- Email: [email protected]
c mice
Comparative Immunology, Microbiology & Infectious Diseases 24 (2001) 39±55 www.elsevier.com/locate/cimid
Enteropathogenicity of Plesiomonas shigelloides and Aeromonas spp. in experimental monoand coinfection with Cryptosporidium parvum in the intestine of neonatal BALB/c mice Jiri Vitovec a, Eva Aldova b, Petr Vladik c, Karel Krovacek d,* a
University of South Bohemia, Faculty of Agriculture, Ceske Budejovice, Czech Republic b National Institute of Public Health, Praha, Czech Republic c State Veterinary Institute, Ceske Budejovice, Czech Republic d Department of Veterinary Microbiology, Box 583, Biomedical Centre, 751 23 Uppsala, Sweden
Abstract Enteropathogenicity of Plesiomonas shigelloides, Aeromonas hydrophila, A. caviae and A. sobria was studied both in monoinfections and in coinfections with coccidium Cryptosporidium parvum in neonatal BALB/c mice. In monoinfection experiments, neonatal BALB/c mice were orally infected with 7 107 or 7 108 CFU, respectively, of a strain of P. shigelloides or a strain of an Aeromonas spp. In coinfection experiments, the neonatal mice were, in addition to being orally infected with one of the four bacterial species, orally infected with an inoculum containing 105 oocysts of C. parvum. Results from monoinfections with P. shigelloides revealed long-term colonisation of the neonatal mouse intestine by this pathogen, along with associated pathological lesions. The lesions varied in severity from atrophy to necrosis of the mucosal inner surface of the ileum and colon, with predilection to the colon and brush border of colonic enterocytes. The eects of coinfection of P. shigelloides with C. parvum were characterised by bacteremia and heavy colonisation of the intestine by P. shigelloides. In addition, extensive necrotising in¯ammatory changes in the ileum and colon were accompanied by diarrhoea and deaths of coinfected mice. In contrast, the results from monoinfections of neonatal mice with Aeromonas spp. showed only a short-term colonisation of the intestine by the pathogen. However, when mice were coinfected with A. hydrophila and C. parvum, then the growth of the bacterial species was
* Corresponding author Tel: +46-18-4714592; fax: +46-18-504461. E-mail address: [email protected] (K. Krovacek). 0147-9571/01/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 7 - 9 5 7 1 ( 0 0 ) 0 0 0 1 2 - 6
40
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
prolonged, and occurred in both the spleen and intestine. However, no substantial clinical or histopathological changes were observed in mice, whether monoinfected with Aeromonas spp. or coinfected with C. parvum. Our study suggests that experimental monoinfections of neonatal BALB/c mice with P. shigelloides, Aeromonas spp. and C. parvum, together with coinfections (each bacterial species with the protozoan C. parvum ), may serve as a useful model to study the initial steps of gastrointestinal colonisation and diarrhoeal disease syndromes caused by enteropathogenic bacteria and protozoa, individually and in combination. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Enteropathogenicity; Plesiomonas shigelloides; Aeromonas hydrophila; A. caviae; A. sobria; Cryptosporidium parvum; Experimental infection; Coinfection; Neonatal BALB/c mice
ReÂsume L'eet enteÂropathogeÁne de Pleisomonas shigelloides, Aeromonas hydrophila, A. caviae et A. sobria a eÂte eÂtudie en situation de mono ou co-infection avec le coccidium Crystosporidium parvum dans des souris BALB/c au stade neÂonatal. Dans les cas de monoinfection, les animaux ont eÂte infecteÂs par voie orale avec 7 107 et 7 108 CFU de P. shigelloides ou Aeromonas respectivement. Dans les cas de co-infection, conjointement aÁ l'infection par l'une des quatre espeÁces bacteÂriennes, les animaux ont recu par voie orale eÂgalement un inoculum contenant 105 oocystes d'un protozoaire, le coccidium C. parvum. Les reÂsultats obtenus aÁ partir des mono-infections par P. shigelloides montrent une colonisation aÁ long terme de l'intestin par le pathogeÁne, ainsi que des leÂsions pathologiques associeÂes. Les leÂsions sont de seÂveÂrite variable, allant de l'atrophie aÁ la neÂcrose de la surface interne de la muqueuse de l'ileum et du colon, avec une preÂdilection pour le colon et la bordure apicale des enteÂrocytes du colon. La co-infection avec C. parvum se caracteÂrise par une septiceÂmie et une colonisation majeure de l'intestin par P. shigelloides. D'importantes in¯ammations et neÂcroses de l'ileum et du colon, accompagneÂes de diarrheÂe et conduisant aÁ la mort des souris co-infecteÂes ont eÂgalement eÂte observeÂes. Au contraire, les reÂsultats obtenus aÁ partir des mono-infections avec les dieÂrentes espeÁces d'Aeromonas montrent seulement une colonisation aÁ court terme de l'intestin par ces bacteÂries alors que les souris co-infecteÂes par A. hydrophila et C. parvum preÂsentent une peÂriode de croissance prolongeÂe des bacteÂries dans la rate et l'intestin. Cependant, aucun changement notable sur le plan clinique et histopathologique n'a eÂte observe aussi bien chez les souris mono-infecteÂes par Aeromonas spp. que celles co-infecteÂes avec C. parvum. Notre eÂtude montre donc qu'une infection de souris BALB/c au stade neÂonatal par inoculation orale de P. shigelloides, Aeromonas spp. et C. parvum, ainsi que des co-infections bacteÂries/protozoaires, constitue un treÁs bon modeÁle animal pour l'eÂtude des eÂtapes initiales de colonisation gastrointestinale et des syndromes de diarrheÂe causeÂs par ces pathogeÁnes bacteÂriens et protozoaires. 7 2000 Elsevier Science Ltd. All rights reserved. Mots-cleÂf: EnteÂropathogeÂniciteÂ; Plesiomonas shigelloides; Aeromonas hydrophila; A. caviae; A. sobria; Cryptosporidium parvum; Infection expeÂrimentale; Co-infection; Souris BALB/c neÂonatales
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
41
1. Introduction Plesiomonas shigelloides and Aeromonas species belong to bacterial groups of aquatic pathogens recently recognised as potential human enteropathogens [1±4]. They are Gram negative, motile, non-spore forming bacilli, which are facultatively anaerobic and oxidase positive. According to the Ninth Edition of Bergey's Manual of Determinative Bacteriology, the genus Plesiomonas and Aeromonas spp. are still included in the family of Vibrionaceae [5]. Aeromonas spp. and P. shigelloides are commonly found in a wide range of aquatic environments and foods, and have been isolated from lakes, rivers, coastal waters and drinking water [6±8]. They have also been isolated from humans and domestic animals, such as dogs, cats, goats, sheep and cows [1±4,6±8]. P. shigelloides and the motile Aeromonas spp. have been implicated as agents of human gastroenteritis. They can cause three major types of gastroenteritis: (i) a secretory, watery form, (ii) an invasive, dysentery-like form, and (iii) a subacute or chronic form lasting between 2 weeks and 3 months [3,7,9]. Although extraintestinal infections such as bacteremia, cellulitis, polyarthritis, and meningitis caused by P. shigelloides and Aeromonas spp. are rarely reported, they have been associated with infections in immunocompromised patients and in patients with an underlying disease [1,3,9± 11]. Both P. shigelloides and the motile Aeromonas spp. possess several seemingly pathogenic properties, such as the production of heat-stabile and heat-labile exotoxins, cytotoxin, haemolysin, hemaglutinin and other potentially virulence factors [7,11±15]. However, the exact mechanisms of enteropathogenicity for P. shigelloides and Aeromonas spp. are yet to be elucidated and understood. Laboratory investigations, for instance, have failed to consistently identify such mechanisms or to lead to an animal model that faithfully reproduces the disease symptoms of infected humans and animals. The aim of this study was to investigate and study the enteropathogenicity of P. shigelloides, A. hydrophila, A. caviae and A. sobria both in monoinfections and in coinfections with the protozoan parasite Cryptosporidium parvum in neonatal BALB/c mice.
2. Materials and methods 2.1. Animals A total of 110 pregnant BALB/c mice for the study were obtained from AnLab (Czech Republic). All were individually kept in plastic cages housed in ¯exible ®lm isolation facilities installed with high eciency particulate ®lters (BEM, Znojmo, Czech Republic). In the course of 2 days, the pregnant female mice delivered, on average, seven neonates. Each litter was kept together with its mother during the entire period of the experiments. The mice were subdivided into groups, one
42
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
group for each experiment. Within each group, the mice were kept in pairs. Cages, food, water and bedding were sterilised before use. 2.2. Bacterial strains and parasite used for inoculation Bacterial strains were obtained from the National Collection of Microorganisms at Prague, Czech Republic. A Plesiomonas shigelloides strain (29480, serotype 027:H8) was isolated from a child with diarrhoea. An Aeromonas hydrophila strain (29529), an A. caviae strain (29188), and an A. sobria strain (29529) were isolated from other cases of human diarrhoea. A bovine strain of the protozoan parasite Cryptosporidium parvum was isolated from a naturally-infected scouring calf at the research facilities of the Faculty of Agriculture (University of South Bohemia, Ceske Budejovice, Czech Republic). 2.3. Detection of potential virulence factors in bacterial strains used for experimental infection The strains of P. shigelloides, A. hydrophila, A. caviae and A. sobria were examined for dierent putative virulence factors, such as adhesive and invasive properties, and for the production of cytotoxins, cytotonic toxins, enterotoxin and hemolysin. These factors and properties were investigated using methods described in earlier studies [11±14]. 2.4. Preparation of bacterial suspensions for mice oral infection Strains were streaked on Blood Agar Base No. 2 (Oxoid) containing 7% de®brinated sheep blood, and then incubated at 378C for 20 h before being inoculated into nutrient broth (Oxoid). The broth was incubated at 378C for 20 without agitation. Cultures were then harvested by centrifugation (16,000 g) at 48C for 30 min. The bacterial pellets were washed twice in sterile Phosphate Buer Saline (PBS, pH 7.2), and then resuspended in sterile PBS to a cell density of 7 109 CFU/ml (P. shigelloides ) and 7 1010 CFU/ml (Aeromonas spp.) before being used to orally infect neonatal mice (0.1 ml per mouse). 2.5. Experimental design and oral infection of mice A total of 700 mice of both sexes, 6±7 days old, were used for the infection experiments in this study. For each experiment with P. shigelloides, two trials were carried out, but for the infection experiments with the three Aeromonas spp., only one trial was carried out. For the P. shigelloides experiments, 140 neonatal mice were divided into six groups corresponding to six dierent treatments (Table 1). Within each group, mice were kept in pairs. An analogous set of experimental groups was implemented for each Aeromonas spp. In the coinfection experiments, an inoculum of 105 oocysts of C. parvum was orally administered to each mouse using methods described earlier [15]. In the
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
43
Table 1 Experimental design and oral infections of neonatal BALB/c mice by P. shigelloides and/or C. parvum. An analogous set of experimental groups was implemented for each Aeromonas spp. Group
Type of infection
PS CP PS/CP
Monoinfection with P. shigelloides Monoinfection with C. parvum Coinfection (infection with P. shigelloides followed 2 days later by administration of C. parvum ) Coinfection (infection with C. parvum was followed 2 days later by administration of P. shigelloides ) Coinfection (the inoculum of P. shigelloides and C. parvum were administered simultaneously) Each mouse in group was given 0.1 ml of sterile PBS
CP/PS CP+PS Control
simultaneous experiments, an inoculum of 105 oocysts of the C. parvum was ®rst administered followed immediately by inoculation of respective bacterial suspensions. 2.6. Autopsy and collection of the samples for bacteriological and histopathological investigations On dierent days of postinoculation (DPI), two neonatal mice, one from an infected group of mice and one from a control group, were randomly selected and euthanised with ether before performing gross pathological investigations. For transmission electron microscopy (TEM), scanning electron microscopy (SEM), and histopathological examination, intestinal segments of ileum (near the ostium ileocecale); the distal part of jejunum, apex caeci, and caecum; and parts of the colon were collected. For bacteriological examination, tissue samples were also taken aseptically from the distal part of the jejunum, colon and spleen. 2.7. Bacteriological examinations Tissue samples from small intestine, colon and spleen were separately homogenised under aseptic conditions and inoculated into nutrient broth (NB, Oxoid) in culture tubes. The NB culture tubes were incubated overnight at 378C before material from the nutrient broth (NB) was streaked onto desoxycholate agar (DC, Oxoid), inositol brilliant green bile salts agar, and blood agar (7% sheep erythrocytes in blood agar No.2, Oxoid). The agar plates were incubated overnight at 378C. For further identi®cation, 3±5 colonies were randomly selected from each agar plate. The bacteria in each of these colonies were identi®ed as Gram negative, oxidase positive rods. For further identi®cation, Plesiomonas and Aeromonas spp. were typed using a conventional biochemical tube test. To identify the bacteria, acid and gas production from glucose; acid production from lactose, sucrose, mannitol, salicin, arabinose and inositol; indole production; and Voges±
44
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
Proskauer and Mollers decarboxylases (lysine, ornithine, arginine) were used. To obtain viable counts for samples from small intestine incubated in NB, serial dilutions in sterile NB were prepared, plated later on blood agar and DC agar, and then incubated overnight at 378C. 2.8. Histological examination Parts of proximal, central and distal regions of jejunum, caecum and colon, along with parts of spleen and liver of the infected mice, were ®xed in 10% buered formaldehyde. These ®xed tissues were embedded in paran wax, sectioned (5 mm thick), mounted on glass slides, and stained using haematoxylin and eosine (HE), or Giemsa, according to standard procedures for histopathological examination. 2.9. Transmission electron microscopy (TEM) examination Tissue samples from ileum and colon of infected mice were ®xed in 2.5% glutaraldehyde in 0.1 M cacodylate buer, post®xed in 1% OsO4, dehydrated in a graded series of ethanol concentrations, transferred to Beem-capsules, and embedded in Durcupan. The ultrathin sections were post stained with uranyl acetate and lead citrate and examined at 80 kV with a Philips EM 420 electron microscope. 2.10. Scanning electron microscopy (SEM) examination Parts of proximal, central and distal regions of jejunum, caecum and colon of the infected mice were ®xed with 2.5% glutaraldehyde in PBS. The specimens were rinsed with PBS, post®xed in 1% osmium tetroxide (OsO4) in distilled water, and rinsed in the same buer. They were then dehydrated in a graded series of ethanol concentrations and dried in a critical-point drying apparatus. The tissue samples were mounted on stubs, coated with gold±palladium, and examined with a JOEL MS 6300 scanning electron microscope. 3. Results 3.1. Putative virulence properties of P. shigelloides and Aeromonas spp. strains An A. hydrophila strain (9072) and an A. sobria strain (29529) were weak cytotoxin producers with titres 1:8, based upon the Vero cell (Green monkey kidney) test, whereas an A. caviae strain (29188) produced cytotoxin with titres 1:128. None of the aeromonads was cytotonic in the Chinese hamster ovary (CHO-K1) tissue culture cell assay. Strains of A. hydrophila and A. caviae were weakly hemolysin positive for bovine erythrocyte (titre 1:1). For all the aeromonads strains, adhesive and invasive properties for human intestinal 407
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
45
tissue culture cells were not observed. Strain 29480 of P. shigelloides expressed none of the virulence factors investigated. 3.2. Clinical observations of infected mice For each experiment, all mice were monitored daily for death or clinical signs of the development of diarrhoea. All clinical observations were made in reference to days of postinoculation (DPI). However, this terminology needs further elaboration because, for coinfection experimental groups e.g., in CP/PS group, there is a 2-day dierence in inoculation times for the protozoan C. parvum (CP) and the bacterial species P. shigelloides (PS). To illustrate in terms of a hypothetical experiment for the CP/PS group, 7 DPI for C. parvum will be 5 DPI for P. shigelloides. To handle this diculty in specifying the DPI for both pathogens on a certain day, we refer to that day of our hypothetical experiment for the CP/PS group as 7/5 DPI. The PS group and control group of mice showed no signs of diarrhoea, but diarrhoea did occur for the CP/PS group during the period from 5/3 to 13/11 DPI. During these 8 days of infection, the growth of the diarrhoeic mice was retarded, and 16 mice died between 9/7 and 11/9 DPI. In contrast, although mild diarrhoea was observed in the PS/CP group from 7/5 to 13/11 DPI, and in the CP+PS group from 5 to 11 DPI, there were no deaths among the infected mice. 3.3. Bacteriological ®ndings 3.3.1. PS groups In the PS group, the plesiomonads were isolated in pure cultures from the spleen, small intestine and colon of infected mice from 1 to 5 DPI. Results of the two trials for the experiment showed that pure cultures of the pathogen could also be isolated from the intestines from 1 to 11 DPI (trial 1) and 1±17 DPI (trial 2). Samples collected after these periods for the two trials, and until the end of the experiments (21 DPI), were overgrown with Proteus spp, which compounded diculties in performing further viable count studies. For the (CP/PS) group of mice, from 3/1 to 5/3 DPI the plesiomonads grew in pure cultures in the spleen and the intestine. The number of bacteria (CFU/ml), escalated on subsequent days of the coinfection, was as follows: 1 104 (7/5 DPI), 3 104 (9/7 DPI), and then 1 107 (11/9 DPI). In the PS/CP group of mice, from 9/7 and 11/9 DPI, pure cultures of P. shigelloides (about 3 105 CFU/ml) were isolated from intestinal samples. For the CP+PS group, pure cultures of P. shigelloides of 3 105 CFU/ml or more were isolated from samples of mice at 11 DPI. At 13 DPI, 1 105 CFU/ml were recorded from samples of this group. No growth of P. shigelloides was found in the CP group and control group. 3.3.2. AH groups For the AH group, A. hydrophila was isolated from the spleen and intestine only at 1 DPI. In all three combinations of coinfection with C. parvum and A.
46
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
hydrophila (i.e., CP/AH, AH/CP and CP+AH), A. hydrophila was isolated from the spleen and intestine in the course of the second week after infection (14±21 DPI). For the CP/AH group, at 9/7 DPI, there were about 2 104 CFU/ml of A. hydrophila, and for the AH/CP group, about 4 102 and 3 105 CFU/ml at 7/5 DPI and 11/9 DPI, respectively. For the AH+CP group, there were about 4 103 CFU/ml of A. hydrophila at 9/7 DPI. 3.3.3. AC and AS groups In the case of group AC, pure cultures of A. caviae could be isolated from the spleen and intestine from 1 to 5 DPI. In the case of group AS, pure cultures of A. sobria were found from 1 to 7 DPI. In the CP+AC and CP+AS groups, pure cultures of A. caviae and A. sobria, respectively, were isolated from 1 to 5 DPI. They were also isolated in pure cultures from samples taken from the AC/CP and AS/CP groups from 3/1 to 5/3 DPI. 3.4. Gross pathological ®ndings Necropsy ®ndings for the CP/PS group revealed that during 9/7 to 11/9 DPI the ileum and colon of the infected mice were dilated, congested and thickened. The spleens of the infected mice were also slightly enlarged. In contrast, neither macroscopic changes were observed in any of the groups of mice monoinfected with Aeromonas spp, nor in any groups coinfected with Aeromonas spp. and C. parvum. 3.5. Histopathological examination The histological results for the PS group showed propagation and activation of goblet cells, vacuolisation of absorptive epithelium in the distal part of jejunum and colon, and activation of lymphoid tissue in the ileum. For the CP group, a moderate infection developed at 2 and 3 DPI in the distal part of jejunum and ileum. Moderate signs of infection were also in the caecum and colon. At 5 DPI, we observed moderate and/or severe infection of the distal part of jejunum, and moderate infection of the caecum and colon, whereas at 7 DPI, only moderate infection of the caecum and colon was recorded, and the same for the distal part of jejunum and ileum. At 9 DPI, infection became sporadic in the colon and in the distal part of jejunum and ileum. For the CP/PS group, severe cryptosporidial infection of the distal part of jejunum and ileum and medium infection of caecum and colon was observed at 5/ 3 DPI. At 9/7 DPI, necrotising in¯ammatory changes of the ileum and colon were noticed, with occasional ®ndings of cryptosporidia. For a period after 11/9 DPI, reparative postin¯ammatory changes were found, along with occasional ®ndings of C. parvum in the ileum, caecum and colon. After this period there were no observable pathohistological changes. For the CP+PS group, similar results to those for the CP/PS group were observed at 11 DPI in the ileum, caecum and colon, but with less extensive
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
47
pathological changes. Again in contrast, no histological changes were observed in mono- and coinfection experiments for aeromonads. 3.6. Transmission electron microscopy (TEM) During 1 and 2 DPI for the PS group, normal structure of the ileum and the colon was found. Tall columnar absorptive cells were characterised by numerous long regular microvilli along their luminal borders (Fig. 1). On the villous absorptive cells of the small intestine, microvilli were more numerous, regular, and elongated than on the surface of the absorptive cells in the colon. Focal degeneration and necrobiosis of the supranuclear parts of the absorptive cells took place from 3 to 7 DPI. Similar observations were noticed for the CP/PS group from 3/1 to 7/5 DPI. However, in addition, there was swelling, vacuolisation, and expansion of intermembranous spaces between the outer and inner membranes of mitochondria. Furthermore, the number and size of microvilli were reduced. In the brush borders, numerous meronts and gamonts of cryptosporidia were present. For the PS group, numerous absorptive cells underwent necrobiosis from 9 to 11 DPI, and from 9/7 to 13/11 DPI for the CP/PS group. At the same time, dilatation and degeneration of mitochondria occurred. The terminal web was widened, and the number of tono®laments was reduced. These extended into the intestinal lumen, which often formed pedestals beneath the attached bacteria (Fig. 2a). The degenerated brush border zone was found to have sparse, short,
Fig. 1. Normal structure of the neonatal mouse colon showing long, regular microvilli along luminal borders of the absorptive cells (TEM 19,500).
48
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
irregular microvilli, and a few trophozoits (Fig. 2b). Formation of cup-like depressions occurred in the zone associated with bacteria, along with eacement of the brush borders (Fig. 2c). No penetration of bacteria into the mucosa was observed.
Fig. 2. Neonatal mouse colon coinfected with C. parvum/P. shigelloides after dierent days of postinoculation (DPI): (a) 11/9 DPI, tono®laments extended into intestinal lumen forming pedestals beneath attached bacteria (TEM 41,500); (b) 7/5 DPI, trophozoit in the degenerated brush border zone of absorptive cells (TEM 9800); (c) 11/9 DPI, crater with bacteria and eacement of the brush border absorptive cells (TEM 49,000).
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
49
Fig. 2 (continued).
Fig. 3. Ileum from neonatal mouse monoinfected with A. hydrophila, 2 DPI, showing disconnection of interdigitating folds and dilated intercellular spaces among the absorptive cells situated above the capillary in the neighbouring propria (TEM 10,700).
50
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
For all Aeromonas spp groups of mice, focal disconnection of interdigitating folds occurred during 1 to 3 DPI in the ileum and colon, accompanied by wide dilated intercellular spaces between enterocytes positioned beneath zonula occludens. These focal disconnections of interdigitating folds were observed speci®cally above the capillaries with oedematous and hypertrophied walls in the propria (Fig. 3). The capillaries in the lamina propria showed hypertrophy. Abundant bacteria in the intestinal lumen were observed that did not adhere to or penetrate the absorptive epithelial cells. 3.7. Scanning electron microscopy (SEM) For the PS group, during the ®rst week (1±7 DPI), oedematous swelling of the distal third of small intestine and colon mucosal surface occurred and numerous goblet cells had distended openings. Samples taken at 9 DPI indicated that the colonic mucosa was covered with necrotic material that predominantly contained fusiform bacteria. For all three forms of coinfection combinations of P. shigelloides and C. parvum, and particularly for intestinal samples from the CP/PS group collected at 4/2 DPI, massive cryptosporidial infection and villous atrophy was evident. Samples collected later revealed that the mucosal surface of ileum, caecum and colon were covered with necrotic material containing indigenous micro¯ora. Intestinal samples collected between 1 and 3 DPI from the three groups of mice that were monoinfected with Aeromonas spp. revealed the presence of a group of absorptive cells in the ileum and colon that had no distinct intercellular borders, covered with focal ®ne membranes containing short rods (Fig. 4a). For all Aeromonas spp. groups of coinfected mice, at 5/3 DPI there was local fusion between adjacent absorptive cells as well as sparse vesicular or polypoid protrusions above the inner mucosal surface (Fig. 4b). Moreover, development of focal shortening of microvilli, the presence of abundant indigenous micro¯ora, and the existence of groups of short rods dispersed on the brush borders of the absorptive epithelium were also observed. 4. Discussion Plesiomonas shigelloides and some members of the genus of Aeromonas have been involved in gastrointestinal and extra-intestinal infections, both in humans and animals [1,4,9,11,16]. However, the identi®cation and understanding of their true role in enteric disease has been hampered by the lack of an animal model incorporating reproducible ful®lment of Koch's postulates. In addition, human volunteer studies and primate challenge studies with Aeromonas spp. have also been inadequate [17,18] in this matter. On the other hand, the presence of these pathogens in stools of patients with diarrhoea in the absence of other enteric pathogens indicates their possible role in the gastrointestinal diseases [6]. The protozoan parasite C. parvum is an important and widely distributed
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
51
enteric pathogen in humans and young livestock, as indicated by increasing documentation of disease outbreaks which are associated with presence of cryptosporidia in public water supplies [19,20]. Since both animals and humans can be coinfected with dierent pathogenic
Fig. 4. SEM micrographs: (a) ileum from neonatal mouse infected with A. hydrophila (2 DPI) showing absorptive cells on the villous tip covered with focal ®ne membranes (scale bar=10 mm); (b) mouse colon coinfected with A. hydrophila and C. parvum (5/3) DPI) showing vesicular protrusion above the inner mucosal surface (scale bar=1 mm).
52
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
microorganisms, and since various relationships can exist between the immunoresponse generated by the dierent pathogens, the pathogenicity of the infection of one pathogen can be in¯uenced by the presence of another. Coinfections with other enteropathogens and a potential predisposing factor in P. shigelloides infections have been described [9]. In addition, spontaneous coinfection of P. shigelloides with C. parvum and other coccidia has also been reported in animals [21]. The basic aim of this study was to investigate and studied the enteropathogenicity of P. shigelloides, A. hydrophila, A. caviae and A. sobria both in monoinfections and in coinfections with the protozoan parasite C. parvum in neonatal BALB/c mice. Implicit in our experiments and results is a model that includes long-term colonisation of P. shigelloides in the small and large intestine, which was found to be frequently accompanied with bacteremia, since P. shigelloides was isolated from the spleen of infected mice from the 1st to 5th day of postinoculation (1±5 DPI). Histological ®ndings revealed pathological lesions, which varied in severity from atrophy to necrosis of the mucosal inner surface of the ileum and colon, with predilection to the colon and brush border of colonic enterocytes. Ultrastructural TEM observations of samples from infected intestinal mucosa revealed shrinkage in the size of enterocytes and reduction in the number and size of microvilli during the initial stages of the pathological process. Even mitochondria and micro®laments in the widened terminal web were reduced in number. These initial ultrastructural pathological alterations of the intestine were simultaneously accompanied by necrobiosis, predominately observed in the brush borders and supranuclear parts of the enterocytes. Interestingly, the eects of coinfection with C. parvum were characterised by more intense and heavier colonisation of the intestine by P. shigelloides. In addition, extensive necrotising in¯ammatory changes in the ileum and colon, accompanied by bacteremia, diarrhoea and deaths of the mice were also recorded. Our results agree with those of Vitovec et al. [22], who observed similar interactions between Campylobacter jejuni and C. parvum in experimentally infected BALB/c neonatal mice. These investigators found that the synergistic interaction of C. jejuni and C. parvum resulted in heavier colonisation of the inner surface of the small and large intestine by C. jejuni. With the experimental designs described in this article as a backdrop, our results along with those of previous coinfection studies on C. jejuni and C. parvum, suggest that the colon and the distal third of small intestine are the target sites for P. shigelloides and C. jejuni infections. These observations seem to resemble rather well the pathological observations often encountered in cases of human colitis for which spontaneous plesiomonads infections are involved [23,24]. The virulence in Aeromonas spp. may well be multifactoral and polygenic in nature, involving a complex interaction and bacterial and host factors [1,11]. For instance, protein de®ciency in experimental mice led to bowel colonisation by aeromonads [25], and streptomycin treatment predisposed mice to A. sobria and A. hydrophila colonisation, whereas control mice that were not treated with
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
53
streptomycin were in turn not colonised [26]. It is also known that coinfection by other enteropathogens frequently occurs in Aeromonas-associated diarrhoea in infants [27]. Moreover, coinfection with cryptosporidia might indeed have contributed to the severe diarrhoea encountered in 9% of foals infected by A. hydrophila [28]. These observations and others suggest that the neonatal BALB/c mouse model that is implicit in our study may serve well as a tool for studying the initial steps of gastrointestinal colonisation of Aeromonas spp. The results from the Aeromonas spp. monoinfections showed only a short term colonisation of mouse intestine by aeromonads, whereas mice coinfected with A. hydrophila and C. parvum were found to have a prolonged growth period for these bacteria in the spleen and intestine. However, coinfection with C. parvum did not substantially further in¯uence the intestinal colonisation. Within the experimental groups of enterocytes in our study, the TEM observations of samples infected by aeromonads revealed a disconnection of interdigitating folds, and they also revealed, beneath the zonula occludens, the formation of widely dilated intercellular spaces between enterocytes. These were usually found above the capillaries, and speci®cally with oedematous and hypertrophied walls in the propria. In addition, neither clinical signs of diarrhoea nor substantial pathomorphological changes on the inner mucosal surface, nor in¯uence on colonisation and patent period of C. parvum, were observed in the mice infected by the Aeromonas spp. strains that were used. In conclusion, the results from the present investigation indicate that experimental infection of neonatal BALB/c mice with P. shigelloides, Aeromonas spp. and C. parvum, including coinfections of the bacteria with the protozoan, may serve well as an initial experimental animal model for studying the initial steps of gastrointestinal colonisation and the diarrhoeal disease syndrome caused by these bacterial and protozoan pathogens. Such a model can serve as a step leading toward improved models for understanding gastrointestinal colonisation and the diarrhoeal disease syndrome.
Acknowledgements This study was supported by a grant from the Royal Swedish Academy of Sciences and Czech Academy of Sciences (KK and JV). The authors gratefully acknowledge Dr. B. Koudela for assistance in preparing C. parvum and Dr. A. Faris for his critical comments and suggestions during the preparation of the manuscript and Dr. Valerie Amarger for the French translation of the resumeÂ.
References [1] Janda JM. Recent advances in the study of the taxonomy, pathogenicity, and infectious syndromes associated with the genus Aeromonas. Clin Microbiol Rev 1991;4:397±410.
54
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
[2] Abbott SL, Kokka RP, Janda JM. Laboratory investigations on the low pathogenic potential of Plesiomonas shigelloides. J Clin Microbiol 1991;29:148±53. [3] Brenden RA, Miller MA, Janda JM. Clinical disease spectrum and pathogenic factors associated with Plesiomonas shigelloides infections in humans. Rev Infect Dis 1988;10:303±16. [4] Herrington DA, Tzipori S, Robins-Browne R, Tall BD, Levine MM. In vitro and in vivo pathogenicity of Plesiomonas shigelloides. Infect Immun 1987;55:979±85. [5] Holt JG, Krieg NR, Sneath PHA, Staley JT, Williams ST, editors. Bergey's manual of determinative bacteriology. 9th ed. Baltimore, USA: Williams & Wilkins, 1994. [6] Khardori N, Fainstein V. Aeromonas and Plesiomonas as etiological agents. Ann Rev Microbiol 1986;42:395±419. [7] KuÈhn I, Allestam G, Huys P, Jannssen K, Kersters K, Krovacek K, StenstroÈm TA. Diversity, stability and virulence properties of Aeromonas strains isolated from drinking water distribution systems in Sweden. Appl Environ Microbiol 1997;63:2708±15. [8] Janda JM, Duey PS. Mesophilic aeromonads in human disease: current taxonomy, laboratory identi®cation and infectious disease spectrum. Rev Infect Dis 1988;10:980±97. [9] Clark RB, Westby GR, Spector H, Soricelli RR, Young CHL. Fatal Plesiomonas shigelloides septicaemia in a splenectomised patient. J Infect 1991;23:89±92. [10] Ljungh A, WadstroÈm T. Aeromonas toxins. In: Dorner F, Drews J, editors. Pharmacology of bacterial toxins, vol. 119. Oxford: Pergamon Press, 1986. p. 289±301. [11] Krovacek K, Conte M, Galderisi P, Morelli G, Postiglione A, Dumontet S. Fatal septicaemia caused by Aeromonas hydrophila in a patient with cirrhosis. Comp Immun Microbiol Infect Dis 1993;16:267±72. [12] Krovacek K, Pasquale V, Baloda SB, Soprano V, Conte M, Dumontet S. Comparison of putative virulence factors in Aeromonas hydrophila strains isolated from the marine environment and human diarrheal cases in southern Italy. Appl Environ Microbiol 1994;60:1379±82. [13] Krovacek K, Faris A, Mansson I. In vitro invasion of Aeromonas spp. to HEp-2 tissue culture cells. Acta Vet Scand 1991;32:39±43. [14] Krovacek K, Dumontet S, Eriksson E, Baloda SB. Isolation, and virulence pro®les of Aeromonas hydrophila implicated in an outbreak of food poisoning in Sweden. Microbiol Immunol 1995;39:655±61. [15] Vitovec J, Koudela B. Location and pathogenicity of Cryptosporidium parvum in experimentally infected mice. J Vet Med A 1988;35:515±24. [16] Hathcock TL, Schumacher J, Wright JC, Stringfellow J. The prevalence of Aeromonas species in faeces of horses with diarrhoea. Vet Intern Med 1999;13:357±60. [17] Morgan DR, Johnson SR, Chattopadhyyay B. Lack of correlation between known virulence properties of Aeromonas hydrophila and enteropathogenicity for humans. Infect Immun 1985;50:62±5. [18] Pitarangsi C, Echeverria P, Whitmire R, Tirapat C, Formal S, Dammin GJ, Tingtalapong M. Enteropathogenicity of Aeromonas hydrophila and Plesiomonas shigelloides: prevalence among individuals with and without diarrhea in Thailand. Infect Immun 1982;35:666±73. [19] Dubey JP, Speer CA, Fayer R. Cryptosporidiosis of man and animals. Boston: CRC Press, 1990. [20] Fayer R. Cryptosporidium and cryptosporidiosis. Boca Raton, FL: CRC Press, 1997. [21] GluÈnder G, Zum von Vorkommen. Plesiomonas shigelloides bei Wild-und Zoovogeln. Berl MuÈnch TieraÈrztl Washr 1988;101:334±7. [22] Vitovec J, Koudela B, Vladik P, Hausner O. Interaction of Cryptosporidium parvum and Campylobacter jejuni in experimentally infected neonatal mice. Zbl Bakt 1991;274:548±59. [23] Loon FP, Rahim Z, Chowdhuri KA, Kay BA, Rahman SA. Case report of Plesiomonas shigelloides-associated persistent dysentery and pseudomembranous colitis. J Clin Microbiol 1989;27:1913±5. [24] Mandal BK, Whale K, Morson BC. Acute colitis due to Plesiomonas shigelloides. Brit Med J 1982;285:1539±40. [25] Pazzaglia G, Winoto I, Jennings G. Oral challenge with Aeromonas in protein-malnourished mice. J Diarrhoeal Dis Res 1994;12:108±12.
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
55
[26] Sanderson K, Ghazali FM, Kirov SM. Colonization of streptomycin-treated mice by Aeromonas species. J Diarrhoeal Dis Res 1996;14:27±32. [27] Pazzaglia G, Sack RB, Salazar E, Yi A, Chea E, Leon-Barua R, Guerrero CE, Palomino J. High frequency of coinfecting enteropathogens in Aeromonas-associated diarrhea of hospitalized peruvian infants. J Clin Microbiol 1991;29:1151±6. [28] Browning GF, Chalmers RM, Snodgrass DR, Batt RM, Hart CA, Ormarod SE, Leadon D, Stoneham SJ, Rossdale PD. The prevalence of enteric pathogens in diarrhoeic thoroughbred foals in Britain and Ireland. Equine Vet J 1991;23:405±9.
Enteropathogenicity of Plesiomonas shigelloides and Aeromonas spp. in experimental monoand coinfection with Cryptosporidium parvum in the intestine of neonatal BALB/c mice Jiri Vitovec a, Eva Aldova b, Petr Vladik c, Karel Krovacek d,* a
University of South Bohemia, Faculty of Agriculture, Ceske Budejovice, Czech Republic b National Institute of Public Health, Praha, Czech Republic c State Veterinary Institute, Ceske Budejovice, Czech Republic d Department of Veterinary Microbiology, Box 583, Biomedical Centre, 751 23 Uppsala, Sweden
Abstract Enteropathogenicity of Plesiomonas shigelloides, Aeromonas hydrophila, A. caviae and A. sobria was studied both in monoinfections and in coinfections with coccidium Cryptosporidium parvum in neonatal BALB/c mice. In monoinfection experiments, neonatal BALB/c mice were orally infected with 7 107 or 7 108 CFU, respectively, of a strain of P. shigelloides or a strain of an Aeromonas spp. In coinfection experiments, the neonatal mice were, in addition to being orally infected with one of the four bacterial species, orally infected with an inoculum containing 105 oocysts of C. parvum. Results from monoinfections with P. shigelloides revealed long-term colonisation of the neonatal mouse intestine by this pathogen, along with associated pathological lesions. The lesions varied in severity from atrophy to necrosis of the mucosal inner surface of the ileum and colon, with predilection to the colon and brush border of colonic enterocytes. The eects of coinfection of P. shigelloides with C. parvum were characterised by bacteremia and heavy colonisation of the intestine by P. shigelloides. In addition, extensive necrotising in¯ammatory changes in the ileum and colon were accompanied by diarrhoea and deaths of coinfected mice. In contrast, the results from monoinfections of neonatal mice with Aeromonas spp. showed only a short-term colonisation of the intestine by the pathogen. However, when mice were coinfected with A. hydrophila and C. parvum, then the growth of the bacterial species was
* Corresponding author Tel: +46-18-4714592; fax: +46-18-504461. E-mail address: [email protected] (K. Krovacek). 0147-9571/01/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 7 - 9 5 7 1 ( 0 0 ) 0 0 0 1 2 - 6
40
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
prolonged, and occurred in both the spleen and intestine. However, no substantial clinical or histopathological changes were observed in mice, whether monoinfected with Aeromonas spp. or coinfected with C. parvum. Our study suggests that experimental monoinfections of neonatal BALB/c mice with P. shigelloides, Aeromonas spp. and C. parvum, together with coinfections (each bacterial species with the protozoan C. parvum ), may serve as a useful model to study the initial steps of gastrointestinal colonisation and diarrhoeal disease syndromes caused by enteropathogenic bacteria and protozoa, individually and in combination. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Enteropathogenicity; Plesiomonas shigelloides; Aeromonas hydrophila; A. caviae; A. sobria; Cryptosporidium parvum; Experimental infection; Coinfection; Neonatal BALB/c mice
ReÂsume L'eet enteÂropathogeÁne de Pleisomonas shigelloides, Aeromonas hydrophila, A. caviae et A. sobria a eÂte eÂtudie en situation de mono ou co-infection avec le coccidium Crystosporidium parvum dans des souris BALB/c au stade neÂonatal. Dans les cas de monoinfection, les animaux ont eÂte infecteÂs par voie orale avec 7 107 et 7 108 CFU de P. shigelloides ou Aeromonas respectivement. Dans les cas de co-infection, conjointement aÁ l'infection par l'une des quatre espeÁces bacteÂriennes, les animaux ont recu par voie orale eÂgalement un inoculum contenant 105 oocystes d'un protozoaire, le coccidium C. parvum. Les reÂsultats obtenus aÁ partir des mono-infections par P. shigelloides montrent une colonisation aÁ long terme de l'intestin par le pathogeÁne, ainsi que des leÂsions pathologiques associeÂes. Les leÂsions sont de seÂveÂrite variable, allant de l'atrophie aÁ la neÂcrose de la surface interne de la muqueuse de l'ileum et du colon, avec une preÂdilection pour le colon et la bordure apicale des enteÂrocytes du colon. La co-infection avec C. parvum se caracteÂrise par une septiceÂmie et une colonisation majeure de l'intestin par P. shigelloides. D'importantes in¯ammations et neÂcroses de l'ileum et du colon, accompagneÂes de diarrheÂe et conduisant aÁ la mort des souris co-infecteÂes ont eÂgalement eÂte observeÂes. Au contraire, les reÂsultats obtenus aÁ partir des mono-infections avec les dieÂrentes espeÁces d'Aeromonas montrent seulement une colonisation aÁ court terme de l'intestin par ces bacteÂries alors que les souris co-infecteÂes par A. hydrophila et C. parvum preÂsentent une peÂriode de croissance prolongeÂe des bacteÂries dans la rate et l'intestin. Cependant, aucun changement notable sur le plan clinique et histopathologique n'a eÂte observe aussi bien chez les souris mono-infecteÂes par Aeromonas spp. que celles co-infecteÂes avec C. parvum. Notre eÂtude montre donc qu'une infection de souris BALB/c au stade neÂonatal par inoculation orale de P. shigelloides, Aeromonas spp. et C. parvum, ainsi que des co-infections bacteÂries/protozoaires, constitue un treÁs bon modeÁle animal pour l'eÂtude des eÂtapes initiales de colonisation gastrointestinale et des syndromes de diarrheÂe causeÂs par ces pathogeÁnes bacteÂriens et protozoaires. 7 2000 Elsevier Science Ltd. All rights reserved. Mots-cleÂf: EnteÂropathogeÂniciteÂ; Plesiomonas shigelloides; Aeromonas hydrophila; A. caviae; A. sobria; Cryptosporidium parvum; Infection expeÂrimentale; Co-infection; Souris BALB/c neÂonatales
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
41
1. Introduction Plesiomonas shigelloides and Aeromonas species belong to bacterial groups of aquatic pathogens recently recognised as potential human enteropathogens [1±4]. They are Gram negative, motile, non-spore forming bacilli, which are facultatively anaerobic and oxidase positive. According to the Ninth Edition of Bergey's Manual of Determinative Bacteriology, the genus Plesiomonas and Aeromonas spp. are still included in the family of Vibrionaceae [5]. Aeromonas spp. and P. shigelloides are commonly found in a wide range of aquatic environments and foods, and have been isolated from lakes, rivers, coastal waters and drinking water [6±8]. They have also been isolated from humans and domestic animals, such as dogs, cats, goats, sheep and cows [1±4,6±8]. P. shigelloides and the motile Aeromonas spp. have been implicated as agents of human gastroenteritis. They can cause three major types of gastroenteritis: (i) a secretory, watery form, (ii) an invasive, dysentery-like form, and (iii) a subacute or chronic form lasting between 2 weeks and 3 months [3,7,9]. Although extraintestinal infections such as bacteremia, cellulitis, polyarthritis, and meningitis caused by P. shigelloides and Aeromonas spp. are rarely reported, they have been associated with infections in immunocompromised patients and in patients with an underlying disease [1,3,9± 11]. Both P. shigelloides and the motile Aeromonas spp. possess several seemingly pathogenic properties, such as the production of heat-stabile and heat-labile exotoxins, cytotoxin, haemolysin, hemaglutinin and other potentially virulence factors [7,11±15]. However, the exact mechanisms of enteropathogenicity for P. shigelloides and Aeromonas spp. are yet to be elucidated and understood. Laboratory investigations, for instance, have failed to consistently identify such mechanisms or to lead to an animal model that faithfully reproduces the disease symptoms of infected humans and animals. The aim of this study was to investigate and study the enteropathogenicity of P. shigelloides, A. hydrophila, A. caviae and A. sobria both in monoinfections and in coinfections with the protozoan parasite Cryptosporidium parvum in neonatal BALB/c mice.
2. Materials and methods 2.1. Animals A total of 110 pregnant BALB/c mice for the study were obtained from AnLab (Czech Republic). All were individually kept in plastic cages housed in ¯exible ®lm isolation facilities installed with high eciency particulate ®lters (BEM, Znojmo, Czech Republic). In the course of 2 days, the pregnant female mice delivered, on average, seven neonates. Each litter was kept together with its mother during the entire period of the experiments. The mice were subdivided into groups, one
42
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
group for each experiment. Within each group, the mice were kept in pairs. Cages, food, water and bedding were sterilised before use. 2.2. Bacterial strains and parasite used for inoculation Bacterial strains were obtained from the National Collection of Microorganisms at Prague, Czech Republic. A Plesiomonas shigelloides strain (29480, serotype 027:H8) was isolated from a child with diarrhoea. An Aeromonas hydrophila strain (29529), an A. caviae strain (29188), and an A. sobria strain (29529) were isolated from other cases of human diarrhoea. A bovine strain of the protozoan parasite Cryptosporidium parvum was isolated from a naturally-infected scouring calf at the research facilities of the Faculty of Agriculture (University of South Bohemia, Ceske Budejovice, Czech Republic). 2.3. Detection of potential virulence factors in bacterial strains used for experimental infection The strains of P. shigelloides, A. hydrophila, A. caviae and A. sobria were examined for dierent putative virulence factors, such as adhesive and invasive properties, and for the production of cytotoxins, cytotonic toxins, enterotoxin and hemolysin. These factors and properties were investigated using methods described in earlier studies [11±14]. 2.4. Preparation of bacterial suspensions for mice oral infection Strains were streaked on Blood Agar Base No. 2 (Oxoid) containing 7% de®brinated sheep blood, and then incubated at 378C for 20 h before being inoculated into nutrient broth (Oxoid). The broth was incubated at 378C for 20 without agitation. Cultures were then harvested by centrifugation (16,000 g) at 48C for 30 min. The bacterial pellets were washed twice in sterile Phosphate Buer Saline (PBS, pH 7.2), and then resuspended in sterile PBS to a cell density of 7 109 CFU/ml (P. shigelloides ) and 7 1010 CFU/ml (Aeromonas spp.) before being used to orally infect neonatal mice (0.1 ml per mouse). 2.5. Experimental design and oral infection of mice A total of 700 mice of both sexes, 6±7 days old, were used for the infection experiments in this study. For each experiment with P. shigelloides, two trials were carried out, but for the infection experiments with the three Aeromonas spp., only one trial was carried out. For the P. shigelloides experiments, 140 neonatal mice were divided into six groups corresponding to six dierent treatments (Table 1). Within each group, mice were kept in pairs. An analogous set of experimental groups was implemented for each Aeromonas spp. In the coinfection experiments, an inoculum of 105 oocysts of C. parvum was orally administered to each mouse using methods described earlier [15]. In the
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
43
Table 1 Experimental design and oral infections of neonatal BALB/c mice by P. shigelloides and/or C. parvum. An analogous set of experimental groups was implemented for each Aeromonas spp. Group
Type of infection
PS CP PS/CP
Monoinfection with P. shigelloides Monoinfection with C. parvum Coinfection (infection with P. shigelloides followed 2 days later by administration of C. parvum ) Coinfection (infection with C. parvum was followed 2 days later by administration of P. shigelloides ) Coinfection (the inoculum of P. shigelloides and C. parvum were administered simultaneously) Each mouse in group was given 0.1 ml of sterile PBS
CP/PS CP+PS Control
simultaneous experiments, an inoculum of 105 oocysts of the C. parvum was ®rst administered followed immediately by inoculation of respective bacterial suspensions. 2.6. Autopsy and collection of the samples for bacteriological and histopathological investigations On dierent days of postinoculation (DPI), two neonatal mice, one from an infected group of mice and one from a control group, were randomly selected and euthanised with ether before performing gross pathological investigations. For transmission electron microscopy (TEM), scanning electron microscopy (SEM), and histopathological examination, intestinal segments of ileum (near the ostium ileocecale); the distal part of jejunum, apex caeci, and caecum; and parts of the colon were collected. For bacteriological examination, tissue samples were also taken aseptically from the distal part of the jejunum, colon and spleen. 2.7. Bacteriological examinations Tissue samples from small intestine, colon and spleen were separately homogenised under aseptic conditions and inoculated into nutrient broth (NB, Oxoid) in culture tubes. The NB culture tubes were incubated overnight at 378C before material from the nutrient broth (NB) was streaked onto desoxycholate agar (DC, Oxoid), inositol brilliant green bile salts agar, and blood agar (7% sheep erythrocytes in blood agar No.2, Oxoid). The agar plates were incubated overnight at 378C. For further identi®cation, 3±5 colonies were randomly selected from each agar plate. The bacteria in each of these colonies were identi®ed as Gram negative, oxidase positive rods. For further identi®cation, Plesiomonas and Aeromonas spp. were typed using a conventional biochemical tube test. To identify the bacteria, acid and gas production from glucose; acid production from lactose, sucrose, mannitol, salicin, arabinose and inositol; indole production; and Voges±
44
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
Proskauer and Mollers decarboxylases (lysine, ornithine, arginine) were used. To obtain viable counts for samples from small intestine incubated in NB, serial dilutions in sterile NB were prepared, plated later on blood agar and DC agar, and then incubated overnight at 378C. 2.8. Histological examination Parts of proximal, central and distal regions of jejunum, caecum and colon, along with parts of spleen and liver of the infected mice, were ®xed in 10% buered formaldehyde. These ®xed tissues were embedded in paran wax, sectioned (5 mm thick), mounted on glass slides, and stained using haematoxylin and eosine (HE), or Giemsa, according to standard procedures for histopathological examination. 2.9. Transmission electron microscopy (TEM) examination Tissue samples from ileum and colon of infected mice were ®xed in 2.5% glutaraldehyde in 0.1 M cacodylate buer, post®xed in 1% OsO4, dehydrated in a graded series of ethanol concentrations, transferred to Beem-capsules, and embedded in Durcupan. The ultrathin sections were post stained with uranyl acetate and lead citrate and examined at 80 kV with a Philips EM 420 electron microscope. 2.10. Scanning electron microscopy (SEM) examination Parts of proximal, central and distal regions of jejunum, caecum and colon of the infected mice were ®xed with 2.5% glutaraldehyde in PBS. The specimens were rinsed with PBS, post®xed in 1% osmium tetroxide (OsO4) in distilled water, and rinsed in the same buer. They were then dehydrated in a graded series of ethanol concentrations and dried in a critical-point drying apparatus. The tissue samples were mounted on stubs, coated with gold±palladium, and examined with a JOEL MS 6300 scanning electron microscope. 3. Results 3.1. Putative virulence properties of P. shigelloides and Aeromonas spp. strains An A. hydrophila strain (9072) and an A. sobria strain (29529) were weak cytotoxin producers with titres 1:8, based upon the Vero cell (Green monkey kidney) test, whereas an A. caviae strain (29188) produced cytotoxin with titres 1:128. None of the aeromonads was cytotonic in the Chinese hamster ovary (CHO-K1) tissue culture cell assay. Strains of A. hydrophila and A. caviae were weakly hemolysin positive for bovine erythrocyte (titre 1:1). For all the aeromonads strains, adhesive and invasive properties for human intestinal 407
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
45
tissue culture cells were not observed. Strain 29480 of P. shigelloides expressed none of the virulence factors investigated. 3.2. Clinical observations of infected mice For each experiment, all mice were monitored daily for death or clinical signs of the development of diarrhoea. All clinical observations were made in reference to days of postinoculation (DPI). However, this terminology needs further elaboration because, for coinfection experimental groups e.g., in CP/PS group, there is a 2-day dierence in inoculation times for the protozoan C. parvum (CP) and the bacterial species P. shigelloides (PS). To illustrate in terms of a hypothetical experiment for the CP/PS group, 7 DPI for C. parvum will be 5 DPI for P. shigelloides. To handle this diculty in specifying the DPI for both pathogens on a certain day, we refer to that day of our hypothetical experiment for the CP/PS group as 7/5 DPI. The PS group and control group of mice showed no signs of diarrhoea, but diarrhoea did occur for the CP/PS group during the period from 5/3 to 13/11 DPI. During these 8 days of infection, the growth of the diarrhoeic mice was retarded, and 16 mice died between 9/7 and 11/9 DPI. In contrast, although mild diarrhoea was observed in the PS/CP group from 7/5 to 13/11 DPI, and in the CP+PS group from 5 to 11 DPI, there were no deaths among the infected mice. 3.3. Bacteriological ®ndings 3.3.1. PS groups In the PS group, the plesiomonads were isolated in pure cultures from the spleen, small intestine and colon of infected mice from 1 to 5 DPI. Results of the two trials for the experiment showed that pure cultures of the pathogen could also be isolated from the intestines from 1 to 11 DPI (trial 1) and 1±17 DPI (trial 2). Samples collected after these periods for the two trials, and until the end of the experiments (21 DPI), were overgrown with Proteus spp, which compounded diculties in performing further viable count studies. For the (CP/PS) group of mice, from 3/1 to 5/3 DPI the plesiomonads grew in pure cultures in the spleen and the intestine. The number of bacteria (CFU/ml), escalated on subsequent days of the coinfection, was as follows: 1 104 (7/5 DPI), 3 104 (9/7 DPI), and then 1 107 (11/9 DPI). In the PS/CP group of mice, from 9/7 and 11/9 DPI, pure cultures of P. shigelloides (about 3 105 CFU/ml) were isolated from intestinal samples. For the CP+PS group, pure cultures of P. shigelloides of 3 105 CFU/ml or more were isolated from samples of mice at 11 DPI. At 13 DPI, 1 105 CFU/ml were recorded from samples of this group. No growth of P. shigelloides was found in the CP group and control group. 3.3.2. AH groups For the AH group, A. hydrophila was isolated from the spleen and intestine only at 1 DPI. In all three combinations of coinfection with C. parvum and A.
46
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
hydrophila (i.e., CP/AH, AH/CP and CP+AH), A. hydrophila was isolated from the spleen and intestine in the course of the second week after infection (14±21 DPI). For the CP/AH group, at 9/7 DPI, there were about 2 104 CFU/ml of A. hydrophila, and for the AH/CP group, about 4 102 and 3 105 CFU/ml at 7/5 DPI and 11/9 DPI, respectively. For the AH+CP group, there were about 4 103 CFU/ml of A. hydrophila at 9/7 DPI. 3.3.3. AC and AS groups In the case of group AC, pure cultures of A. caviae could be isolated from the spleen and intestine from 1 to 5 DPI. In the case of group AS, pure cultures of A. sobria were found from 1 to 7 DPI. In the CP+AC and CP+AS groups, pure cultures of A. caviae and A. sobria, respectively, were isolated from 1 to 5 DPI. They were also isolated in pure cultures from samples taken from the AC/CP and AS/CP groups from 3/1 to 5/3 DPI. 3.4. Gross pathological ®ndings Necropsy ®ndings for the CP/PS group revealed that during 9/7 to 11/9 DPI the ileum and colon of the infected mice were dilated, congested and thickened. The spleens of the infected mice were also slightly enlarged. In contrast, neither macroscopic changes were observed in any of the groups of mice monoinfected with Aeromonas spp, nor in any groups coinfected with Aeromonas spp. and C. parvum. 3.5. Histopathological examination The histological results for the PS group showed propagation and activation of goblet cells, vacuolisation of absorptive epithelium in the distal part of jejunum and colon, and activation of lymphoid tissue in the ileum. For the CP group, a moderate infection developed at 2 and 3 DPI in the distal part of jejunum and ileum. Moderate signs of infection were also in the caecum and colon. At 5 DPI, we observed moderate and/or severe infection of the distal part of jejunum, and moderate infection of the caecum and colon, whereas at 7 DPI, only moderate infection of the caecum and colon was recorded, and the same for the distal part of jejunum and ileum. At 9 DPI, infection became sporadic in the colon and in the distal part of jejunum and ileum. For the CP/PS group, severe cryptosporidial infection of the distal part of jejunum and ileum and medium infection of caecum and colon was observed at 5/ 3 DPI. At 9/7 DPI, necrotising in¯ammatory changes of the ileum and colon were noticed, with occasional ®ndings of cryptosporidia. For a period after 11/9 DPI, reparative postin¯ammatory changes were found, along with occasional ®ndings of C. parvum in the ileum, caecum and colon. After this period there were no observable pathohistological changes. For the CP+PS group, similar results to those for the CP/PS group were observed at 11 DPI in the ileum, caecum and colon, but with less extensive
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
47
pathological changes. Again in contrast, no histological changes were observed in mono- and coinfection experiments for aeromonads. 3.6. Transmission electron microscopy (TEM) During 1 and 2 DPI for the PS group, normal structure of the ileum and the colon was found. Tall columnar absorptive cells were characterised by numerous long regular microvilli along their luminal borders (Fig. 1). On the villous absorptive cells of the small intestine, microvilli were more numerous, regular, and elongated than on the surface of the absorptive cells in the colon. Focal degeneration and necrobiosis of the supranuclear parts of the absorptive cells took place from 3 to 7 DPI. Similar observations were noticed for the CP/PS group from 3/1 to 7/5 DPI. However, in addition, there was swelling, vacuolisation, and expansion of intermembranous spaces between the outer and inner membranes of mitochondria. Furthermore, the number and size of microvilli were reduced. In the brush borders, numerous meronts and gamonts of cryptosporidia were present. For the PS group, numerous absorptive cells underwent necrobiosis from 9 to 11 DPI, and from 9/7 to 13/11 DPI for the CP/PS group. At the same time, dilatation and degeneration of mitochondria occurred. The terminal web was widened, and the number of tono®laments was reduced. These extended into the intestinal lumen, which often formed pedestals beneath the attached bacteria (Fig. 2a). The degenerated brush border zone was found to have sparse, short,
Fig. 1. Normal structure of the neonatal mouse colon showing long, regular microvilli along luminal borders of the absorptive cells (TEM 19,500).
48
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
irregular microvilli, and a few trophozoits (Fig. 2b). Formation of cup-like depressions occurred in the zone associated with bacteria, along with eacement of the brush borders (Fig. 2c). No penetration of bacteria into the mucosa was observed.
Fig. 2. Neonatal mouse colon coinfected with C. parvum/P. shigelloides after dierent days of postinoculation (DPI): (a) 11/9 DPI, tono®laments extended into intestinal lumen forming pedestals beneath attached bacteria (TEM 41,500); (b) 7/5 DPI, trophozoit in the degenerated brush border zone of absorptive cells (TEM 9800); (c) 11/9 DPI, crater with bacteria and eacement of the brush border absorptive cells (TEM 49,000).
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
49
Fig. 2 (continued).
Fig. 3. Ileum from neonatal mouse monoinfected with A. hydrophila, 2 DPI, showing disconnection of interdigitating folds and dilated intercellular spaces among the absorptive cells situated above the capillary in the neighbouring propria (TEM 10,700).
50
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
For all Aeromonas spp groups of mice, focal disconnection of interdigitating folds occurred during 1 to 3 DPI in the ileum and colon, accompanied by wide dilated intercellular spaces between enterocytes positioned beneath zonula occludens. These focal disconnections of interdigitating folds were observed speci®cally above the capillaries with oedematous and hypertrophied walls in the propria (Fig. 3). The capillaries in the lamina propria showed hypertrophy. Abundant bacteria in the intestinal lumen were observed that did not adhere to or penetrate the absorptive epithelial cells. 3.7. Scanning electron microscopy (SEM) For the PS group, during the ®rst week (1±7 DPI), oedematous swelling of the distal third of small intestine and colon mucosal surface occurred and numerous goblet cells had distended openings. Samples taken at 9 DPI indicated that the colonic mucosa was covered with necrotic material that predominantly contained fusiform bacteria. For all three forms of coinfection combinations of P. shigelloides and C. parvum, and particularly for intestinal samples from the CP/PS group collected at 4/2 DPI, massive cryptosporidial infection and villous atrophy was evident. Samples collected later revealed that the mucosal surface of ileum, caecum and colon were covered with necrotic material containing indigenous micro¯ora. Intestinal samples collected between 1 and 3 DPI from the three groups of mice that were monoinfected with Aeromonas spp. revealed the presence of a group of absorptive cells in the ileum and colon that had no distinct intercellular borders, covered with focal ®ne membranes containing short rods (Fig. 4a). For all Aeromonas spp. groups of coinfected mice, at 5/3 DPI there was local fusion between adjacent absorptive cells as well as sparse vesicular or polypoid protrusions above the inner mucosal surface (Fig. 4b). Moreover, development of focal shortening of microvilli, the presence of abundant indigenous micro¯ora, and the existence of groups of short rods dispersed on the brush borders of the absorptive epithelium were also observed. 4. Discussion Plesiomonas shigelloides and some members of the genus of Aeromonas have been involved in gastrointestinal and extra-intestinal infections, both in humans and animals [1,4,9,11,16]. However, the identi®cation and understanding of their true role in enteric disease has been hampered by the lack of an animal model incorporating reproducible ful®lment of Koch's postulates. In addition, human volunteer studies and primate challenge studies with Aeromonas spp. have also been inadequate [17,18] in this matter. On the other hand, the presence of these pathogens in stools of patients with diarrhoea in the absence of other enteric pathogens indicates their possible role in the gastrointestinal diseases [6]. The protozoan parasite C. parvum is an important and widely distributed
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
51
enteric pathogen in humans and young livestock, as indicated by increasing documentation of disease outbreaks which are associated with presence of cryptosporidia in public water supplies [19,20]. Since both animals and humans can be coinfected with dierent pathogenic
Fig. 4. SEM micrographs: (a) ileum from neonatal mouse infected with A. hydrophila (2 DPI) showing absorptive cells on the villous tip covered with focal ®ne membranes (scale bar=10 mm); (b) mouse colon coinfected with A. hydrophila and C. parvum (5/3) DPI) showing vesicular protrusion above the inner mucosal surface (scale bar=1 mm).
52
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
microorganisms, and since various relationships can exist between the immunoresponse generated by the dierent pathogens, the pathogenicity of the infection of one pathogen can be in¯uenced by the presence of another. Coinfections with other enteropathogens and a potential predisposing factor in P. shigelloides infections have been described [9]. In addition, spontaneous coinfection of P. shigelloides with C. parvum and other coccidia has also been reported in animals [21]. The basic aim of this study was to investigate and studied the enteropathogenicity of P. shigelloides, A. hydrophila, A. caviae and A. sobria both in monoinfections and in coinfections with the protozoan parasite C. parvum in neonatal BALB/c mice. Implicit in our experiments and results is a model that includes long-term colonisation of P. shigelloides in the small and large intestine, which was found to be frequently accompanied with bacteremia, since P. shigelloides was isolated from the spleen of infected mice from the 1st to 5th day of postinoculation (1±5 DPI). Histological ®ndings revealed pathological lesions, which varied in severity from atrophy to necrosis of the mucosal inner surface of the ileum and colon, with predilection to the colon and brush border of colonic enterocytes. Ultrastructural TEM observations of samples from infected intestinal mucosa revealed shrinkage in the size of enterocytes and reduction in the number and size of microvilli during the initial stages of the pathological process. Even mitochondria and micro®laments in the widened terminal web were reduced in number. These initial ultrastructural pathological alterations of the intestine were simultaneously accompanied by necrobiosis, predominately observed in the brush borders and supranuclear parts of the enterocytes. Interestingly, the eects of coinfection with C. parvum were characterised by more intense and heavier colonisation of the intestine by P. shigelloides. In addition, extensive necrotising in¯ammatory changes in the ileum and colon, accompanied by bacteremia, diarrhoea and deaths of the mice were also recorded. Our results agree with those of Vitovec et al. [22], who observed similar interactions between Campylobacter jejuni and C. parvum in experimentally infected BALB/c neonatal mice. These investigators found that the synergistic interaction of C. jejuni and C. parvum resulted in heavier colonisation of the inner surface of the small and large intestine by C. jejuni. With the experimental designs described in this article as a backdrop, our results along with those of previous coinfection studies on C. jejuni and C. parvum, suggest that the colon and the distal third of small intestine are the target sites for P. shigelloides and C. jejuni infections. These observations seem to resemble rather well the pathological observations often encountered in cases of human colitis for which spontaneous plesiomonads infections are involved [23,24]. The virulence in Aeromonas spp. may well be multifactoral and polygenic in nature, involving a complex interaction and bacterial and host factors [1,11]. For instance, protein de®ciency in experimental mice led to bowel colonisation by aeromonads [25], and streptomycin treatment predisposed mice to A. sobria and A. hydrophila colonisation, whereas control mice that were not treated with
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
53
streptomycin were in turn not colonised [26]. It is also known that coinfection by other enteropathogens frequently occurs in Aeromonas-associated diarrhoea in infants [27]. Moreover, coinfection with cryptosporidia might indeed have contributed to the severe diarrhoea encountered in 9% of foals infected by A. hydrophila [28]. These observations and others suggest that the neonatal BALB/c mouse model that is implicit in our study may serve well as a tool for studying the initial steps of gastrointestinal colonisation of Aeromonas spp. The results from the Aeromonas spp. monoinfections showed only a short term colonisation of mouse intestine by aeromonads, whereas mice coinfected with A. hydrophila and C. parvum were found to have a prolonged growth period for these bacteria in the spleen and intestine. However, coinfection with C. parvum did not substantially further in¯uence the intestinal colonisation. Within the experimental groups of enterocytes in our study, the TEM observations of samples infected by aeromonads revealed a disconnection of interdigitating folds, and they also revealed, beneath the zonula occludens, the formation of widely dilated intercellular spaces between enterocytes. These were usually found above the capillaries, and speci®cally with oedematous and hypertrophied walls in the propria. In addition, neither clinical signs of diarrhoea nor substantial pathomorphological changes on the inner mucosal surface, nor in¯uence on colonisation and patent period of C. parvum, were observed in the mice infected by the Aeromonas spp. strains that were used. In conclusion, the results from the present investigation indicate that experimental infection of neonatal BALB/c mice with P. shigelloides, Aeromonas spp. and C. parvum, including coinfections of the bacteria with the protozoan, may serve well as an initial experimental animal model for studying the initial steps of gastrointestinal colonisation and the diarrhoeal disease syndrome caused by these bacterial and protozoan pathogens. Such a model can serve as a step leading toward improved models for understanding gastrointestinal colonisation and the diarrhoeal disease syndrome.
Acknowledgements This study was supported by a grant from the Royal Swedish Academy of Sciences and Czech Academy of Sciences (KK and JV). The authors gratefully acknowledge Dr. B. Koudela for assistance in preparing C. parvum and Dr. A. Faris for his critical comments and suggestions during the preparation of the manuscript and Dr. Valerie Amarger for the French translation of the resumeÂ.
References [1] Janda JM. Recent advances in the study of the taxonomy, pathogenicity, and infectious syndromes associated with the genus Aeromonas. Clin Microbiol Rev 1991;4:397±410.
54
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
[2] Abbott SL, Kokka RP, Janda JM. Laboratory investigations on the low pathogenic potential of Plesiomonas shigelloides. J Clin Microbiol 1991;29:148±53. [3] Brenden RA, Miller MA, Janda JM. Clinical disease spectrum and pathogenic factors associated with Plesiomonas shigelloides infections in humans. Rev Infect Dis 1988;10:303±16. [4] Herrington DA, Tzipori S, Robins-Browne R, Tall BD, Levine MM. In vitro and in vivo pathogenicity of Plesiomonas shigelloides. Infect Immun 1987;55:979±85. [5] Holt JG, Krieg NR, Sneath PHA, Staley JT, Williams ST, editors. Bergey's manual of determinative bacteriology. 9th ed. Baltimore, USA: Williams & Wilkins, 1994. [6] Khardori N, Fainstein V. Aeromonas and Plesiomonas as etiological agents. Ann Rev Microbiol 1986;42:395±419. [7] KuÈhn I, Allestam G, Huys P, Jannssen K, Kersters K, Krovacek K, StenstroÈm TA. Diversity, stability and virulence properties of Aeromonas strains isolated from drinking water distribution systems in Sweden. Appl Environ Microbiol 1997;63:2708±15. [8] Janda JM, Duey PS. Mesophilic aeromonads in human disease: current taxonomy, laboratory identi®cation and infectious disease spectrum. Rev Infect Dis 1988;10:980±97. [9] Clark RB, Westby GR, Spector H, Soricelli RR, Young CHL. Fatal Plesiomonas shigelloides septicaemia in a splenectomised patient. J Infect 1991;23:89±92. [10] Ljungh A, WadstroÈm T. Aeromonas toxins. In: Dorner F, Drews J, editors. Pharmacology of bacterial toxins, vol. 119. Oxford: Pergamon Press, 1986. p. 289±301. [11] Krovacek K, Conte M, Galderisi P, Morelli G, Postiglione A, Dumontet S. Fatal septicaemia caused by Aeromonas hydrophila in a patient with cirrhosis. Comp Immun Microbiol Infect Dis 1993;16:267±72. [12] Krovacek K, Pasquale V, Baloda SB, Soprano V, Conte M, Dumontet S. Comparison of putative virulence factors in Aeromonas hydrophila strains isolated from the marine environment and human diarrheal cases in southern Italy. Appl Environ Microbiol 1994;60:1379±82. [13] Krovacek K, Faris A, Mansson I. In vitro invasion of Aeromonas spp. to HEp-2 tissue culture cells. Acta Vet Scand 1991;32:39±43. [14] Krovacek K, Dumontet S, Eriksson E, Baloda SB. Isolation, and virulence pro®les of Aeromonas hydrophila implicated in an outbreak of food poisoning in Sweden. Microbiol Immunol 1995;39:655±61. [15] Vitovec J, Koudela B. Location and pathogenicity of Cryptosporidium parvum in experimentally infected mice. J Vet Med A 1988;35:515±24. [16] Hathcock TL, Schumacher J, Wright JC, Stringfellow J. The prevalence of Aeromonas species in faeces of horses with diarrhoea. Vet Intern Med 1999;13:357±60. [17] Morgan DR, Johnson SR, Chattopadhyyay B. Lack of correlation between known virulence properties of Aeromonas hydrophila and enteropathogenicity for humans. Infect Immun 1985;50:62±5. [18] Pitarangsi C, Echeverria P, Whitmire R, Tirapat C, Formal S, Dammin GJ, Tingtalapong M. Enteropathogenicity of Aeromonas hydrophila and Plesiomonas shigelloides: prevalence among individuals with and without diarrhea in Thailand. Infect Immun 1982;35:666±73. [19] Dubey JP, Speer CA, Fayer R. Cryptosporidiosis of man and animals. Boston: CRC Press, 1990. [20] Fayer R. Cryptosporidium and cryptosporidiosis. Boca Raton, FL: CRC Press, 1997. [21] GluÈnder G, Zum von Vorkommen. Plesiomonas shigelloides bei Wild-und Zoovogeln. Berl MuÈnch TieraÈrztl Washr 1988;101:334±7. [22] Vitovec J, Koudela B, Vladik P, Hausner O. Interaction of Cryptosporidium parvum and Campylobacter jejuni in experimentally infected neonatal mice. Zbl Bakt 1991;274:548±59. [23] Loon FP, Rahim Z, Chowdhuri KA, Kay BA, Rahman SA. Case report of Plesiomonas shigelloides-associated persistent dysentery and pseudomembranous colitis. J Clin Microbiol 1989;27:1913±5. [24] Mandal BK, Whale K, Morson BC. Acute colitis due to Plesiomonas shigelloides. Brit Med J 1982;285:1539±40. [25] Pazzaglia G, Winoto I, Jennings G. Oral challenge with Aeromonas in protein-malnourished mice. J Diarrhoeal Dis Res 1994;12:108±12.
J. Vitovec et al. / Comp. Immun. Microbiol. Infect. Dis. 24 (2001) 39±55
55
[26] Sanderson K, Ghazali FM, Kirov SM. Colonization of streptomycin-treated mice by Aeromonas species. J Diarrhoeal Dis Res 1996;14:27±32. [27] Pazzaglia G, Sack RB, Salazar E, Yi A, Chea E, Leon-Barua R, Guerrero CE, Palomino J. High frequency of coinfecting enteropathogens in Aeromonas-associated diarrhea of hospitalized peruvian infants. J Clin Microbiol 1991;29:1151±6. [28] Browning GF, Chalmers RM, Snodgrass DR, Batt RM, Hart CA, Ormarod SE, Leadon D, Stoneham SJ, Rossdale PD. The prevalence of enteric pathogens in diarrhoeic thoroughbred foals in Britain and Ireland. Equine Vet J 1991;23:405±9.