxin was reported by Koupal and Deibel (51, 80). The toxin, most of which is strongly bound in the outer membrane of the cell wall of S. enteriditis, was obtained from the organism in its exponential phase of growth in brain heart infusion (BHI) broth shaker culture. The activity of the enterotoxin (determined by the suckling mouse model) was destroyed by pronase, but was resistant to a amylase, trypsin, lysozyme, phospholipase A, C, or D, acidity, and alkalinity. Uti- lizing the rabbit ileal loop assay, Koupal and Deibel (80) l8 also detected enterotoxin activity from E. typhimurium grown in various complex and defined media. Activity was observed in culture filtrates of the organism in mid stationary growth in BHI, 2% Casamino Acids media, synthetic amino acids medium, and a minimal glucose-salts medium. The toxin appears stable to low temperatures but a decrease in activity is observed when the preparations are frozen and thawed. Attempts at molecular weight characterization were inconclusive. An additional factor that might be involved in Salmonella gastroenteritis has been described by Sandefur and Peterson (75, 76). Initially characterized as a delayed permeability factor, it proved to be heat-labile, causing erythema and induration after 18 hours in a rabbit skin test, and effecting the elongation of Chinese hamster ovary cells. The culture filtrate, containing this factor, must be passed through a Sephadex G-100 column in order to remove an inhibitor-like substance that was found to mask factor activity. Even though intestinal secretory activity was not demonstrated, it appears similar to cholera enterotoxin since it can be neutralized by cholera antitoxin and there- fore may play an active role in Salmonella gastroenteritis. The biological properties of the substances produced by Salmonella (i.e., skin permeability alteration, CHO cell elongation, rabbit intestinal and suckling mouse fluid accumulation) are shared by both cholera and E. coli l9 enterotoxins. Even though evidence suggests they may be responsible for fluid and electrolyte loss during intestinal infections with y. cholerae and E. ggli, the actual patho- genic role of these substances is unknown in human and equine salmonellosis. MATERIALS AND METHODS Cultures The Salmonella typhimurium (case numbers 3CH, 22CH, 875H, 18CH, 20CH, 844H, T24CH and E9288), Salmonella anatum (case numbers 25CH, ZCH, 166H, 017H, E312,153, 200H, 184H and A24CH) and Salmonella cholerasuis NADL used in this study were obtained from Dr. E. V. Morse, School of Veteri- nary Medicine, Purdue University, Lafayette, Indiana. All organisms were isolated from asymptomatic equine carriers and from clinically active horses suffering from both acute and chronic diarrhea. The clinical histories of some but not all patients were available and are published (60). In certain phases of the study, a toxigenic E, coli strain, obtained from Dr. J. W. Peterson, University of Texas, Galveston, was included for control purposes. Media and Maintenance of Cultures The original cultures of Salmonella were maintained on tryptose agar slants (Difco Laboratories, Detroit, Michigan) at 4 C. For storage, all Salmonella were grown in 10 ml of BHI broth (Difco Laboratories, Detroit, Michigan), pH 6.8 in stationary culture. After 24 h at 37 C, the cultures were centrifuged at 7,710 x g for 20 min on a 20 21 Sorvall RC-S Superspeed Refrigerated Centrifuge (Ivan Sorvall Inc., Norwalk, Connecticut). The sedimented cells were washed twice in 0.85% NaCl, resuspended in 5 ml of physiological saline, added to 5 ml of glycerol (Mallinck- rodt, St. Louis, Missouri) and frozen at -20 C. Organisms treated in this manner served as the source of inocula for all cultures. rain heart infusion (BHI), filtered BHI, or Evan's casamino acid yeast extract (CA-YE) broth (19) were used as growth media. The CA-YE contained 20 g Casamino Acids (Difco Laboratories, Detroit, Michigan), 6 g yeast extract (Difco Laboratories, Detroit, Michigan), 2.5 g NaCl, 8.71 g 0.05 M KZHPO4-HZO, and 1 ml of trace salts (5% MgSO -7H 0, 4 2 0.5% MnCl -4H 0, and 0.5% FeCl S04). 2 2 2 2 Volume was brought to 1 liter and pH adjusted to 8.5 with dissolved in 0.001 N H 0.5 N NaOH. When appropriate BHI was filtered with XM300 membrane (Amicon, Lexington, Massachusetts) and pH adjusted to 7.85 with 0.5 N NaOH. Serial tenfold dilutions of culture filtrates were plated on tryptose agar for bacterial count determination. Preparation of Enterotoxin Samples Enterotoxin material was prepared by inoculating broth with glycerol storage cultures. Usually 5 ml of broth in 150 x 16 mm tubes was inoculated with 1 drop of cells and incubated at 37 C in a Wheaton Roller Culture Apparatus (Wheaton Scientific, Millville, New Jersey). Speed setting 22 1/2 rev/min. After 18 h the cultures were centrifuged at 12,100 x g for 20 min on a Sorvall RC-5 Superspeed Refrig- erated Centrifuge. The supernatant fluids from the broths were used as the crude preparations for inoculation of mice, or were filtered through a Swinnex-l3 membrane filter (Millipore Corp., Bedford, Massachusetts) with a pore size of 0.45 um before testing. Since previous studies (30) suggested that filtered and unfiltered supernatant fluids behave identically in suckling mice, the bulk of studies were done with unfiltered supernatant fluids. For comparative growth studies, overnight cultures of Salmonella were also inoculated into filtered BHI, the pH adjusted to 7.85, optical density 0.060 at 520 nm, and incubated at 37 C in a Gyrotory shaker water bath (New Brunswick Scientific, New Brunswick, New Jersey), 130 rev/ min, for various periods of time and enterotoxin material was prepared. Effect of Heat on Enterotoxic Activigy Sealed portions (2ml) of the test sample were heated at 100 C in a water bath for 30 min. After heating, the samples were cooled to 37 C and assayed. Chemicals Enterotoxin preparations were diluted in 0.01 M Tris (hydroxymethyl) aminomethane buffer (Sigma, St. Louis, Missouri), pH 8.0. 23 One hundred mg of Salmonella typhimurium W lipopoly- saccharide (Difco Laboratories, Detroit, Michigan) was resuspended in 10 ml of 0.85% saline. The total volume was split and dispensed into 10 separate vials and stored at —5 C until assayed for enterotoxin activity. Twofold dilu- tions of LPS were made using physiological saline as the diluent. £43.99: Breeding colonies of two month old 18 to 20 g HA/ICR mice (Spartan Research Animals, Haslett, Michigan) were established and their litters, approximately 10 to 16 mice per litter, were used in the suckling mouse assay. The mice were housed 6 per cage, with pine wood chips as bedding in a room at 24 C. Food (Wayne Lab-Blox, Allied Mills Inc., Chicago, Illinois) and water were available 3E libitum. Suckling Mouse Assay The suckling mouse assay described by Dean et a1. (13) was modified as follows. Newborn suckling mice (2 to 5 days old) were separated from their mothers immediately before use and randomly divided into groups of three. Each mouse was inoculated (intragastric, percutaneous injection with a 30 1/2 gauge needle) with 0.1 m1 of a crude culture filtrate or control material containing 2 drOps of 2% Evans Blue dye/ml. After exposure, mice were placed in boxes divided into compartments (6 x 10 x 6 cm deep) and kept at 24 room temperature. Two and 1/2 h after inoculation the mice were killed by cervical dislocation, the abdomen opened, and the entire intestine (not including the stomach) removed with a hemostat. The intestine from each mouse was weighed. The ratio of gut weight to remaining carcass weight was cal- culated and an average value was obtained. Each experimental sample was assayed using three mice and each experiment was at least done in duplicate. Activity in the assay is expressed as the ratio of intestinal weight/body weight (IW/BW). Values greater than or equal to 0.084 were con- sidered positive, values less than 0.075 negative and those between 0.075 and 0.084 indeterminate. These values are comparable to those reported by other investigators (30, 51). Statistics The mean of IW/BW ratios : 1 standard error of the mean for each sample tested was calculated (5). RESULTS Production of Enterotoxin-Like Activity Seventeen Salmonella isolates from horses (Table l) were examined for enterotoxin-like activity in the suckling mouse assay. Cultures grown for 18 h in BHI broth in roller tubes were centrifuged at 12,100 x g for 20 min. Non-sterile supernates were assayed immediately after harvesting cells (Table 2). A positive response (IW/BW 1 0.084) of fluid accumulation was found in 9 out of the 17 isolates. Of these, 6 were Salmonella typhimurium, 2 were Salmonella anatum, and one was Salmonella cholerasuis. Only one strain, Salmonella anatum 2CH, was clearly negative. The remaining 7 isolates were indeterminate (0.075 to 0.084) by IW/BW ratios. Some physiological parameters of this toxin activity and its production are described below. In subsequent tests, only positive strains 200H, 20CH, T24CH, E9288 and NADL and negative strain 2CH were used. Temperature Stability of Enterotoxin- Like Activity Culture supernates of the positive Salmonella species were frozen at -70 C not less than 1 or more than 6 weeks. 25 26 Table 1.--Salmonella Species and Source of Isolation. VCase Species Source .umber ZSCH E, anatum Asymptomatic carrier 2CH E, anatum Asymptomatic carrier SCH E, typhimurium Asymptomatic carrier 22CH E, typhimurium Asymptomatic carrier 875H .§' typhimurium Asymptomatic carrier, stress 18CH E, typhimurium Asymptomatic carrier, stress 166H E, anatum Stress, intermittent diarrhea 017H E, anatum Stress, diarrhea, septicemia £312,153 E, anatum Stress, acute diarrhea 20CH E, typhimurium Stress, acute diarrhea, foal 200H E, anatum Stress, acute diarrhea 844H E, typhimurium Stress, severe diarrhea, septicemia 184H E, anatum Chronic diarrhea A24CH E, anatum Stress, chronic diarrhea, fatal T24CH E, typhimurium Stress, chronic diarrhea, fatal E9288 .§° typhimurium Fatal NADL E, cholerasuis Unknown 27 Table 2.--Production of Enterotoxin-like Activity in BHI Roller Tube Cultures of Salmonellaa. E2:::3§:;in Case Number IW/BW Ratio Negative Broth Control .072 :_.003 (10) 2CH .071 :_.003 (5) Indeterminate A24CH .075 :_.002 (3) 3CH .077 :.°003 (2) E312,153 .077 I .004 (2) ZSCH .078 :_.006 (2) 22CH .078 :_.002b 017H .080 :_.003 (3) 166H .083 :_.002 (4) Positive 875H .084 :_.005 (5) 200B .086 :_.001 (4) 184H .089 :_.007 (2) ZOCH .090 :_.006 (3) T24CH .092 :_.005 (6) 844H .093 :_.005 (3) E9288 .097 :_.002 (7) 18CH .100 :_.004 (2) NADL .100 + .005 (5) aFigures represent mean + 1 standard error of the mean of at least 2 separate assays of toxin—preparations tested immediately upon harvesting filtrates. Numbers in parentheses indicate the number of tests. Each test used 3 mice. bFigure represents mean :_1 standard deviation. 28 Thawed preparations were tested for toxin activity in the infant mouse assay. Freeze-thaw ratios over the entire storage period were equivalent and consequently the data has been grouped. The data (Table 3) show that Salmonella enterotoxin-like activity is not stable to freezing. Roller culture supernates of 3 Salmonella strains possessing enterotoxin-like activity were subjected to 100 C for 30 min. The toxic factor was stable to heat (Table 4). Production of Enterotoxin-Like Activity Under Various Condi- tions of Bacterial Growth An overnight culture of E. Eyphimurium E9288 was inoculated into 70 ml of filtered BHI (initial 0. D. 0.060, 520 nm) and incubated in a Gyrotory shaker water bath at 37 C. The storage culture of E9288 was also inoculated into 5 ml of BHI and incubated in a roller tube apparatus at 37 C. Growth curves were plotted for both cultures (Figure l). Shaker cultures of E9288 entered logarithmic phase of growth immediately, whereas a lag in cell division occured in roller cultures. By the 6th to 8th h, both cultures reached the stationary phase, with final cell num- ber greater in the shaker culture. Culture supernatant fluids from E9288 were titrated for toxin activity during various stages of growth in shaker and roller tube cultures. The relative toxin level in the culture supernates was determined by use of serial twofold dilutions of the toxin in Tris buffer. Experimentally, at 29 Table 3.--Loss of Enterotoxin-like Activity After a 6 Week Period of Storage at -70 Ca. IW/BW Ratio Case Number Before After 200B .086 I .001 (4) .069 :_.002 (3) 20CH .090 :_.006 (3) .080 :_.003 (3) T24CH .092 I .005 (6) .073 I.'007 (2) E9288 .097 :_.002 (7) .080 :_.004 (3) NADL .100 :.°005 (5) .081 :_.003 (3) E_. glib NTC .100 1.010 (2) Broth Control .072 :_.003 (10) .070 + .006 (3) aFigures represent mean + 1 standard error of the mean of at least 2 separate assays of toxin—preparations grown in BHI in roller tubes for 18 h. Numbers in parentheses indicate the number of tests. Each test used 3 mice. bE, coli grown in CA-YE and BHI broth in roller tubes for 18 h. cNot tested. 30 Table 4.--Retention of Enterotoxin-like Activity after Heat Treatmenta. IW/BW Ratio Case Number No Heat 100 C, 30 min 20CH .084 :_.010 (2) .088 :_.006 (3) E9288 .096 :_.006 (3) .094 :_.002 (S) 844Hb .098 :_.003 .102 :_.007 Broth Control .074 + .003 (S) .078 :_.004 (2) aFigures represent mean :_1 standard error of the mean of at least 2 separate assays of toxin preparations grown in 5 m1 BHI broth in roller tubes. Numbers in parentheses indicate the number of tests. Each test used 3 mice. bFigures represent mean :_1 standard deviation. Figure l. 31 Growth Curve of E. typhimurium E9288 in Roller Culture (0) and Shaker Culture (A). LOG'O VIABLE C ELLS/ML 32 8- 7 e E .L 3 J 3 I2 I6 20 TIME (HOURS) FWRE I 33 designated intervals, 5 ml of the shaker culture were removed or a roller tube culture was harvested. The last dilution giving an IW/BW ratio greater than or equal to 0.084 was considered positive and was plotted against time (Figure 2). At least 2 separate assays of toxin prepara- tions were tested at each time point for each dilution. Maximum enterotoxin-like activity was observed in culture supernates prepared from 8 to 14 h roller cultures and from 6 h shaker cultures. No enterotoxin-like activity could be demonstrated in 18 h shaker cultures compared to 18 h roller cultures of Salmonella. Loss of Enterotoxin-Like Activity Upon Filter Sterilization of Samples One ml of crude culture supernate prepared from 18 h roller tube cultures, of toxigenic E9288 and non-toxigenic 2CH, was filtered through a Millipore membrane. The sterile filtrate was examined for enterotoxin-like activity in suckling mice. The used Millipore filter was resuspended in 1 ml of BHI and tested for toxin activity (Table 5). Salmonella typhimurium E9288 preparations lost their entero- toxin-like activity upon filtration. The factor responsible for the activity observed in the mice was retained by the Millipore filter. Both the sterile filtrate of Salmonella anatum 2CH and the membrane used to filter 2CH failed to exhibit enterotoxin-like activity. This strain was also negative in the unfiltered preparations (Table 2 and 5). 34 Figure 2. Titration of E. typhimurium E9288 Enterotoxin- like Activity in Roller Culture (0) and Shaker Culture (A). 35 .- . J-1 8 lea-I)- |34 ... 2 A>:>_._.o< wx_4..z.xo._.ommhzwv m m .5 h 20 TIME (HOURS) FIGURE 2 36 Table S.--Loss of Enterotoxin-like Activity After Filter Sterilization of Supernatant Fluidsa. C Crude Filtrate Sterile Filtrate 358 Number IW/BW Ratio Countb IW/BW Ratio E9288 .097 :_.002 (7) 1.9 x 106 .071 :_.002 (7) filterc .097 :_.004 (3) E9288 2CH .071 1 .003 (5) 3.2 x 104 .064 :_.000 (2) filterc .070_: .001 (2) 2CH aFigures represent mean + 1 standard error of the mean of at least 2 separate assays of toxin—preparations from 18 h roller tube cultures. Numbers in parentheses indicate the number of tests. Each test used 3 mice. bAverage number viable Salmonella/0.1 ml supernatant fluids. cMillipore filter (used to sterilize 1.5 m1 of supernate) resuspended in 1 m1 of BHI. 37 Activity of Lipopolysaccharide (LPS) in the Suckling Mouse Assay The possibility of LPS being associated with enterotoxin-like activity in the culture supernatant fluids of Salmonella led to testing of commercially prepared LPS in the infant mouse assay. Doses up to l mg/mouse failed to enhance fluid secretion (Table 6). Table 6.-—Inability of Salmonella LPS to Stimulate Enterotoxin-like Activity in Suckling Mice Dose (Hg) IW/BW Ratio 62.75 .077 i .002 (3) 125.00 .073 i .002 (3) 250.00 .079 i .002 (3) 500.00 .073 i .001 (3) 1000.00 .076 i .003 (3) 0.85% NaCl .067 i .001 (5) aFigures represent mean + 1 standard error of the mean of at least 2 separate assays of LPS resuspended in 0.85% saline. Numbers in parentheses indicate the number of tests. Each test used 3 mice. Activity of Live Salmonella in the Suckling Mouse Assay An investigation of the number of viable organisms remaining in the enterotoxin preparations was undertaken. Bacterial counts were determined for 18 h roller tube culture supernates. Salmonella typhimurium E9288 had an 38 average of 1.9 x 107 viable cells/ml of supernate. Sal- monella anatum 2CH had an average of 3.2 x lOS/ml of super- nate (Table 5). The sedimented cells from E9288 and 2CH were washed twice in physiological saline and resuspended in 5 ml of saline. Salmonella were diluted and the resulting cell suspensions were examined for enterotoxin-like activity in suckling mice (Table 7). Greater than 4.0 x 106 E. typhi- murium E9288/mouse were necessary to stimulate intestinal secretion. Salmonella anatum 2CH were unable to elicit a positive response in the mice, even at doses as high as 1.8 x 108 bacteria per mouse. 39 Table 7.--Reactivity of Suckling Mice to Washed Cells of Salmonellaa. IW/BW Ratio Doseb E9288 8 8 2.5 x 10 - 7.8 x 10 .093 i .002 (4) 1 1 x 107 - 7 8 x 107 .091 i .003 (5) 1.1 x 106 - 4.0 x 106 .067 i .006 (3) 1.1 x 102 - 1.1 x 105 .063 i .001 (4) 2CH 1.8 x 108 .067 i .009C 1.0 x 107 - 6.0 x 107 .072 i .002 (3) 1.0 x 106 - 6.0 x 106 .063 i .002 (3) 1.0 x 105 - 6.0 x 105 .065 :_.002 (2) aFigures represent mean i 1 standard error of the mean of at least 2 separate assays of washed cell prepara- tions. Numbers in parentheses indicate the number of tests. Each test used 3 mice. bDose range recorded as the number of bacteria injected per mouse. CFigure represent mean i 1 standard deviation. DISCUSSION Swine, calves and lambs suffer from an acute diarrheal disease brought on by toxigenic E. coli. Horses experience a similar clinical syndrome, frequently asso- ciated with Salmonella, but enterotoxin-like activity has never been associated with equine salmonellosis. Using strains of human origin, Koupal and Deibel (51) have shown that the suckling mouse assay is sensitive to Salmonella enterotoxin. This assay was chosen to survey for toxin activity in the 8 strains of E. typhimurium, 8 strains of S. anatum and one strain of E. cholerasuis. For E. coli, 16 to 24 h roller tube cultures in CA-YE is Optimal for toxin production (30). In this study BHI broth was used since most investigators find this medium best for screening Salmonella for enterotoxin (51, 75). In the present study, 9 strains were shown to produce enterotoxin-like activity in the suckling mouse assay. Six (875H, 18CH, 844H, 20CH, T24CH, and E9288) out of the 8 E. typhimurium cultures gave positive responses. Of these, 4 were from diarrheal cases and 2 were from carriers. Two (200H and 184H) out of the 8 E. anatum cultures gave positive responses, both isolated from diarrheal cases. The 40 41 S. cholerasuis also gave a positive response. No case his- tory was available. Enterotoxin-like activity was not demon- strated in 8 cultures. Half were from carriers and the other half from diarrheal cases. It appears that no direct correlation between the ability of Salmonella to produce an enterotoxin—like reaction and severity of the disease state can be made. Knowledge of the effects of various storage condi- tions on enterotoxin activity is critical because suckling mice are often not available when the samples are prepared. E. ggli ST is stable at -20 C for up to 6 months (30) and is stable at -70 C up to 3 weeks (13). According to Koupal and Deibel (51), E. enteriditis toxin is stable at 4 C, but pro- gressively less stable when frozen at -20 C. In this study, toxigenic supernates were stored at ~70 C and assayed for activity over the following 6 weeks. Enterotoxin-like activity could not be demonstrated suggesting that toxin activity was not stable to freezing. The toxicity of cholera enterotoxin is also lost after freezing (Dr. G. Yang, 1978, personal communication). The enterotoxins of E. 221E are differentiated primarily by their reactivity after heating. The heat-stable toxin is stable at 100 C but is destroyed by autoclaving (86). The heat-labile toxin is inactivated by heating to 60 C for 30 min (39). Sandefur and Peterson (75) using Salmonella observed that the rapid permeability factor in 42 rabbit skin tests was heat-stable (100 C) up to 4 hours and the delayed permeability factor was heat—labile (75 C, 30 min). Infant mouse assays were not performed by these investigators. Koupal and Deibel (51) characterized an enterotoxic factor in suckling mice. Activity was lost at 80 C. Heat reactivity of the toxin—like activity of EEEEQ- BELLE was examined in this study. Culture supernates of 3 different Salmonella were exposed to 100 C for 30 min. Positive IW/BW ratios were obtained indicating that the toxins examined were heat-stable. Since vegetative cells in the supernatant fluids are killed within minutes by exposure to boiling water, something other than live EElEQ' EEllE.mUSt have been responsible for the enterotoxin-like activity observed after this treatment. Sedlock et a1. (80), using the rabbit ileal loop assay, noted that not all media exhibited an equivalent ability to support production of Salmonella enterotoxin. Likewise, variability with regard to production of rapid permeability factor produced by Salmonella in different media was observed by Sandefur and Peterson (75). Repro- ducible results were only obtained when exponential phase cultures of Salmonella were tested in suckling mice (51). Giannella noted similar effects with E. coli in infant mice (30). The dependence of enterotoxin-like activity on cultural conditions of Salmonella was, therefore, examined. In this study, enterotoxin-like activity in 43 roller culture tubes was detected by the 4th hour of growth. The appearance of activity may depend on cultures reaching a critical cell number (1.9 x 108/m1 at 4th hour) before toxin can be detected. Enterotoxin-like activity peaked in the stationary phase of growth (8 to 14th hour) where one log increase in cell number was noted (Figure 1). Maximum levels of toxin were detected sooner in shaker culture (6 h), but it dropped at a much faster rate than in roller culture. Activity was no longer detectable at 18 h. Sandefur and Peterson (75) discovered an inhibitor-like substance in culture supernatants which masked the effects of the delayed permeability factor in rabbit skin tests. Whether or not a similar inhibitor exists in this system is not known. Some confusion exists in the literature regarding the need for presence of viable organisms for successful reactivity of enterotoxin assays. Giannella (30) reported that filtered and unfiltered preparations of E. 32;; enterotoxin react identically. Sedlock and Deibel have shown that sterile culture filtrates of Salmonella do not consistently stimulate fluid secretion in rabbit loops (79) but that reproducible results can be obtained if invasive, nontoxin producing Salmonella are injected concurrent with sterile toxin-containing preparations. They suggest the invasive process reduces the presence of intestinal mucin which physically inhibits the enterotoxin from reaching its 44 target cells. In this study 2 to 5 ml of Salmonella preparations were filtered through a 0.45 pm membrane 13 mm in diameter. The filters clogged rapidly. Enterotoxin-like activity could not be demonstrated in the sterile filtrates. That the filter did hold back the fluid accumulating prin- ciple was shown when positive IW/BW ratios were obtained with the suspension made from the membrane. Viable Egigge ggllg_are present on the filter. Giannella et a1. (33) have presented substantial evidence which supports the concept that penetration of the epithelial surface by the bacteria is a necessary step in the pathogenesis of infection. Their studies indicate that strains lacking this ability do not cause fluid accumulation in rabbit loops. The possibility exists that invasion may play a role in stimulating enterotoxin-like activity in the suckling mice. To determine whether there was any correlation between the number of viable bacteria/m1 of supernatant fluid and the ability to promote fluid accumulation, bac- terial counts were made. E9288, a strain possessing enterotoxin-like activity, had an average of 1.9 x 107 viable cells/ml of supernate. 2CH, a strain never demon- strating activity, had an average of 3.2 x lOS/ml. The 2 log difference was not responsible for the differences in reactivity since as many as 1.8 x 108 E. anatum 2CH were not able to elicit fluid accumulation. Hence, cell number alone is not responsible for the enterotoxin-like activity 45 observed herein. Even when the membrane used to sterilize 2CH was resuspended in BHI and assayed in the mice, activity was not observed. An apparent difference exists between E. anatum 2CH and E. typhimurium E9288. Crude toxin preparations of E9288 (an average of 1.9 x 106 bacteria/0.1 ml) stimulated fluid secretion. When up to 4 x 106 washed E9288 were injected per mouse, fluid accumulation was not observed. Higher doses of EEEEQ- EEllE E9288 did, however, induce positive IW/BW ratios. Koupal and Deibel (51) determined that the active entero- toxin material in Salmonella originates from the outer mem- brane of Salmonella and that it is released during growth. The difference in bacterial number and the fact that entero- toxin—like activity was retained by Millipore filters sug- gest that enterotoxin is primarily cell bound and is released into the growth medium in low concentrations. Since endotoxin is present in all 17 preparations and only 9 gave positive responses in the mouse assay, a causal relationship for LPS in diarrhea production could be tentatively eliminated. To confirm this assumption further, commercially prepared endotoxin was tested in infant mice. Negative responses were observed with up to 1 mg of LPS. Previous investigations with E. coli and Salmonella endotoxin have also demonstrated that LPS does not provoke positive responses (39, 51, 86). Salmonella multiplication and the presence of live bacteria in the intestine appears essential for the 46 development of the diarrheal syndrome. The precise role of enterotoxin needs additional characterization but the present study clearly supports the notion that toxins dis- tinct from lipopolysaccharide exist in Salmonella and con- ceivably play a role in enteric disease. LITERATURE CITED 10: LI TERATURE C ITED Alderete, J. F., and D. C. Robertson. 1978. Purifi- cation and Chemical Characterization of the Heat- Stable Enterotoxin Produced by Porcine Strains of Enterotoxigenic Escherichia coli. Infect. Immun. l2: 1021-1030. Axon, A. T. R., and D. Poole. 1973. Salmonellosis Presenting with Cholera-Like Diarrhea. Lancet. 745-746. Baker, J. R. 1970. Salmonellosis in the Horse. Brit. Vet. J. 126: 100-105. Basu, S., and M. J. Pickett. 1969. Reaction of Vibrio cholerae and Choleragenic Toxin in Ileal Loop of Laboratory Animals. J. Bact. 100: 1142-1143. Bhattachanjya, G. K., and R. A. Johnson. 1977. Infer- ences About a Population. p. 233-243. In: Stati- stical Concepts and Methods. New York: John Wiley and Sons, Inc. Black, P. H., L. J. Kunz, and M. N. Swartz. 1960. Salmonellosis - A Review of Some Unusual Aspects. New Eng. J. Med. 262: 811-817. Bryans, J. T., E. H. Fallon, and B. O. Shephard. 1961. Equine Salmonellosis. Cornell Vet. EE: 467-477. Carpenter, C. C. J. 1972. Cholera and Other Entero- toxin-Related Diarrheal Diseases. J. Infect. Dis. 126: 551-564. ' Cordy, D. R., and R. W. Davis. 1946. An Outbreak of Salmonellosis in Horses and Mules. J. Amer. Vet. Med. Assoc. 108: 20-24. Craig, J. P. 1965. A Permeability Factor (Toxin) Found in Cholera Stools and Culture Filtrates and its Neutralization by Convalescent Cholera Sera. Nature. E21: 614-616. 47 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 48 Dack, G. M., E. 0. Jordan, and W. L. Wood. 1929. "Food Poisoning" Produced in Monkeys by Feeding Living Salmonella Cultures. J. Prevent. Med. 3: 153-158. — De, S. N., and D. N. Chatterje. 1953. An Experimental Study of the Mechanism of Action of Vibrio cholerae on the Intestinal Mucous Membranes. J. Path. Bact. EE; 559-562. Dean, A. G., Y. Ching, R. G. Williams, and L. B. Harden. 1972. Test for Escherichia coli Entero- toxins Using Infant Mice: Application in a Study of Diarrhea in Children in Honolulu. J. Infect. Dis. EEE: 407-411. Donta, S. T., M. King, and K. Sloper. 1973. Cholera Enterotoxin Induction of Steroidogenesis in Tissue Culture. Nature. 243: 246-247. Donta, S. T., H. W. Moon, and S. C. Whipp. 1974. Detection of Heat-Labile Escherichia coli Entero- toxin with the Use of Adrenal Cells in Tissue Culture. Science. EEE: 334-335. Dorn, C. R., J. R. Coffman, D. A. Schmidt, H. E. Garner, J. B. Addison, and E. L. McCune. 1975. NeutrOpenia and Salmonellosis in Hospitalized Horses. J. Amer. Vet. Med. Assoc. EEE: 65-67. Dorner, F. 1975. Escherichia coli Enterotoxin Puri- fication and ParEial Characterization. J. Biol. Chem. 250: 8712-8719. Edwards, P. R. 1958. Salmonellosis: Observations on Incidence and Control. Ann. N. Y. Acad. Sc. 19: 598-613. Evans, D. G., D. J. Evans, and S. L. Gorbach. 1973. Identification of Enterotoxic Escherichia coli and Serum Antitoxin Activity by the Vascular Perme- ability Factor Assay. Infect. Immun. E: 731-735. Evans, D. G., D. J. Evans, and N. F. Pierce. 1973. Differences in the Response of Rabbit Small Intes- tine to Heat-Labile and Heat-Stable Enterotoxins of Escherichia coli. Infect. Immun. 1: 873-880. Evans, D. G., R. P. Silver, D. J. Evans, D. G. Chase, and S. L. Gorbach. 1975. Plasmid-Controlled Colo- nization Factor Associated with Virulence in Esche- richia coli Enterotoxigenic for Humans. Infect. Immun. lg: 656-667. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 49 Evans, D. J., L. C. Chen, G. T. Curlin, and D. G. Evans. 1972. Stimulation of Adenyl Cyclase by Escherichia coli Enterotoxin. Nature N. Biol. EEE: 137-138. Evans, D. J., D. G. Evans, and S. L. Gorbach. 1973. Production of Vascular Permeability Factor by Enterotoxigenic Escherichia coli Isolated from Man. Infect. Immun. E: 725-730. Evans, D. J., D. G. Evans, and S. L. Gorbach. 1974. Polymyxin B-Induced Release of Low-Molecular-Weight, Heat-Labile Enterotoxin from Escherichia coli. Infect. Immun. 19‘ 1010-1017. Field, M. 1971. Ion Transport in Rabbit Ileal Mucosa. II. Effects of Cyclic 3', 5'-AMP. Am. J. Physiol. 221: 992-997. Field, J., D. Fromm, Q. Al-Awqati, and W. B. Greenough. 1972. Effect of Cholera Enterotoxin on Ion Trans- port Across Isolated Ileal Mucosa. J. Clin. Invest. El: 796-804. Finkelstein, R. A., J. J. LoSpalluto. 1969. Pathoge- nesis of Experimental Cholera: Preparation and Iso- lation of Choleragen and Choleragenoid. J. Exp. Med. E32: 185-202. Fromm, D., R. A. Giannella, S. B. Formal, R. Quijano, and H. Collins. 1974. Ion Transport Across Iso- lated Ileal Mucosa Invaded by Salmonella. Gastro. EE: 215-225. Gartner. 1888. Pathogene and Saprophytische Bacterien in ihrem Verhéltniss zum Wasser, insonderlich zum Trinkwasser. Cor.-B1. D. allg. érztl. ver. v. Thuringen, Weimar. El: 233-245. Giannella, R. A. 1976. Suckling Mouse Model for Detection of Heat-Stable Escherichia coli Entero- toxin: Characteristics of the Model. Infect. Immun. 14: 95-99. Giannella, R. A., S. A. Broitman, and N. Zamcheck. 1971. Salmonella Enteritis. I. Role of Reduced Gastric Secretion in Pathogenesis. Am. J. Dig. Dis. EE: 1000-1006. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 50 Giannella, R. A., S. A. Broitman, and N. Zamcheck. 1971. Salmonella Enteritis. II. Fulminant Diarrhea in and Effects on the Small Intestine. Am. J. Dig. Dis. EE: 1007-1013. Giannella, R. A., S. B. Formal, G. J. Dammin, and H. Collins. 1973. Pathogenesis of Salmonellosis. Studies of Fluid Secretion, Mucosal Invasion, and Morphologic Reaction in the Rabbit Ileum. J. Clin. Invest. EE: 441-453. Giannella, R. A., R. E. Gots, A. H. Charney, W. B. Greenough, and S. B. Formal. 1975. Pathogenesis of Salmonella-Mediated Intestinal Fluid Secretion. Activation of Adenylate Cyclase and Inhibition by Indomethacin. Gastro. E2: 1238-1245. Guerrant, R. L., L. L. Brunton, T. C. Schnaitman, L. I. Rebhun, and A. G. Gilman. 1974. Cyclic Adeno- sine Monophosphate and Alteration of Chinese Hamster Ovary Cell Morphology: a Rapid, Sensitive EE_Vitro Assay for the Enterotoxins of Vibrio cholerae and Escherichia coli. Infect. Immun. 10: —320-—327‘. " Gustafson, D. P., and E. H. Page. 1976. Chronic Equine Diarrhea. Proc. Am. Assoc. Equine Practnr. 167-175. Gyles, C. L. 1971. Discussion: Heat-Labile and Heat- Stable Forms of the Enterotoxin from E. coli Strains Enteropathogenic for Pigs. Ann. N. Y. Acad. Sci. llé: 314-322. Gyles, C. L. 1974. Immunological Study of the Heat- Labile Enterotoxin of Escherichia coli and Vibrio cholerae. Infect. Immun. E: 564-570. Gyles, C. L., and D. A. Barnum. 1969. A Heat-Labile Enterotoxin from Strains of Escherichia coli Entero- pathogenic for Pigs. J. Infect. Dis. 120: 419-426. Gyles, C. L., M. So., and S. Falkow. 1974. The Enterotoxin Plasmids of Escherichia coli. J. Infect. Dis. 130: 40-49._ Jones, G. W., and J. M. Rutter. 1972. Role of K88 Antigen in the Pathogenesis of Neonatal Diarrhea Caused by Eschegichia coli in Piglets. Infect. Immun. E: 918-927. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 51 Kantor, H. S., P. Tao, and S. L. Gorbach. 1974. Stimulation of Intestinal Adenyl Cyclase by Escheri- chia coli Enterotoxin: Comparison of Strains from an Infant and Adult with Diarrhea. J. Infect. Dis. 129: 1-9. Kent, T. H., S. B. Formal, and E. H. LaBrec. 1966. Acute Enteritis Due to Salmonella gyphimurium in Opium-Treated Guinea Pigs. Arch. Path.EE: 501-508. Kent, T. H., S. B. Formal, and E. H. LaBrec. 1966. Salmonella Gastroenteritis in Rhesus Monkeys. Arch. Path. EE: 272-279. Kimberg, D. V., M. F. Field, J. Johnson, A. Henderson, and E. Gershon. 1971. Stimulation of Intestinal Mucosa Adenyl Cyclase by Cholera Enterotoxin and Prostaglandins. J. Clin. Invest. E9: 1218-1230. Kimberg, D. V., M. F. Field, E. Gershon, and A. Hender- son. 1974. Effects of Prostaglandins and Cholera Enterotoxin on Intestinal Mucosal Cyclic AMP Accumu- lation. Evidence Against an Essential Role for Prostaglandins in the Action of the Toxin. J. Clin. Invest. EE: 941-949. Klipstein, F. A., and R. F. Engert. 1976. Purifica- tion and PrOperties of Klebsiella pneumoniae Heat- Stable Enterotoxin. Infect. Immun. EE: 373-381. Klipstein, F. A., and R. F. Engert. 1976. Partial Purification and Properties of Enterobacter cloacae Heat-Stable Enterotoxin. Infect. Immun. EE: 1307- 1314. Klipstein, F. A., and R. F. Engert. 1977. Immunologi- cal Interrelationships Between Cholera Toxin and the Heat-Labile and Heat-Stable Enterotoxins of Coliform Bacteria. Infect. Immun. EE: 110-117. Klipstein, F. A., I. R. Horowitz, R. F. Engert, and E. A. Schenk. 1975. Effect of Klebsiella pneu- moniae Enterotoxin on Intestinal Transport in the Rat. J. Clin. Invest. §§: 799-807. Koupal, L. R., and R. H. Deibel. 1975. Assay, Char- acterization, and Localization of an Enterotoxin Produced by Salmonella. Infect. Immun. 11: 14-22. Lariviere, S., C. L. Gyles, and D. A. Barnum. 1973. Preliminary Characterization of the Heat-Labile Enterotoxin of Escherichia coli F11(P155). J. Infect. Dis. EEE: 312-320. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 52 LoSpalluto, J. J., and R. A. Finkelstein. 1972. Chemical and Physical Properties of Cholera Exo- enterotoxin (Choleragen) and its Spontaneously Formed Toxoid (Choleragenoid). Biochim. Biophys. Acta. 2E1: 158-166. Maenza, R. M., D. W. Powell, G. R. Plotkin, S. B. Formal, H. R. Jarvis, and H. Sprinz. 1970. Experi- mental Diarrhea: Salmonella Enterocolitis in the Rat. J. Infect. Dis. 121: 475-485. McCullough, N. B., and W. Eisele. 1951. Experimental Human Salmonellosis. I. Pathogenicity of Strains of E. meleagridis and E. anatum Obtained from Spray- Dried Whole Egg. J. Infect. Dis. 22: 278-289. McCullough, N. G., and W. Eisele. 1951. Experimental Human Salmonellosis. III. Pathogenicity of Strains of E. newEort, E. derby, and E. bareilly Obtained from Spray-Dried Whole Egg. J. Infect. Dis. 22; 209-213. Moon, H. W., P. Y. Fung, S. C. Whipp, and R. E. Isaac- son. 1978. Effects of Age and Ambient Temperature on the Response of Infant Mice to Heat-Stable Entero- toxin of Escherichia coli: Assay Modifications. Infect. Immun. 22: 36-39. Morse, E. V., M. A. Duncan. 1974. Salmonellosis--An Environmental Health Problem. J. Amer. Vet. Med. Assoc. 165: 1015-1019. Morse, E. V., M. A. Duncan, J. F. Fessler, and E. H. Page. 1976. The Treatment of Salmonellosis in Equidae. Mod. Vet. Practice. El: 47-51. Morse, E. V., M. A. Duncan, E. A. Page, and J. F. Fessler. 1976. Salmonellosis in Eguidae: A Study of 23 Cases. Cornell Vet. 22: 198- 13. Olsen, N. E. 1966. Acute Diarrheal Disease in the Horse. J. Amer. Vet. Med. Assoc. 148: 418-421. Orskov, I., F. Orskov, H. W. Smith, and W. J. Sojka. 1975. The Establishment of K99, a Thermolabile, Transmissable Eschericia coli K Antigen, Previously Called "Kco", Possessed by Calf and Lamb Entero- pathogenic Strains. Acta. Pathol. Microbiol. Scand. 83: 31-36. Owen, R. 1975. Post Stress Diarrhea in the Horse. Vet. Rec. 22: 267-270. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 53 Pai, C. H., and V. Mors. 1978. Production of Entero- toxin by Yersinia enterocolitica. Infect. Immun. 22: 908-911. Pierce, N. F., W. B. Greenough, and C. C. Carpenter. 1971. Vibrio cholerae Enterotoxin and its Mode of Action. Bact. Rev. 2E: 1-13. Platt, H. 1976. Septicemia in the Foal. A Review of 61 Cases. Brit. Vet. J. 129: 221-229. Powell, D. W., G. R. Plotkin, R. M. Maenza, L. I. Solberg, D. H. Catlin, and S. B. Formal. 1971. Experimental Diarrhea. I. Intestinal Water and Electrolyte Transport in Rat Salmonella Enterocoli- tis. Gastro. 22: 1053-1064. Rooney, J. 1977. The Newborn Foal. p. 76-77. In: The Sick Horse. New York: A. S. Barnes and Co., Inc. Rooney, J. R., J. T. Bryans, and E. R. Doll. 1963. Colitis "X" of Horses. J. Amer. Vet. Med. Assoc. 142: 510-511. Rout, R., R. A. Giannella, R. S. Formal, and G. Damin. 1973. The Pathophysiology of Salmonella Diarrhea in Rhesus Monkeys. Gastro. 22: 793. Abstract. Sack, R. B. 1975. Human Diarrheal Disease Caused by Enterotoxigenic Escherichia coli. Ann. Rev. Micro- biol. 22: 333-353. Sakazaki, R., and S. Miura. 1956. The Enteric Bac- terial Flora of the Intestinal Tract of Healthy Horses. Jap. J. Vet. Res. 2: 59-63. Sakazaki, R., K. Tamura, A. Nakamura, and T. Kurata. 1974. EnterOpathogenic and Enterotoxigenic Activi- ties on Ligated Gut Loops in Rabbits of Salmonella and Some Other Enterobacteria Isolated from Human Patients with Diarrhea. Jap. J. Med. Sci. Bio. 21: 45-48. Notes. Sakazaki, R., K. Tamura, A. Nakamura, T. Kurata, A. Gohda, and Y. Kazuno. 1974. Studies on Entero- pathogenic Activity of Vibrio parahaemolyticus using Ligated Gut Loop Models in Rabbits. Jap. J. Med. Sci. Biol. 21; 35-43. Sandefur, P. D., and J. W. Peterson. 1976. Isolation of Skin Permeability Factors from Culture Filtrates of Salmonella typhimurium. Infect Immun. 22: 671- 679. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 54 Sandefur, P. D., and J. W. Peterson. 1977. Neutrali- zation of Salmonella Toxin-Induced Elongation of Chinese Hamster Ovary Cells by Cholera Antitoxin. Infect. Immun. EE: 988-992. Saphra, I., and J. W. Winter. 1957. Clinical Mani- festations of Salmonellosis in Man: Evaluation of 7779 Human Infections Identified at New York Salmo- nella Center. New Eng. J. Med. 2E2: 1128-1134. Schenkein, I., R. F. Green, D. S. Santos, and W. K. Mass. 1976. Partial Purification and Characteri- zation of a Heat-Labile Enterotoxin of Escherichia coli. Infect. Immun. 22: 1710-1720. Sedlock, D. M., and R. H. Deibel. 1978. Detection of Salmonella Enterotoxin Using Rabbit Ileal Loops. Can. J. Microbiol. 22: 268-273. Sedlock, D. M., L. R. Koupal, and R. H. Deibel. 1978. Production and Partial Purification of Salmonella Enterotoxin. Infect. Immun. 22: 375-380. Seligman, E., I. Saphra, and Wasserman. 1943. Salmo- nella Infection in Man: Analysis of 1000 Cases Bac- teriologically Identified by New York Salmonella Center. Am. J. Hyg. 22: 226-249. Smith, H. W. 1972. The Production of Diarrhea in Baby Rabbits by the Oral Administration of Cell-Free Preparations of EnterOpathogenic Escherichia coli and Vibrio cholerae: The Effect of Antisera. J. Med. Microbiol. E: 299-303. Smith, H. W., C. L. Gyles. 1970. The Relationship Between Two Apparently Different Enterotoxins Pro- duced by Enteropathogenic Strains of Escherichia coli of Porcine Origin. J. Med. Microbiol. 2: 387-401. Smith, H. W., and C. L. Gyles. 1970. The Effect of Cell-Free Fluids Prepared from Culture of Human and Animal Enteropathogenic Strains of Escherichia coli on Ligated Intestinal Segments of Rabbits and Pigs. J. Med. Microbiol. 2: 403-409. Smith, H. W., and S. Halls. 1967. Observations by the Ligated Intestinal Segment and Oral Inoculation Methods on Escherichia coli Infections in Pigs, Calves, Lambs and Rabbits. J. Path. Bacteriol. 22: 499-529. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 55 Smith, H. W., and S. Halls. 1967. Studies on Escheri- chia coli Enterotoxin. J. Path. Bacteriol. 22: 531-543. Smith, H. W., and S. Halls. 1968. The Transmissable Nature of the Genetic Factor in Escherichia coli that Controls Enterotoxin Production. J. Gen. Microbiol. E2: 319-334. Smith, W. L., and W. E. M. Lands. 1971. Stimulation and Blockage of Prostaglandin Biosynthesis. J. Bio. Chem. 246: 6700-6704. Stark, R. L., and C. L. Dunca. 1972. Purification and Biochemical Properties of Clostridium perfringes Type A Enterotoxin. Infect. Immun. E: 662-673. Stirk, S. A. 1976. Therapy of a Horse with Diarrhea of Unknown Etiology. Equine Vet. J. 2: 86-88. Takeda, T., Y. Takeda, T. Miwatani, and N. Ohtomo. 1978. Detection of Cholera Enterotoxin Activity in Suckling Hamsters. Infect. Immun. 22: 752-754. Takeuchi, A. 1967. Electron Microscope Studies of Experimental Salmonella Infection. 1. Penetration into the Intestinal Epithelium by Salmonella typhi- murium. Am. J. Path. E2: 109-136. Takeuchi, A., and H. Sprinz. 1967. Electron-Micro- scope Studies of Experimental Salmonella Infection in the Preconditioned Guinea Pig. II. Response of the Intestinal Mucosa to the Invasion by Salmo- nella typhimurium. Am. J. Path. 22: 137-161. Taylor, J., and M. P. Wilkins. 1961. The Effect of Salmonella and Shigella on Ligated Loops of Rabbit Gut. Ind. J. Med. Res. 22: 544-549. Waddell, W. R., and L. J. Kunz. 1956. Association of Salmonella Enteritis with Operations on the Stomach. New Eng. J. Med. 255: 555-559. Waller, S. L. 1973. Prostaglandins and the Gastro- intestinal Tract. Gut. 22: 402-417. Waterman, A. 1977. A Review of the Diagnosis and Treatment of Fluid and Electrolyte Disorders in the Horse. Equine Vet. J. 2: 43-48. nICHIGnN STATE UNIV. LIBRnRIES 1|lHHHHmWIWIWIIHIWIIWHH1| 31293011070376