Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
General enquiries on this form should be made to: Defra, Science Directorate, Management Support and Finance Team, Telephone No. 020 7238 1612 E-mail: [email protected] SID 5 Research Project Final Report Note In line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects. A SID 5A form must be completed where a project is paid on a monthly basis or against quarterly invoices. No SID 5A is required where payments are made at milestone points. When a SID 5A is required, no SID 5 form will be accepted without the accompanying SID 5A. This form is in Word format and the boxes may be expanded or reduced, as appropriate. ACCESS TO INFORMATION The information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors. SID 5 (2/05) Project identification 1. Defra Project code 2. Project title OZ0710 The role of goats and pigs in the maintenance and transmission of VTEC including EHEC O157:H7 3. Contractor organisation(s) Veterinary Laboratories Agency Woddham Lane New Haw Addlestone Surrey KT15 3NB 54. Total Defra project costs 5. Project: Page 1 of 18 £ start date ................ 01 October 2002 end date ................. 30 September 2005 6. It is Defra’s intention to publish this form. Please confirm your agreement to do so. ................................................................................... YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow. Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer. In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. (b) If you have answered NO, please explain why the Final report should not be released into public domain Executive Summary 7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work. Enterohaemorrhagic Escherichia coli O157:H7 infections of man have been associated with consumption of unpasteurised goat’s milk and direct contact with kid goats on petting farms and yet little is known about colonisation of goats with this organism. To assess coloniosation and transmission in goats a series of deliberate oral inoculation studies were performed with goats of differing ages. Four conventionally reared goats aged 6 days were inoculated orally with approximately 1010 colony-forming units (cfu) of a non-verotoxigenic strain of Escherichia coli O157:H7. All remained clinically normal. Tissues were sampled under terminal anaesthesia at 24 (two animals), 48 and 72 h postinoculation (hpi). E. coli O157:H7 was cultured from ileum, caecum, colon and rectum from all animals, but the number of bacteria recovered at these sites varied between animals. Attaching effacing (AE) lesions associated with O157 organisms, as confirmed by immunolabelling, were observed in the ileum of one of the two animals examined at 24 hpi, and in the ileum, caecum and proximal colon of an animal examined at 72 hpi. E. coli O157 organisms were detected at ≥ 105 cfu/g of tissue at these sites. In addition, AE lesions associated with unidentified bacteria were observed at various sites in the large bowel of the same animals. Lesions containing both E. coli O157 and unidentified bacteria (non-O157) were not observed. Non-O157 AE lesions were also observed in the large bowel of one of two uninoculated control animals. This indicated that three (one control and two inoculated) animals were colonized with an unidentified AE organism before the commencement of the experiment. The O157-associated AE lesions were observed only in animals colonized by non-O157 AE organisms and this raises questions about individual host susceptibility to AE lesions and whether non-O157 AE organisms influence colonization by E. coli O157. Eight-week-old conventionally-reared goats were inoculated orally in separate experiments with 1 x 1010 colony-forming units (cfu) of a non-verotoxigenic strain of E. coli O157:H7 (strain NCTC12900 Nalr), an aflagellate derivative (DMB1) and an intimin deficient derivative (DMB2). At 24h after inoculation the three E. coli O157:H7 strains were shed at approx. 5 x 104 cfu per gram of faeces from all animals. Significantly fewer initimin deficient bacteria were shed only on days 2 (p=0.003) and 4 (p=0.014) whereas from day 7 to 29 days there were no differences. Tissues from three animals inoculated with wild-type E. coli O157:H7 strain NCTC 12900 Nalr were sampled at 24, 48 and 96 hours after inoculation and the organism was cultured from the large intestine of all three animals and from the duodenum and the ileum of the one animal examined at 96 hours. Tissues were examined histologically but AE lesions were not observed at any intestinal site of the animals examined at 24 or 48 hours. However, the animal examined at 96 hours, which uniquely shed approximately, 1 x 107 E. coli O157:H7 per gram of faeces for the preceding three days, showed a heavy, diffuse infection with cryptosporidia and abundant, multifocal AE SID 5 (2/05) Page 2 of 18 lesions in the distal colon, rectum and at the recto-anal junction. These AE lesions were confirmed by immunohistochemistry to be associated with E. coli O157:H7. Five 5-year-old Nanies with kids at foot were dosed orally with 10 10 cfu O157:H7 at about six days after parturition. Four 4 of the 5 nannies were colonised and that faecal shedding was variable up to day 11 after dosing. Several samples containing greater than 1 x 104 E. coli O157:H7 cfu per gram. This short lived suggested antagonistic factors may have prevented colonisation. One possible route for kid goats to become infected with E. coli O157:H7 is by sucking at their nanny on teats contaminated with faeces. In this study the cleanliness of the teats and udders may be attributed to sucking and, if it is assumed that some faeces containing E. coli O157:H7 organisms had been on the teats, even transiently, it is probable that the kids were challenged. Faeces containing E. coli O157:H7 organisms shed into the environment were a source of infection of the kids also. No kids became positive for E. coli O157:H7 organisms. Thus, it may be argued that sucking kids are refractive to colonisation by E. coli O157:H7 possibly because the challenge dose was low and the kids had access to milk, that is known to be highly protective. However, these hypotheses need to be examined further. Isolation of Shiga-toxin (Stx) positive Escherichia coli O157:H7 from commercially-grown pigs has been reported and experimental infection studies have demonstrated that this pathogen can persist in 12week-old experimentally orally inoculated conventional pigs for up to 2 months. Intimin is not required for persistence, but we have shown that the flagellum of Stx-negative E. coli O157:H7 does not have a role to play in pathogenesis in ruminant models whereas, in poultry, the flagellum of E. coli O157:H7 was important for long-term persistent infection. The contribution of the flagellum of Stx-negative E. coli O157 in the colonisation of pigs was investigated by adherence assays on porcine (IPI-21) cell line and IVOC and experimental oral inoculation of conventional 14-week-old pigs. E. coli O157:H7 NCTC12900nalr and isogenic aflagellate and intimin deficient mutants adhered equally well to IPI-21 cell line cells. In porcine IVOC association assays E. coli O157:H7 NCTC12900nalr was associated in significantly higher numbers to tissues from the caecum and the terminal rectum than other sites. The aflagellate and intimin deficient mutants adhered to tissues from the duodenum, jejunum and ileum in significantly higher numbers than the parent. Groups of 14-week-old pigs were dosed orally with 1010 CFU/10ml of either E. coli O157:H7 NCTC12900nalr or isogenic aflagellate and intimin deficient mutants and recovery of each test strain was similar. Histological analysis of pig tissues at post mortem examination revealed that E. coli O157 specifically stained bacteria were associated with the mucosa of the ascending and spiral colon . These data suggest that for Stx-negative E. coli O157:H7 to colonise and persist in pigs, involves mechanisms that do not require the flagellum or intimin. SID 5 (2/05) Page 3 of 18 Project Report to Defra 8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer). Purpose Study VTEC, especially Escherichia coli O157:H7 may cause Haemolytic Uraemic Syndrome (HUS), Haemorrhagic Colitis (HC), Thrombotic Thrombocytopaenic Purpura and sporadic Infantile Diarrhoea in humans. These syndromes are potentially fatal in the young, elderly and immuno-compromised. Recent trace back and other epidemiological studies indicate that goats may be a significant source for the transmission and maintenance of E. coli O157:H7. Experimental infection of gnotobiotic and neonatal (12h of age) pigs with high doses of E. coli O157 is associated with neuronal damage (mediated by VT toxins) and diarrhoea (mediated by colonisation of the GI mucosa). However, nothing is known of the colonisation, persistence and excretion of E. coli O157:H7 or other VTEC in conventional either goats or pigs. In order to assess the risk to human health arising from un-pasteurised goat’s milk and pig meat, it is essential that DEFRA is informed of the behaviour of VTEC including E. coli O157:H7 in these hosts. The aims of this proposal are (a) to study the pathogenicity in terms of colonisation, persistence and excretion of VTEC including E. coli O157:H7 in goats and pigs, (b) to study potential transmission routes and (c) the possible mode(s) of mammary infection in the goat study. Please note that all references cited in the project report are listed in a separate appendix entitled OZ0710 literature. Also, all figures and tables cited in the body of the text are provided as in a separate appendix entitled OZ0710 supplementary data. GOAT STUDIES Detailed Introduction Escherichia coli serotype O157:H7 infection was first recognised in the early 1980’s to be associated with haemorrhagic colitis [HC], haemolytic-uraemic syndrome [HUS] and thrombocytopaenic purpura in man (Karmali et al., 1983; Riley et al., 1983). Human infection and the sequelae of infection have been well-documented subsequently (Smith & Scotland, 1993; Boyce et al., 1995; Swinbanks, 1996) and E. coli O157:H7, classified as belonging to the recently defined Enterohaemorrhagic E. coli [EHEC] pathotype, is regarded worldwide as the leading cause of both HC and HUS. Transmission of E. coli O157:H7 is faecal-oral (Pepin et al., 1997; Locking et al., 2001) with cattle considered as the primary reservoir (Griffin & Tauxe, 1991) and sheep recognised as a significant reservoir also (Chapman et al., 1997; Heuvelink et al., 1998; Meng et al., 1998; Fegan & Desmarchelier, 1999). Other potential sources of these organisms have been described including rabbits and seagulls amongst other species (Griffin & Tauxe, 1991; Pritchard et al., 2001). Several recent reports cite goats as potential sources of E. coli O157:H7 infection. Shukla et al. (1995) reported four human cases of bloody diarrhoea affecting one adult and three children aged two to four years, which were traced to a farm visitor centre in Leicestershire, UK. Strains of E. coli O157:H7 phage type (PT) 2, all with identical restriction fragment length polymorphism (RFLP) patterns, were isolated from the cases and from four cattle and six goats. Three other incidents in the UK have implicated goats on visitor farms as potential sources of infection (Chapman et al., 2000; Pritchard et al., 2000; Payne et al., 2003). On continental Europe, Heuvelink et al. (2002) showed that two of eleven petting zoos had goats that were positive for E. coli O157:H7. In Bohemia, four cases of HUS in children were reported and the source was identified as unpasteurised goat’s milk, with E. coli O157:H7 PT2 isolated from a goat and an asymptomatic human carrier who drank goat’s milk (Bielaszewska et al., 1997). An outbreak of E. coli O157:H7 associated with unpasteurised goat’s milk has been described in Canada also (McIntyre et al., 2002). Foschino et al. (2002) reported a 1.7% incidence rate of E. coli SID 5 (2/05) Page 4 of 18 O157:H7 in samples of raw goat’s milk intended for cheese making in the Bergamo region of Italy, although it was not associated with infection in man. Despite these incidents involving the goat as the primary source, the biology of E. coli O157:H7 in this host is not understood. Little is known of the prevalence of E. coli O157:H7 in this host with only a few reports suggesting low prevalence rates in the region of 0.2% (Blanco et al., 2003; Dontorou et al., 2004). Attaching-effacing E. coli (AEEC) belonging to non-O157 serogroups are found frequently in diarrhoeic goats (Duhamel et al., 1992; Drolet et al., 1994) and putative AEEC were reported in healthy goats (Cid et al., 2001; de la Fuente et al., 2002; Orden et al., 2003a,b). Furthermore, E. coli belonging to serogroup O145 induced attaching-effacing lesions in an adult goat (Barlow et al., 2004). Collectively, these observations suggest that goats may be a permissive host for AEEC including E. coli O157:H7. Recently, it was shown that six-day-old conventional kids were colonised by E. coli O157:H7 and that small attaching-effacing (AE) lesions in the mucosa of the ileum and large intestine were observed following oral inoculation (Wales et al., 2005, in press). These data suggest that goats are susceptible to colonisation by E. coli O157:H7, probably in an intimin dependent manner, but whether this is the case for older goats remains untested. In this study, we wished to assess the contribution of two bacterial factors, intimin and flagella, in colonisation. Intimin is considered to play a role as a key effector in the induction of attaching effacing lesions whereas flagella have been shown to contribute to persistence, particularly of commensal E. coli, in many species. E. coli O157:H7 strain NCTC12900 Nalr which we have shown in previous studies colonises small ruminants and isogenic intimin and flagella deficient mutants of this strain were selected for this study. Weaned, conventional, eight-week-old goats were orally inoculated with the test strains and, to assess colonisation, faecal shedding was monitored and serial post mortem examinations were performed to locate the organism within the gastro-intestinal tract. Recently, we showed that six-day-old conventional kids may be colonised by E. coli O157:H7 and that small attaching-effacing (AE) lesions induced by E. coli O157:H7 in the mucosa of large intestine were observed after deliberate oral inoculation (Wales and others 2004). Whilst these data suggest that kid goats are susceptible to colonisation by E. coli O157:H7, what remains unclear, however, is whether older goats are also susceptible to colonisation by E. coli O157:H7 and whether transmission from affected nannies to presumed susceptible neonate kids at foot is a risk factor in the management of goats on petting zoos and farms. Here we describe an experiment in which nannies with kids at foot were inoculated orally with E. coli O157:H7 and the fate of the dosed organisms was followed in order to assess these issues. All animal procedures were approved following ethical review at the VLA and complied with the animals scientific procedures act (1986), and were performed under Home Office Licence 70/5722. Previous studies in another small ruminant model, lambs and sheep, made use of E. coli O157:H7 strain NCTC 12900 Nalr. This strain does not produce verocytotoxins and can be used at containment level II. It also generates a localised adherence phenotype on cultured HEp-2 and bovine primary epithelial cells with actin rearrangements characteristic of the attaching effacing (AE) lesion (Dibb-Fuller and others 2001, Cookson and others 2003). Furthermore, this strain forms AE lesions in the large intestine of orally inoculated six-day-old goats (Wales and others 2004), in the large intestine of six-week-old lambs (Wales and others 2001a) and on the mucosa in ligated large intestinal loops in six-month-old lambs (Wales and others 2002). Additionally, this strain persists in six-week-old and six-month-old conventional lambs (Cookson and others 2002, Wales and others 2001b, Wales and others 2004, Woodward and others 2003). METHODS Bacterial strains and inocula A derivative of E. coli O157:H7 strain NCTC12900 that does not possess either stx1 or stx2 verocytotoxin genes (NCTC, Health Protection Agency, Colindale, London) (Best et al., 2005, in press) was made resistant to -1 by passage on complex medium supplemented with the antibiotic and designated nalidixic acid at 15 E. coli O157:H7 strain NCTC12900 Nalr . Mutants defective for the elaboration of flagella and intimin were constructed in E. coli O157:H7 strain NCTC12900 Nalr following the methods described previously by Best et al. (2005, in press). The flagella deficient derivative was made with a streptomycin resistance gene cassette inserted in the fliC gene, fliC::strr, and was designated DMB1. This derivative was non-motile in 0.35% nutrient agar, did not elaborate flagella as determined by transmission electron microscopy and did not agglutinate H7 anti-sera. The intimin deficient derivative was made with a chloramphenicol resistance gene cassette inserted in the eae gene, eae::camr, and was designated DMB2. This derivative failed to induce AE lesions in HEp-2 cells. The growth rates in Luria-Bertani (LB) broth for both isogenic derivatives were the same as for the parent strain. Strain NCTC12900 Nalr and isogenic derivatives were stored in Heart Infusion broth (HIB) medium supplemented with 30% (w/v) glycerol on beads at –80oC and working stocks were stored at room temperature on Dorset’s egg medium. NCTC12900 Nalr and isogenic derivatives were streaked from Dorset’s egg medium onto SMAC plates containing nalidixic acid (15 ml-1) and well-isolated colonies were inoculated separately into 100 ml aliquots of LB broth in 250 ml conical flasks. After incubation for 16 hours at 37˚C with gentle agitation the bacterial cells were harvested by centrifugation (3000 g for 10 min) and resuspended in PBS (pH 7.4). The bacterial suspensions contained approximately 1 x 109 cfu ml-1 as determined by serial dilution and plating. SID 5 (2/05) Page 5 of 18 Animals Six day old kid studies. Six goat kids (A-F; Table 1) were allowed to suckle for 3 or 4 days after parturition, and then received milk replacer by bottle four times a day. Two days before inoculation with E. coli O157:H7, the rectum of each animal was swabbed and cultured for E. coli O157 as described below. Four of these animals (C-F) were inoculated orally with E. coli O157:H7 NCTC 12900 Nal r administered by syringe at 6 (three animals) or 7 (one animal) days of age. Two animals, (A, B), housed separately, remained as uninoculated controls for histopathological examination. Animals were observed at least four times a day by a veterinary surgeon. Six week old kid studies. Post-partum, twenty neonatal goat kids were allowed to suckle and after weaning at about 6 weeks of age were separated from their respective nannies. The animals were selected randomly into three groups comprising eight (A), six (B) and six (C) animals with each group penned separately. Each animal was identified with duplicate eartags encoding a unique four digit identification number and housed indoors for a further two weeks provided with standard rations and water ad libitum. Prior to inoculation with E. coli O157:H7, faeces were taken from the rectum of each animal and cultured for E. coli O157, as described below. All animals in each of the three groups were inoculated orally with one challenge strain; group A was challenged with E. coli O157:H7 NCTC12900 Nalr, group B with DMB2 and group C with DMB1. The challenge doses (10ml) were administered orally using a worming gun, ensuring delivery of the inoculum to the pharynx. Rectal faecal samples were taken from each animal 1, 2, 4, 7, 8, 10, 11, 15, 18, 22, 25 and 29 days after challenge. Animals were observed at least twice daily to monitor clinical status. Nannies with kinds at foot studies. Faecal samples were collected from pregnant nanny goats of about five years of age and shown to be free of O157 by bacteriological analysis of faeces. These nannies were then brought into a contained class II environment. After parturition, the kids were permitted to suckle nomally. Nannies were then dosed orally as described above with 1010 cfu of NCTC12900 Nalr. All animal procedures were approved following ethical review at the VLA and complied with the animals scientific procedures act (1986), and were performed under Home Office Licence 70/5722. Necropsy procedures From six-week-day kids, tissue samples were collected under terminal pentobarbitone anaesthesia from the inoculated goats at 24 (two animals), 48 or 72 h post-inoculation (hpi). On each occasion, the goat was placed in left lateral recumbency and the right flank and inguinal areas were soaked with a chlorhexidine-based surgical scrub. An incision in the exposed flank, followed by dissection of the pelvic cavity, enabled samples to be collected aseptically from the gastrointestinal tract. Samples, including the intestinal wall and associated contents, were collected from the abomasum, duodenum, jejunum, ileum, caecum, proximal ascending colon, mid-spiral and descending colon, proximal, mid and distal rectum and the recto-anal junction (RAJ), for bacteriological and histological examination. In addition, samples from the mesenteric lymph nodes, liver and kidney were collected for histological examination only. To prevent cross-contamination of samples, clips and plastic bags were used to enclose the cut ends of the intestinal tract as dissection proceeded, as described previously in sheep (Wales et al., 2001a). The necropsy procedures were as described previously (Wales et al., 2005). Briefly, from group A, goats were selected at random at 24 (animal 1019), 48 (animal 1001) and 96 (animal 1003) hours after challenge and tissue samples were collected immediately after euthanasia with barbiturate overdose. Samples were collected from the rumen, duodenum, jejunum, ileum, caecum, ascending colon, spiral colon and from six sites excised at approximately 2cm apart measuring from the recto-anal junction (RAJ), along the rectum toward the distal colon. The samples were for bacteriological, histological examination and electron microscopy as described below. Clips and plastic bags were used to enclose the cut ends of the intestinal tract as dissection proceeded, as described previously in sheep (Wales et al., 2001a), to prevent cross-contamination of samples. Bacteriological examination Faecal samples taken prior to oral inoculation were examined for the presence of E. coli O157 following previously described methods (Wales et al., 2001a,b; 2002; Woodward et al., 2003). Faeces (1g) were resuspended in 9 ml Buffered Peptone Water (BPW), incubated at 37oC for six hours and O157:H7 organisms recovered by O157 specific IMS and plated onto CT-SMAC plates. To detect the strains used to orally inocluate the animals, previously described methods were followed (Wales et al., 2001a,b; 2002; Woodward et al., 2003). Faeces (1g) were re-suspended in 9 ml BPW by vortexing -1). Additionally, and serial dilutions were plated directly onto SMAC plates containing nalidixic acid (15 dilutions were retained overnight at 4C and, if there were no direct counts of confirmed O157 organisms, the dilutions were incubated at 37C for six hours and samples were plated on SMAC plates containing nalidixic acid -1). The serogroup of bacteria recovered by these processes was checked by O157-specific latex agglutination (Oxoid, Basingstoke, UK). Tissue samples collected from kids for bacteriological examination were homogenised in BPW (1 g in 9 ml) and processed as described above. SID 5 (2/05) Page 6 of 18 Pathological studies Light microscopy. Tissues were placed immediately into 10% neutral buffered formalin at room temperature and left to fix for at least 24 hours. Trimmed tissues were processed routinely to paraffin wax, and 4 μm sections were stained with Haematoxylin and Eosin (H & E) and observed using an Olympus CX41 microscope. Immunohistochemistry. Tissue blocks were fixed in 10% neutral buffered formalin, processed to wax and sectioned at 4 μm on a rotary microtome in preparation for immuno-histochemistry. Tissue sections were dewaxed in xylene, dehydrated in absolute alcohol before immersion in freshly prepared 3% hydrogen peroxidemethanol for 10 minutes to inhibit endogenous peroxidase activity. Sections were rehydrated in running tap water prior to assembly into the Shandon Sequenza ® staining system. The slides were washed with 2 x Sodium Chloride Tris Buffered Saline (0.005M TBS, pH 7.6 1.7% NaCl) for 5 minutes before incubation at room temperature with a normal goat serum (Vector Labs, UK) for 20 minutes. E. coli O157 specific polyclonal antibody, raised in rabbits (VLA, Weybridge), was then applied (1/1000 and 1/5000 diluted in 2 x NaCl TBS supplemented with 5% normal rabbit serum) to the sections for 1 hour at room temperature (Wales et al., 2001b). E. coli antigens were visualised following incubation with biotinylated goat anti-rabbit IgG (1/200 dilution, with normal goat serum 1/66 dilution in 2 x NaCl TBS supplemented with 3% normal sheep serum) for 30 minutes at room temperature, an avidin–biotin-peroxidase conjugate (Vector Labs, UK) for 30 minutes at room temperature and citrate-buffered diaminobenzidine (DAB) for 10 minutes at room temperature. The sections were then counter-stained in Meyer’s Haematoxylin, before being dehydrated in absolute alcohol, cleared in xylene and coverslipped using DPX mountant. The stained tissue sections were then examined by light microscopy. Electron microscopy. After removal from the animal, tissues were fixed in 3% gluteraldehyde prepared in a 0.1M phosphate buffer. Following microscopical examination of the HE sections, at sites where lesions were identified, corresponding gluteraldhyde fixed tissues were cut to 1-2mm in thickness for TEM examination. Briefly the tissues were then washed in 0.1M phosphate buffer, post fixed in 1% osmium tetroxide, dehydrated by immersion in a series of alcohol solutions increasing to 100% alcohol and placed in propylene oxide prior to embedding in araldite resin. The resin was polymerised at 600C for 48 hours. One-micron sections were stained with toluidine blue for light microscopical examination. Ultra-thin sections at 70-90 nm thickness were then prepared onto copper grids using a diamond knife and stained with uranyl acetate and lead citrate prior to examination using a Phillips CM10 TEM. Statistical analyses The sensitivity of detection by direct plating was approximately 500 organisms per gram of faeces. Samples positive by enrichment were considered to have up to 500 organisms per gram, and those samples in which no organisms were detected were given an arbitrary value of 1 to avoid the issue of a zero value giving results to infinity. The non-parametric Kruskal-Wallis test was used to compare the mean group counts for each occasion. When this was significant at p≤0.05 the individual group means were compared by a multiple comparison test. The percentages of animals positive were also compared by Fisher’s Exact test at each timepoint. RESULTS: SIX-DAY-OLD KIDS Clinical Findings. The kids remained normal throughout the experiment, with no evidence of pyrexia or diarrhoea. Bacteriological Findings. All rectal swabs collected from the kids before the experimental procedure were negative for E. coli O157. The oral inoculation doses were 9 x 109 cfu for kid C and 3 x 1010 cfu for kids D, E and F. E. coli O157 was only recovered from intestinal tissues, and there were marked differences between animals in respect of the numbers of O157 organisms recovered from tissues (Table 2), particularly between the two animals (C and D) examined at 24 hpi. Kid C yielded counts of between 103 and 105 cfu/g from each of seven tissue samples representing different sites from the ileum to the mid rectum; kid D, which had received approximately three times the dose of kid C, also gave counts of that order but only from the ileum, caecum and proximal colon. Kids E and F, examined at 48 and 72 hpi respectively, yielded consistently high counts of E. coli O157 organisms from eight sites (ileum to distal rectum). Histopathological Findings. Most of the histological sections of intestine demonstrated a good state of preservation, but a minority showed evidence of slight autolysis (minor separation of the luminal epithelium from the underlying submucosa). In some infected animals, AE lesions of varying size were seen on the intestinal mucosal epithelium at a number of sites (Table 2). The number of lesions also varied substantially, ranging from one per section to one or more per high power field (x 40 objective); confluent lesions were not seen. AE lesions were arbitrarily categorized as small (covering up to nine adjacent enterocytes in the plane of the section), medium (covering 10 to 20 enterocytes) or large (covering more than 20 enterocytes). AE lesions were present in the ileum of two animals, namely, kid C (24 hpi), in which they were small and sparse, and kid F (72 hpi), in which SID 5 (2/05) Page 7 of 18 they were more common and of small or medium size (Fig. 1). Lesions were not observed on the follicleassociated epithelium over Peyer’s patches in the ileum. In the caecum, AE lesions were present in control kid A (common, small or medium-sized), in kid C (24 hpi; sparse and small) and in kid F (72 hpi; common, small or medium-sized). Similar lesions were present at moderate or high density, in a range of sizes from small to large, in the colons of kids A (control; Fig. 2), C (24 hpi) and F (72 hpi). In the same three kids, AE lesions of varying size were common in proximal, middle and distal rectal sections, and on rectal mucosa immediately adjacent to the RAJ. Only in kids C and F were AE lesions seen at the RAJ; these lesions were not specifically associated with lymphoid tissue. AE lesions were not seen in tissues from kid B (control), D (24 hpi) or E (48 hpi). Mild or moderate fusion and atrophy of duodenal villi was present in control kids A and B and in kids E (48 hpi) and F (72 hpi). There was an extensive layer of adherent Gram-positive, coccoid bacteria on a few duodenal villi of kid D. Diffuse, and occasionally focal, infiltrates of eosinophils and neutrophils were observed at low or moderate density in the small intestinal lamina propria of control and inoculated animals. In the large intestines, enterocyte degeneration and detachment was evident at the site of many AE lesions. Mucosal epithelial erosions that did not have overlying AE lesions in the plane of section were observed occasionally. However, such erosions were seen only in samples in which AE lesions were present elsewhere. Eosinophil infiltrates in the large intestinal lamina propria were generally of low density, and were minimal or absent in the rectal sections. Neutrophil infiltrates in the lamina propria of large intestinal sections containing AE lesions were often moderately dense, and diffuse, patchy or concentrated subepithelially. Fewer neutrophils were present in large intestinal sections that did not contain AE lesions. However, there were exceptions, notably the proximal ascending colon of kid B and the distal rectum of kid D, which lacked detectable AE lesions but had relatively dense neutrophil infiltrates; and the descending colon of kid C, which contained numerous AE lesions but few neutrophils. The most marked acute inflammatory changes were seen in the tissues of the RAJ, in both control and inoculated animals. These were associated in some animals with microabscesses in the dermis or epidermis of the perianal skin and sometimes the inflamed areas contained bacterial colonies. Immunohistochemical Findings. Immunolabelling (Table 2) demonstrated that bacteria positive for O157 antigen had formed the AE lesions in the ileum of kid C (24 hpi) and in the ileum (Fig. 1), caecum and proximal ascending colon of kid F (72 hpi). At all other sites shown by HE or Giemsa staining to have AE lesions, O157 antigen was not detected, indicating that the lesions were associated with non-O157 bacteria. Antisera against O26, O45, O71, O115 and O145 antigens were used on sections of caecum from kid A (control) in an attempt to serogroup organisms associated with the non-O157 AE lesions. These antisera were selected because of observed AE lesions associated with E. coli O26 and O115 in cattle and sheep (Pearson et al., 1999; Gunning et al., 2001; Cookson et al., 2002a,b; Wales et al., 2002), and because of the isolation of E. coli O45, O71 and O145 with AE potential from an adult goat (Barlow et al., 2004). Positive labelling was not observed with any of these five antisera, and lesions in which both E. coli O157 and non-O157 organisms were co-located were not detected at any site. At no location in which AE lesions were seen was there any relation between the detection of O157 antigen and the degree of mucosal damage. Bacterial colonies in the perianal skin were not associated with O157 antigen. Transmission Electron Microscopy. Bacterial lesions seen in the ileum (Fig. 3) and caecum of kid F had been shown by immunolabelling to be associated with E. coli O157. Such lesions seen in the spiral colon (Fig. 4) of kid A, on the other hand, had not shown labelling with the O157 antiserum. All such lesions had a typical AE appearance. The bacteria were intimately attached to the mucosa with effacement of microvilli, and some cells were degenerate. Some epithelial cells, with attached organisms, appeared to be in the process of extrusion from the mucosa (Fig. 4). In tissue from the ileum of kid C, in which only small, sparse lesions had been observed on histological examination, lesions were not present in the section examined. RESULTS: EIGHT-WEEK-OLD KIDS Clinical findings. All animals in all study groups remained clinically normal throughout the experiment, with no evidence of pyrexia or diarrhoea. Gross pathological changes were not observed at necroscopy in the three animals examined at 24, 48 and 96 hours after oral inoculation. Faecal shedding of E. coli O157:H7. All faeces samples collected from the animals prior to the experimental procedure were negative for E. coli O157 as assessed by immuno-magnetic separation. The oral inoculation doses were determined to be 1 x 1010 cfu for each kid in each of the three groups. The faecal shedding data are shown in figure 5. At 24 hours after challenge, all animals in all groups were shedding the inoculated strain (range 2 x 104 to 9.2 x 106 cfu per gram faeces), there being no statistically significant differences between the geometric mean values of the three groups at this time. On days 2 and 4 after inoculation, differences were observed. On these days all animals in groups A and C dosed with E. coli O157:H7 strain NCTC12900 Nalr and DMB1 (isogenic aflagellate derivative), respectively, shed the inoculated strain at about the same numbers as on day 1 whereas DMB2 (isogenic intimin deficient derivative) was detected in four and three of the six animals in group B on days 2 and 4 after challenge, respectively. The only statistically significant difference in numbers of animals shedding was at day 4 (p = 0.021). On days 2 and 4, the geometric mean shedding scores in groups A and C (~ 10 5 cfu per gram of faeces) were SID 5 (2/05) Page 8 of 18 significantly higher than the geometric mean shedding score for group B animals (~ 5 x 10 1 and 1 x 102 cfu per gram of faeces) (day 2, p = 0.003; day 4, p = 0.014). The geometric mean counts from group B animals were lower than those from the other two groups on days 7 and 8, but these differences were not significant (p = 0.057 and p = 0.062, respectively). From day 10 onwards, only a few animals shed O157 organisms in each group and there were no statistically significant differences between groups or strains. However, it was noted that E. coli O157:H7 strain NCTC12900 Nalr was not detected in faecal samples tested on days 22, 25 and 29 whereas both DMB1 and DMB2 were, albeit in low numbers, on these days. Bacteriological findings at necropsy. Three animals in group A, inoculated with E. coli O157:H7 strain NCTC12900 Nalr, were selected at random for post mortem examination on days 1 (animal 1019), 2 (animal 1001) and 4 (animal 1003). On the day of post mortem examination, faeces were taken and examined for E. coli O157:H7. Animal 1019 had 2 x 104 cfu per gram of faeces and animal 1001 had 2 x 10 4 cfu per gram on day 1 and 7 x 105 cfu per gram on day 2. By contrast, animal 1003 had 9.2 x 10 6 cfu per gram of faeces on day 1, 1.7 x 107 cfu per gram on day 2 and 3.5 x 107 cfu per gram on day 4. There were differences in the numbers of E. coli O157:H7 strain NCTC12900 Nalr recovered from tissues in individual animals (Fig. 6). E. coli O157:H7 was not detected in the small intestine of animal 1019. However, E. coli O157:H7 was detected in the jejunum of animal 1001 and the duodenum, jejunum and ileum of animal 1003. In the large intestine, by contrast, E. coli O157:H7 was detected in most sites in all three animals. In the case of the animal 1003, high numbers of E. coli O157:H7 were found at all intestinal sites examined with up to 106 cfu E. coli O157:H7 per gram of tissue in the mid-section of the rectum. The rumen was negative for all three animals. Histopathological findings. Significant changes were confined to the distal colon, rectum and recto-anal junction (RAJ) of animal 1003, examined on day 4. The most prominent feature was the presence of large numbers of cryptosporidium organisms that were identified attached to the mucosa, along with scattered, focal, attaching and effacing lesions associated with adherent bacteria (Fig 7). Cryptosporidia and associated bacteria appeared to be co-located on the mucosal surface. Associated with these lesions, the lamina propria was infiltrated with moderate numbers of mixed lymphoid cells and polymorphonuclear leukocytes. The remaining large intestine samples (ascending colon and spiral colon) were negative for cryptosporidia and attached bacteria. Sections of small intestine examined were less well preserved than the large intestine. The epithelium in the majority of samples was still intact, although some artefactual separation of the underlying lamina propria had occurred. Small numbers of coccidia were identified in the ileum, predominantly in the crypt epithelium of animal 1001, examined post mortem at day 2. Cryptosporidia and attached bacteria were not identified. Abnormalities were not detected in the rumen. Attempts to isolate the cryptosporidia from animal 1003 could not be made as the samples had been disposed of by the time the results from the histological examination were known. Immunohistochemistry. Multifocal, variable sized colonies of E. coli O157:H7 organisms were identified at all sites of the distal colon, rectum and recto-anal junction in animal 1003 (Fig 8). These tended to cover 1 or 2, to several adjacent epithelial cells. In addition, lesions identified in the mucosa at the recto-anal junction were found on both lymphoid-associated epithelium and non-lyphoid-associated epithelium. Transmission electron microscopy. Cryptosporidia and attaching-effacing lesions associated with bacteria were confirmed on the intestinal mucosa (Fig. 9 & 10). Both organisms were frequently co-located on the same cell (Fig. 10), confirming the light microscopical findings. RESULTS: NANNIES WITH SUCKING KIDS AT FOOT Four of the 5 nannies were colonised after oral dosing and faecal shedding thereafter was variable. Only five faecal samples from two animals (figure 11, animal 1003 on day 1 and animal 1007 on days 1, 3, 7 and 11) contained E. coli O157:H7 organisms with greater than 1 x 104 cfu per gram up to and including day 11. One possible route for kid goats to become infected with E. coli O157:H7 is by sucking at their nanny on teats contaminated with faeces. However, the observed cleanliness of the teats and udders may be attributed to sucking. If it is assumed that some faeces containing E. coli O157:H7 organisms had been on the teats, even transiently, it is probable that the kids were challenged. Faeces containing E. coli O157:H7 organisms shed into the environment were a source of infection of the kids also. No kids became positive for E. coli O157:H7 organisms. DISCUSSION Six-day-old kid studies. Conventionally reared 6-day-old kid goats aged 6 days remained clinically normal when inoculated orally with approximately 1010 cfu of a non-verocytotoxigenic E. coli O157:H7 strain (NCTC 12900 Nalr). At 24 to 72 hpi, E. coli O157 organisms were not detected in the duodenum of any animal and low numbers were detected in the jejunum of one animal only; large numbers, however, were detected in the ileum and SID 5 (2/05) Page 9 of 18 proximal large intestine of all animals. These data suggest that the inoculum had passed to the lower intestinal tract within the time scale of the experiments. It is possible that the inoculum had not travelled the entire length of the intestinal tract of animal D, which was examined at an early stage (24 hpi). At many sites, the density of E. coli O157 organisms was greater than 105 cfu g-1, which is the density of organisms required to induce detectable lesions (Dean-Nystrom et al., 1999). Indeed, AE lesions attributable to O157 organisms were located in the ileum, caecum and proximal colon, where large numbers of E. coli O157 organisms were detected. This is the first time that E. coli O157-associated AE lesions have been detected in the goat. AE lesions associated with unidentified non-O157 bacteria were detected in the large intestines of the two inoculated kids that also had E. coli O157-associated lesions, and in one uninoculated control kid. This indicates colonization by non-O157 AE organisms before the experiment. It is possible that pre-existing AE lesions in the more distal parts of the large intestine of kids C and F prevented the formation of lesions in these parts by E. coli O157. There was no evidence of AE lesions containing both E. coli O157 and non-O157 bacteria. Cookson et al. (2002a,b) demonstrated that E. coli O115 had induced AE lesions in the caecum, colon and rectum of 6-week-old sheep dosed orally with E. coli O157:H7, but no O157:H7 lesions were observed. Potential antagonism between AE organisms in the intestinal tract has yet to be adequately explored. Naylor et al. (2003) suggested that, in cattle, adhesion of E. coli O157:H7 to the lymphoid-associated epithelium at the recto-anal junction was a significant factor in the persistence of the organism. In the present study, despite identifying lymphoid tissue at the recto-anal junction of all four inoculated kids, no such association between E. coli O157:H7 and follicle-associated epithelium was seen. Indeed, specific colonization of any tissues around the RAJ by E. coli O157 was not seen, even in the kid (F) that had high numbers of the organism in this location. AE lesions were reported in a diarrhoeic 2-month old goat (Duhamel et al., 1992), and in association with a naturally occurring outbreak of enteritis in one-week-old kids (Drolet et al., 1994). Recently, AE lesions associated with E. coli O145 were recognized in the small and large intestines of an adult goat (Barlow et al., 2004). In the present study, AE lesions were detected in the intestines of three of six kids, including one uninoculated control animal. When present, the lesions were distributed densely enough to be readily detected, and often occurred in several parts of the intestine of an animal. The kids without detectable AE lesions came from the same source as those with lesions and, in two of three cases, were inoculated with E. coli O157 and had intestinal concentrations of this bacterium that were comparable to those in kids with lesions. The present findings suggest, therefore, that at this age there is substantial individual variation in susceptibility to AE lesions, and this susceptibility appears not to be specific to the serogroup of challenge infecting AE organism. This is perhaps not surprising, given the underlying similarity of AE bacteria in terms of the mechanisms of lesion production. In such immunologically naïve individuals, factors conferring relative resistance to AE lesions may include genetic susceptibility and maternally derived immunity. There is no pattern in the present data relating the luminal concentration of E. coli O157 to the formation of AE lesions. Whilst kid D, with generally low bacterial counts, lacked detectable AE lesions, kid E also lacked lesions despite having higher counts, similar to those of kids (C and F) with lesions. The difference in luminal counts of E. coli O157 between the two kids sampled at 24 hpi (C and D) was striking, but in this early period post-inoculation, factors such as alimentary tract motility may be significant. Despite the numerous AE lesions observed, no animal showed clinical evidence of enteric disease. Since three of six animals were affected in the present study, including an uninoculated control kid, it would appear that AE lesions may be relatively common amongst young, healthy goat kids. This is supported by Orden et al. (2003a,b), who reported that 40 % of healthy goats harboured E. coli with AE potential, the strains belonging to 19 different serotypes and often to untypable categories. The O157 AE lesions in the present goat kids were more numerous and larger than those seen in lambs of a similar age, inoculated with E. coli O157:H7 strains producing verocytotoxins, but having similar intestinal concentrations of the organism (Wales et al., 2001b). Furthermore, the O157 AE lesions in the kids were located proximally (ileum to proximal colon), while those in the lambs were present in the caecum, colon and rectum. Although the degree of neutrophilic inflammation in the large intestines of the kids was generally related to the presence of AE lesions, in some sections such inflammation was present in the absence of detectable lesions. However, and perhaps more interesting, a section in which AE lesions were common but in which there was little associated inflammation was also noted. In conclusion, this study demonstrated the attaching-effacing method of adhesion by E. coli O157:H7 in the ileum and large intestine of neonatal goats. There is evidence that an AE capability promotes the persistence of E. coli O157:H7 in cattle and sheep (Cornick et al., 2002; Woodward et al., 2003), and this may also be true of goats. In addition, the present evidence indicates that AE bacteria may commonly adhere intimately to the intestinal mucosa of neonatal goats without causing clinical signs, but there appears to be considerable variation between goats in the susceptibility to such lesions. Six-week-old kid studies: It is established that enterohaemorrhagic E. coli O157:H7 colonises the intestinal tract of cattle and sheep often leading to long term persistent infection (Cray & Moon, 1995; Brown et al., 1997; DeanNystrom, 1999) and that intimin has been identified as a significant bacterial factor in the colonisation of these species (Cornick et al., 2002; Woodward et al., 2003). In this study, faecal shedding was used as an indicator of the extent of intestinal colonisation. Twenty-four hours after oral inoculation the number of organisms, approximately 104 to 105 cfu per gram of faeces, was equivalent for each of the test strains. Strain NCTC12900 SID 5 (2/05) Page 10 of 18 Nalr and the flagella deficient mutant continued to be shed in faeces at about this level by all animals for the following week whereas the shedding of the intimin deficient mutant from day 2 onwards was from fewer animals and with lower numbers of organisms detected. Whilst these data suggest that intimin may play a role in intestinal colonisation in the weaned goat, statistically significant differences in shedding of the intimin deficient mutant were only noted on days 2 and 4 after inoculation, with no significant differences observed between the three test strains thereafter. Indeed, strain NCTC12900 Nalr was eliminated by three weeks whereas the intimin and flagella deficient mutants continued to be shed until the close of the experiment. Therefore, intimin independent mechanisms of colonisation probably exist, for which long polar fimbriae (Jordan et al., 2004) and porcine associated AE lesion (paa) homologues (Batisson et al., 2003) have been cited. E. coli O157:H7 AE lesions were observed in one animal in this study, coincidentally in association with cryptosporidia. Thus, eight-week-old conventionally reared goats are susceptible to colonisation by enterohaemorrhagic E. coli O157:H7 by intimin dependent mechanisms (Nataro & Kaper, 1998). However, whether intimin contributes to colonisation in this model by acting as an adhesin, independent from the mechanisms for AE lesion formation as suggested by Frankel et al. (1996) and Sinclair & O’Brien (2002) is questionable. The data from these experiments do not support a substantive role for H7 flagella in colonisation in this model. Attaching-effacing lesions induced by O157:H7 organisms were not detected in either of the goats examined 1 and 2 days after oral inoculation. It is possible that the number of organisms at the intestinal sites examined may have been insufficient to induce AE lesions since it has been suggested that there may be a minimum density of organisms to trigger AE lesion formation (Dean-Nystrom et al., 1999). In our previous oral inoculation studies, E. coli O157:H7 induced AE lesions that were rare and sparse in neonatal lambs (Wales et al., 2001b), weaned lambs (Woodward et al., 2003) and neonatal kids (see above and Wales et al., 2005). Therefore, the failure to detect AE lesions in these two goats probably relates to the number of tissue sections examined microscopically relative to the overall size of the intestinal tract. Recent studies in cattle reported by Naylor et al. (2003) suggested a preferred site of colonisation and AE lesion formation by E. coli O157:H7 at the recto-anal junction. Sheng et al. (2004) confirmed these findings in cattle but did not cite any data on colonisation of sheep following rectal administration of the inoculum. In the studies reported here, the recto-anal junction and six other sites along the rectum to the distal colon of the goats were examined. Lesions were not detected in this region of the animals examined 1 and 2 days after oral inoculation but were present in the animal examined on day 4 (animal 1003). This animal shed high numbers of E. coli O157:H7, in the region of 106 to 107 cfu per gram faeces, and had approaching 106 organisms per gram of tissue in the mid rectum. This high density of E. coli O157:H7 may have been sufficient to induce the AE lesions. However, the AE lesions were not associated specifically with lymphoid tissue in the RAJ as has been described for cattle (Naylor et al., 2003). The close association of AE lesions with cryptosporidia is an interesting observation. In natural and experimental infections in goats, Cryptosporidium parvum causes severe clinical disease often with high morbidity and mortality (Koudela & Bokova, 1997; Johnson et al., 1999) and severe lesions are induced often in the posterior jejunum and ileum (Koudela & Jiri, 1997). However, asymptomatic carriage of cryptosporidia has been described in adult goats (Noordeen et al., 2002). In the study reported here, the affected goat was asympomatic and yet had cryptosporidia associated with the mucosa of the distal colon, rectum and recto-anal junction, a region of the intestine at which cryptosporidia have been observed in the later stages of infection in calves (Tzipori et al., 1983). The observation of concurrent infections with two or more enteropathogens, including cryptosporidia in association with E. coli, has been made previously in diarrhoeic calves (Moon et al., 1978; Janke et al., 1990; de la Fuente et al., 1999; Gunning et al., 2001). Based on these recorded concurrent infections and data from the experiments reported in this study, a question arises as to the possible significance of the presence of cryptoporidium infection upon colonisation and shedding of E. coli O157:H7. It is possible that either the cryptosporidia were incidental, or that prior infection with cryptosporidia may predispose the host to colonisation by attaching-effacing E. coli. Without further experimentation aimed at addressing this, it is not possible to draw any firm conclusions. Nannies with sucking kids at foot studies. Goats and goat milk are cited as potential sources of E. coli O157:H7 infection of man and the study here showed that 4 of the 5 nannies were colonised and that shedding was variable. This short lived shedding of E. coli O157:H7 suggests antagonistic factors may have prevented colonisation. The age of the animal, the presence of a very mature native gastro-intestinal bacterial flora and immunity through exposure to gut flora that included other E. coli are likely contributing factors. One possible route for kid goats to become infected with E. coli O157:H7 is by sucking at their nanny on teats contaminated with faeces. In this study where we have attempted to establish this ‘near natural’ route of infection, the cleanliness of the teats and udders may be attributed to sucking and, if it is assumed that some faeces containing E. coli O157:H7 organisms had been on the teats, even transiently, it is probable that the kids were challenged. Faeces containing E. coli O157:H7 organisms shed into the environment were a source of infection of the kids also. No kids became positive for E. coli O157:H7 organisms. This finding is in contrast to a study in which a large dose of E. coli O157:H7 organisms (~ 1010 cfu) given as a single bolus to six-day-old kids resulted in colonisation with AE lesion formation in the mucosa of the large intestine (Wales et al 2004). These kids, although of a similar age, had been removed from the nanny two or three days before challenge with feed provided as artificial milk replacer. Thus, it may be argued that sucking kids are refractive to colonisation by SID 5 (2/05) Page 11 of 18 E. coli O157:H7 possibly because the challenge dose was low and the kids had access to milk, that is known to be highly protective (for review, see Kelleher and Lonnerdal 2001). These hypotheses need to be examined further. These data give clues toward limiting the potential hazards to human health caused by the association between goats and E. coli O157:H7 organisms and two possible risk factors have been identified. As adult goats may amplify E. coli O157:H7, contact with contaminated environments or other infected species that may be a source of infection should be reduced or eliminated. Also, as sucking kids are likely to be protected passively, removal from the nanny or from the nanny’s milk should be avoided. However, whilst these suggestions may be made, detailed epidemiological analyses and further infection/transmission studies should be undertaken to assess them. PIG STUDIES Detailed Introduction Infection of humans by enterohaemorrhagic Escherichia coli (EHEC) O157:H7, is associated with diarrhoea that can lead to haemorrhagic colitis and haemolytic uraemic syndrome (Boyce et al., 1995; Karmali et al., 1983; Riley et al., 1983). The reservoir for EHEC O157:H7 is primarily cattle, although carriage in other ruminants such as sheep, goats and deer has been reported (Gansheroff and O'Brien, 2000; Paiba et al., 2003). Occasional isolations have been made from mono-gastric animals including raccoons, rabbits, rats, and wild-birds, such as seagulls and ducks (Wallace et al., 1997; Cizek et al., 1999; Leclercq and Mahillon, 2003). Isolation of EHEC O157:H7 from farmed pigs has been reported from Japan, Europe and North America (Heuvelink et al., 1999; Nakazawa and Akiba, 1999; Johnsen et al., 2001; Eriksson et al., 2003; Feder et al., 2003), although prevalence was low (<3%), and no reported outbreaks in humans have been linked directly to the consumption of porcine derived meat products. However, in Chile high prevalence of EHEC O157:H7 (10.8%) in pigs was reported, whereas prevalence in cattle was lower (2.9%) (Borie et al., 1997). Studies of EHEC O157:H7 colonisation and persistence in gnotobiotic and colostrum deprived neonate piglets, and conventional pre-slaughter weight pigs, have been reported. Oral inoculation of neonate piglets with EHEC O157:H7 induced watery diarrhoea, diffuse colonisation of the caecum and colon and enterocyte effacement (Francis et al., 1986; Tzipori et al., 1986). This model was later used to demonstrate the role of E. coli O157:H7 intimin in colonisation. (Donnenberg et al., 1993; Tzipori et al., 1995). More recent studies have demonstrated that older conventional pigs are a permissive host for EHEC O157:H7 (Booher et al., 2002; Cornick and Helgerson, 2004). Deliberate oral inoculation studies have demonstrated that EHEC O157:H7 can be shed by 12-week-old pigs for 2 months and that the duration of shedding in the faeces was similar to that of ruminants experimentally infected with the same EHEC O157:H7 isolate (Booher et al., 2002). Cornick and Helgerson (2004) have demonstrated that pigs given low doses of 6 x 10 3 – 4 x 104 CFU of EHEC O157:H7 transmitted this pathogen to uninfected pigs of the same age. These findings suggest that EHEC O157:H7 has the potential to transmit at low dose and infect, thereby being sustained within the population to become a possible source of infection to humans. Various studies have demonstrated that the intimin of E. coli O157:H7 is required for persistent infection in cattle and sheep (Cornick et al., 2002; Woodward et al., 2003) although it is not required for persistent infection of poultry (Best et al., 2005). Whilst the intimin of EHEC O157:H7 has been shown essential for attaching effacing colonisation in gnotobiotic and colostrum deprived neonates (Donnenberg et al., 1993; Tzipori et al., 1995; McKee et al., 1995), Jordan and co-workers (2005) reported that EHEC O157:H7 does not require intimin to persist in 12week-old conventional pigs indicating other EHEC O157:H7 factors play a role in persistent infection. In our laboratory we demonstrated that the flagellum of an Stx-negative E. coli O157:H7 isolate contributed to long-term persistent colonisation of SPF chickens (Best et al., 2005) but not in goats and sheep (La Ragione et al., 2005; unpublished data). This observation is the subject of further analysis and, in the study to be reported here, we wished to assess the role of the flagellum and re-assess the role of intimin in persistent colonisation of pre-slaughter age conventional pigs by a previously reported Stx-negative E. coli O157 isolate. Here we present our findings. METHODS Bacterial strains and inocula. Stx-negative E. coli O157:H7 isolate NCTC12900 was obtained from the National Culture Type Collection (Colindale, London). This isolate does not possess stx1 and stx2 and is able to induce actin rearrangements on HEp-2 and bovine primary cell lines. (Dibb-Fuller et al., 2001; Best et al., 2005). As reported previously (La Ragione et al., 2005) a nalidixic acid resistant derivative, designated NCTC12900nal r, was prepared by passage on Luria-Bertani (LB) agar supplemented with nalidixi AE lesions by this nalr isolate has been observed on tissue culture cell lines, the intestinal sheep mucosa and the gastro-intestines of 6-day- and 8-week-old goats (La Ragione et al., 2005; Wales et al., 2005). E. coli O157:H7 NCTC12900nalr mutants unable to elaborate flagella and intimin were constructed as described previously (Best et al., 2005) and have been characterised fully (La Ragione et al., 2005). Briefly, the structural subunit gene of the flagellum, fliC was inactivated by insertion of a streptomycin resistance cassette and was designated DMB1. The intimin gene, eae was inactivated by insertion with a chloramphenicol resistance gene cassette and was designated DMB2. All E. coli O157 isolates were stored in heart infusion broth (HIB) supplemented with 30% glycerol (v/v) at -80oC and were streaked to single colonies on 5% sheep blood agar when required. SID 5 (2/05) Page 12 of 18 Bacterial inocula for tissue culture and IVOC assays were prepared by growth in 20ml and 100ml LB broths, respectively, for 16 h, at 37oC with shaking (225rpm). Bacterial cultures were washed in 0.1M phosphate buffer solution (PBS, pH 7.2) and then resuspended in Dulbecco’s Modified Eagles Medium (DMEM) supplemented with 1% (v/v) non-essential amino acids (Sigma,) and 1% (v/v) L-glutamine (Sigma) to give approximately 1 x 108 CFU/ml immediately prior to inoculation. E. coli O157 inocula for pig experiments were prepared by growth in 100ml LB broth, for 16 h, at 37 oC with shaking (225rpm). Cultures were diluted serially 10-fold in 0.1M PBS (pH7.2) to give the desired bacterial concentration of approximately 1 x 109 or ~ 1 x 1010 CFU/10ml. The inocula (10 ml) were delivered to the pharyx using a conventional dosing gun. Adhesion and invasion assays. Adherence and invasion assays of NCTC12900nalr and aflagellate and intimin mutants were performed on IPI-21 porcine cells (a gift from Phillip Velge, Nouzilly Laboratoire, Pathologie Infectieuse et Immunologie, INRA Tours,37380 Nouzilly, France). IPI-21 cells were cultured in Dulbecco’s Modified Eagles Medium (DMEM, Sigma) supplemented with 10% foetal calf serum (Gibco), 4 mM L-glutamine (Sigma) and 10mg/ml of insulin from bovine pancreas (Sigma). Materials and methods for adhesion and invasion assays were performed as described by Best et al., (2003). Transmission electron microscopy. Transmission electron microscopy (TEM) of E. coli O157 adhering to the cells was performed using methods described previously (La Ragione et al., 2002). Porcine organ culture association assay. In vitro organ culture (IVOC) was adapted from methods described previously for bovine and human IVOC assays (Hicks et al., 1996; Phillips et al., 2000). Approximately 2cm 2 tissue samples were obtained from the gastro-intestines of 14-week-old conventionally reared pre-slaughter weight pigs. Tissue samples included duodenum, jejunum, ileum, ascending colon, spiral colon, caecum, six regions from the rectum (samples were taken approximately 6cm apart starting from the rectal anal junction to the colon and termed rectum 1 to rectum 6, respectively), and the rectal anal junction (RAJ). All tissues were collected immediately after euthanasia and were placed directly into DMEM. All tissue samples were orientated mucosal surface upward by stitching onto sterile medical gauze with a sterile nylon suture) and placed into Duran bottles containing 95ml of DMEM (1% non-essential amino acids and 1% L-glutamine). A bacterial inoculum of 5ml was added to each Duran for each E. coli O157 test isolate to give a final concentration of approximately 1 x 108 CFU/ml. Duran bottles, with loosely fitting lids, were then incubated at 37 oC (5% CO2) for 6 h, with the culture medium being completely replaced every 2 h. After incubation, each tissue sample was washed three times with HBSS. Each tissue sample was either disrupted for 10 min by using a solution of 1% Triton X-100 (Sigma) and vigorous shaking, or placed into 10% neutral buffered formalin at room temperature for 24 h, for histological analysis. After disruption, serial 10-fold dilutions were plated onto LB agar and incubated overnight to determine the number of CFU/ml. Porcine in vivo studies. Two porcine in vivo studies were performed. In experiment 1, eighteen, 14-week-old conventionally reared pre-slaughter weight pigs were housed indoors with food and water ad libitum, and separated into three groups of 6, and allowed to acclimatise to their environment for one week. All pigs in each of the three groups were challenged (1 x 109 CFU/10ml) with either NCTC12900nalr, DMB1 or DMB2. One pig from each challenge group was taken for post mortem examination on days 1 and 3 post inoculation. Rectal faecal samples were taken from all animals on days 1, 3, 7, 11, 15, 18, 22, 25, 29, 31 and 37 post inoculation. In experiment 2, twenty-four 14-week-old conventionally reared pre-slaughter weight pigs were housed indoors and separated into four groups of 6, and allowed to acclimatise to their environment for one week. Three groups were given 20ml of 10% sodium bicarbonate (NaHCO 3) approximately 30 min before E. coli O157 challenge. These three groups were then challenged with 1 x 10 10 CFU/10ml of NCTC12900nalr, DMB1 or DMB2 in separate experiments. The remaining fourth group was not given NaHCO 3 but was inoculated with 1 x 1010 CFU/10ml of NCTC12900nalr. The rectal faecal samples were taken from all animals on days 1, 3, 8, 11, 15, 18, 22, 25, 29, 31 and 37 after inoculation. From experiment 1 and 2, all E. coli O157 isolates were detected using methods described previously (Stevens et al., 2004; La Ragione et al., 2005). Briefly, pig faeces were weighed and re-suspended in 9ml of BPW, mashed with sterile forceps and vortexed. Ten-fold serial dilutions were plated directly onto SMAC plates supplemented with nalidixic acid (15 dilutions were enriched by incubation at 37oC for 6 h and then plated onto SMAC plates supplemented with ssected with a sterile scalpel to give a 1g sample that was re-suspended in 9ml of BPW by vortexing and bacterial counts determined as described above. Histological examination of IVOC and post mortem tissues. After either IVOC or post mortem examination, each tissue sample for histopathological examination was placed immediately in 10% neutral buffered formalin and fixed for at least 24 h at room temperature. Tissues were processed to paraffin wax, and three to five tissue blocks were prepared from each tissue sample. Multiple 4m sections, comprising several levels from each block, were stained with Haematoxylin and Eosin (H & E) and examined by light microscopy. Sections to be immunostained were prepared as described previously (Best et al., 2003; La Ragione et al., 2005). Immuno-staining was SID 5 (2/05) Page 13 of 18 performed on sections taken from the ascending and spiral colon, the caecum and the rectal anal junction from each pig removed for post mortem examination. For IVOC, only immuno-staining was performed on tissues experimental infected with NCTC12900nalr. Statistical analysis. For statistical analyses of adhesion and invasion assays, counts were transformed to log10 values and analyses of variance (ANOVA) performed, followed by the Tukey’s HSD-test. Pig IVOC statistical analysis on bacterial counts were transformed to the log10 and for each organ the isolates compared by a oneway analysis of variance. DMB1 and DMB2 were compared to NCTC12900nal r by t-tests when the overall p value was significant (<0.05). Statistical analysis of rectal faecal sampling counts were compared on each occasion by the non-parametric Kruskal-Wallis test. When this test was significant (p<0.05), DMB1 and DMB2 were compared to NCTC12900nalr using a test based on the t-distribution. Using the same statistical method, DMB1 and DMB2 were also compared, as was NCTC12900nalr, with and without NaHCO3. RESULTS Interaction of E. coli O157 with porcine (IPI-21) cultured cells. NCTC12900 wild-type, NCTC12900nalr (Fig. 1) and DMB1 (aflagellate mutant) formed AE lesions on porcine IP1-21 cell line whereas DMB2 (intimin mutant) did not. NCTC12900nalr. It was noted that the tissue samples were of a poor quality and unequivocal evidence of intimate attachment was not gained, although loosely associated specifically stained E. coli O157 bacteria were observed (data not shown). Therefore, no IHC was performed on tissues experimentally infected with DMB1 and DMB2. E. coli O157:H7 colonisation and persistence in 3-month-old pigs. Two parallel experiments were performed in which pigs were inoculated with E. coli O157 test organisms resuspended in PBS and then given orally either without (experiment 1) or with a NaHCO 3 pre-dose (experiment 2). Experiment 1 (post mortem analysis); The recovery of E. coli O157 from tissues from pigs examined post mortem, 1 and 3 days after inoculation is shown in Table 1. Although the number of animals examined post mortem was limited, a trend was shown that E. coli O157 were not recovered from the duodenum or jejunum and that only one of six ileal tissue samples was positive for E. coli O157. However, E. coli O157, irrespective of which of the three strains were detected, were readily recovered from tissue samples from the large intestines although on day 3 after inoculation the intimin deficient mutant (DMB2) was only recovered from the spiral colon and the caecum. Immuno-Histo-Chemistry (IHC) analysis showed E. coli O157 stained bacteria intimately associated with the mucosa of the ascending and spiral colon and the caecum (Fig.4) of pigs inoculated with NCTC12900nalr, collected on day 3 post inoculation, but none were observed on tissue samples taken on day 1 after inoculation. Similarly, intimate attachment was noted in the ascending and spiral colon and caecal tissues of pigs inoculated with the flagella deficient mutant on day 3, but not day 1 after inoculation. However, it should be noted that AE lesion formation cannot be observed at this magnification. Transmission electron microscopy (TEM) was not done the intimate attachment was seen to be small foci and unlikely to yield AE lesions by TEM from adjacent samples from the same block. Specifically stained E. coli O157 bacterial attachment was noted for the intimin mutant on days 1 and 3 after inoculation, but only for the spiral colon and caecum. Neither NCTC12900nalr or DMB1 and DMB2 were observed attached to the rectal anal junction. Experiment 1; (bacterial shedding). The bacterial-take for 14-week-old pigs inoculated with either NCTC12900nalr, DMB1 or DMB2 was low as determined by assessment of faecal load of E. coli O157 organisms. For the dose given (1 x 109 CFU), it was expected that c. 1 x 104 CFU/g of faeces would be attained, but in this experiment the numbers of E. coli O157 organisms in faeces was low, just above the limits of detection. However, shedding of the three test strains was sporadic for the duration of the experiment, although the last day of detectable shedding was day 31, 22 and 37 for NCTC12900nal r, DMB1 and DMB2, respectively (data not shown). Experiment 2; (bacterial shedding) (Fig. 5). The bacterial-take was far greater than for experiment 1. Statistical analysis did not reveal any significant differences between shedding of NCTC12900nal r and each mutant on any day after inoculation. This was also noted for statistical comparisons made between DMB1 and DMB2. However, for NCTC12900nalr with and without NaHCO3, significantly more NCTC12900nalr was recovered from pigs pre-dosed with NaHCO3 on day 1 after inoculation (p = 0.05?). For all groups pre-dosed with NaHCO3, high numbers of bacteria (106 to 107 CFU/g/faeces) were recovered for each E. coli O157 group 24h after inoculation. NCTC12900nalr, DMB1 and DMB2 continued to be shed for up to day 22 after inoculation, although during this time, only DMB2 was recovered at each time point. Although not statistically significant, more pigs were positive for DMB2 than DMB1 on days 15, 18 and 22 post inoculation. DISCUSSION Porcine cell culture (IPI-21) adherence assays showed that, whilst NCTC12900nalr and the aflagellate mutant (DMB1) adhered intimately and formed AE lesions, whereas the intimin deficient mutant (DMB2) did not, the numbers of each E. coli O157 test isolate adhering was not significantly different. This finding contrasted with a previous tissue culture adherence assay using HEp-2 cells, in which significantly fewer intimin deficient mutant adhered compared to NCTC1200nalr and the aflagellate mutant (Best et al., 2005). These data reaffirm the view SID 5 (2/05) Page 14 of 18 that the host cell influences adherence and that intimin, and presumably LEE related factors, are not the only mechanism for adherence to IPI-21 cells. Also, it has has been noted previously for Stx-negative E. coli O157:H7 association with bovine primary cell lines (Dibb-Fuller et al., 2001), NCTC12900nalr and derivatives DMB1 and DMB2 bound porcine epithelial IPI-21 cells more efficiently than either E. coli K12 or S. Typhimurium. This may indicate a greater avidity of the interaction between E. coli O157 and IPI-21 cells, possibly due to specific host receptors and bacterial ligands. As reported previously, low-level invasion of IPI-21 by E. coli O157 was noted in this study. Whether this occurs passively or by an active mechanism has been discussed previously (McKee et al., 1995; Dibb-Fuller et al., 2001; Best et al., 2003; Best et al., 2005) and has yet to be established. Porcine IVOC adherence assays were performed to test for tissue specific tropism, Overall, the trend for the aflagellate mutant and the intimin mutant was to adhere in significantly higher numbers than NCTC12900nalr to seven of the thirteen porcine tissues tested, most notably at tissues from the small intestine and the proximal rectum. Although poor tissue quality was noted throughout after histological analysis, immunohistochemistry did reveal the presence of specifically stained E. coli O157 associated loosely with the pig IVOC mucosa. This suggests that Stx-negative E. coli O157 has the potential to associate with the porcine mucosa using adherence mechanisms other than the flagellum and intimin. The apparent differences in tropism of NCTC12900nal r and the aflagellate and intimin deficient mutants may indicate that the presence of the flagellum and intimin enhance tropism to specific sites whereas their absence enhances promiscuity in terms of adherence to tissues. Two experimental infection studies with conventionally reared 14-week-old pigs were performed and one objective of the first experiment was to determine whether the tropism observed by IVOC would be similar in vivo. A dose similar to that described by Cornick et al. (2004) for a Stx-positive E. coli O157:H7 isolate was given to each pig, but the take was poor by comparison to the strain used by Cornick and co-workers (2004). Therefore, we did not use any lower challenge doses. Nevertheless, post mortem data revealed a different colonisation pattern to IVOC data. Neither the duodenum nor jejunum and only one of six ileal tissues were colonised by E. coli O157 bacteria. Recovery of E. coli O157 bacteria was greatest from the large intestine. Interestingly, lymphoid follicles are found in the colon and caecum of pigs (Morfitt and Pohlenz, 1989; Kruiningen, 1995). The bacterial enumeration data from our study demonstrates that generally, more E. coli O157:H7 were recovered from these tissues. Interestingly on day 3 after inoculation, the intimin mutant was only recovered from the spiral colon and caecum. This is consistent with previously reported data (Jordan et al., 2005) and may suggest that another colonisation factor with a tropism to lymphoid tissue was expressed and that intimin may still have a lowlevel role to play in other porcine tissues of importance. The rectum was colonised and, combined with the IVOC data, may suggest this tissue may be a preferential site of colonisation as has been noted for other E. coli O157:H7 isolates in bovine models of infection (Naylor et al., 2003; Sheng et al., 2004; Low et al., 2005). Histological analysis did reveal the presence of E. coli O157 attachment of all test isolates to pig mucosa, especially the spiral colon which is consistent with previously published data (Jordan et al., 2005). Since attachment of the aflagellate and intimin deficient mutants were noted this suggests that other mechanisms of attachment may have been involved. The absence of E. coli O157 organisms from tissues from the small intestine is a finding contrary to expectation from the IVOC studies in which the aflagellate and intimin deficient mutants showed an apparent tropism. This suggests that either the IVOC data may be artefactual, although the statistical analysis strongly suggests otherwise, or that passage of E. coli O157 organisms through the stomach was associated with suppression of expression of those E. coli O157 factors required for adherence to the small intestine. For in vivo experiment 2, the administration of sodium bicarbonate (NaHCO 3) before oral inoculation with E. coli O157 did demonstrate a better take of each test isolate, and may be in part due to an increase in the pH of stomach. However, a previous study did report that NaHCO3 stimulated the expression of E. coli O157:H7 LEEencoded genes, albeit in vitro, where the production of intimin, Tir, EspA and EspB was greatly enhanced (Abe et al., 2002). Since the intimin deficient mutant reported in our study was shed in numbers similar to the parent isolate, this might suggest, although highly speculative, that NaHCO 3 may have stimulated other non-LEEdependent genes also. Previous data from our laboratory demonstrated that the flagellum of E. coli O157, but not intimin, was important for persistence in poultry (Best et al., 2005), whereas intimin, but not the flagellum, was important in conventionally weaned lambs (Woodward et al., 2003; unpublished data). This suggests that the function of both surface arrayed structures may be host dependent and may be linked to the availability of specific host-cell receptors. Intimin is thought to interact not only with its own type III secreted receptor, Tir (Campellone and Leong, 2003; Deng et al., 2003), but also mammalian nucleolin for example (Sinclair and O'Brien, 2002; 2004). Whether E. coli O157 intimin can interact with pig nucleolin has yet to be described, but if it cannot, this may explain in part, why intimin was not required for persistence. The flagellum subunit protein flagellin is known to bind with the Toll-Like-Receptor, TLR-5, which triggers proinflammatory and adaptive immune responses (Ramos et al., 2004). Furthermore, the E. coli O157:H7 flagellum has been reported to induce the systemic immune system in both rabbits and humans (Sherman et al., 1988). Therefore, it is possible to hypothesize that an E. coli O157 aflagellate mutant may be able to persist for a time period similar to that of the parent strain, due to decreased stimulation of the immune system. However, collectively the shedding data did demonstrate that the aflagellate mutant DMB1 was recovered less from pig faeces than the wild type and the intimin mutant Whilst not statistically significant, this may represent a trend. However, other reported E. coli O157:H7 virulence factors, such as long polar fimbriae, do contribute to persistence in large animal infection models including the pig model, SID 5 (2/05) Page 15 of 18 and may have contributed to the persistent infection of pigs noted for the intimin- and flagella-deficient mutants reported here (Jordan et al., 2004). The present study conducted here is the first to analyse the role of the E. coli O157:H7 flagellum in commercially-reared pigs. The study also confirmed previous findings (Jordan et al., 2005) of a Stx-positive E. coli O157:H7 isolate that intimin is not required for persistence in pigs. The fact that the flagellum and intimin had little or no role to play in colonisation of 14-week-old conventionally reared pigs which suggests that other factors are required for early colonisation and long-term persistence. These data also suggests that the presence of NaHCO3 may stimulate other EHEC O157:H7 virulence factors important for the early colonisation of pigs. Whether Shiga-toxin has a role to play in early colonisation in pigs remains unknown, although one study has demonstrated that intestinal epithelium is an important determinant for shiga-toxin interaction and this has major implications for the differential consequences of E. coli O157:H7 infection in reservoir hosts and humans (Hoey et al., 2003). Collectively, these data confirm previously reported data that pigs are a potential reservoir for E. coli O157:H7 and a risk to human health. However, the mechanisms involved in persistent infection of this host appear to be not entirely dependent of the flagellum or intimin. SID 5 (2/05) Page 16 of 18 References to published material 9. This section should be used to record links (hypertext links where possible) or references to other published material generated by, or relating to this project. R. M. La Ragione, N. M. Y. Ahmed, A. Best, D. Clifford, U. Weyer, W. A. Cooley, l. Johnson, G. R. Pearson and M. J. Woodward. (2005) Colonisation of eight-week-old conventionally-reared goats by Escherichia coli O157:H7 after oral inoculation. Journal of Medical Microbiology 54, 485-492. A. D. Wales, G. R. Pearson, J. M. Roe, C. M. Hayes, R. M. La Ragione and M. J. Woodward. (2005) Attaching-effacing Lesions Associated with Escherichia coli O157:H7 and Other Bacteria in Experimentally Infected Conventional Neonatal Goats. Journal of Comparative Pathology. 132, 185-194. A. D. Wales, M. J. Woodward and G. R. Pearson. 2005. Attaching-effacing bacteria in animals. Journal of Comparative Pathology. 132, 1-26. R. M. La Ragione, N. M. Y. Ahmed, A. BEST, D. Clifford, U. Weyer, and M. J. Woodward. (2005) Failure to detect transmission of Escherichia coli O157:H7 from orally dosed nannies to sucking kids at foot. Veterinary Record 157. (in press) A. Best, R. M. La Ragione, D. Clifford, N. M. Y. Ahmed, W. A.. Cooley, A. R. Sayers, and M. J. Woodward (2005). A comparison of shiga-toxin negative Escherichia coli O157 aflagellate and intimin deficient mutants in porcine in vitro and in vivo models of infection. Veterinary Microbiology (accepted for publication). SID 5 (2/05) Page 17 of 18 SID 5 (2/05) Page 18 of 18