I. Aktan

Streptococcus bovis biotype I meningoencephalitis in an alpaca ( Lama pacos ) cria

STREPTOCOCCAL infections, in particular the Lancefield Group C Streptococcus equi subspecies zooepidemicus, have been reported in alpacas (Lama pacos) and other South American camelids (Fowler 1998). Group D streptococci such as Streptococcus bovis are less commonly recorded, possibly because they are considered to be opportunistic pathogens (Quinn and others 1994). S bovis bacteraemia has been described in human beings and is associated with gastrointestinal disease, particularly colonic neoplasia, as well as extracolonic malignancy (Gold and others 2004); S bovis meningitis has also been described (Grant and others 2000). Diseases associated with S bovis reported in other species include lactic acidosis or ‘grain engorgement’ of ruminants (Russell and Hino 1985, Gill and others 2000), septicaemia and meningitis in calves (Seimiya and others 1992), endocarditis in mink (Pedersen and others 2003) and septicaemia in pigeons (De Herdt and others 1992). This short communication describes a fatal case of infection with S bovis biotype I in an alpaca cria in the UK. A 10-day-old, 8·5 kg, male, homebred alpaca was found dead without premonitory clinical signs. This cria had been born in an uneventful outdoor parturition and its colostrum intake was considered to have been adequate. It had shared a paddock with 25 other alpacas, including five crias, which had unrestricted access to a shed where they were fed once daily. Two of the other crias had previously had diarrhoea, but no aetiologic diagnosis was made at the time. A total of 150 alpacas of various ages, from different sources, were present on the farm. Postmortem examination was performed the day after death and the carcase was found to be autolysed. The stomach compartments contained a mixture of fibrous material and semiclotted milk. The large intestinal contents were liquid but the rectum was empty. Other organ systems were grossly normal. Samples of brain, heart, lung, liver, kidney and small intestine were fixed in 10 per cent neutral-buffered formalin and processed to paraffin wax, and sections of 4 μm thickness were stained with haematoxylin and eosin. Sections of the brain were also stained using Gram’s technique. Aerobic culture on sheep blood agar of a meningeal swab yielded a pure growth of a Group D streptococcus. Biochemical testing using the API 20 Strep system (bioMerieux) identified the isolate as S bovis biotype I (Table 1). Aerobic culture of the large intestinal contents yielded a mixed bacterial growth, including streptococci, but exact identification was not undertaken. Clostridium perfringens exotoxins were not identified by ELISA testing of small intestinal filtrate. Rotavirus antigen was detected by polyacrylamide gel electrophoresis (PAGE) testing of large intestinal contents. Neither cryptosporidia nor coccidia were identified on parasitological examination. On histopathological examination of the brain (cerebrum, cerebellum, pons and medulla), diffuse meningeal infiltration with predominantly neutrophils and macrophages, interFIG 1: Diffuse purulent meningitis in a sulcus, with extension into the superficial cortex and focal perivascular cuffing, in the cerebrum of an alpaca. Haematoxylin and eosin. x 100

Equine platelets inhibit E. coli growth and can be activated by bacterial lipopolysaccharide and lipoteichoic acid although superoxide anion production does not occur and platelet activation is not associated with enhanced production by neutrophils.

Activated platelets can contribute to host defense through release of products with bactericidal actions such as antimicrobial peptides and reactive oxygen species (ROS), as well as by forming heterotypic aggregates with neutrophils and enhancing their antimicrobial properties. Whilst release of vasoactive mediators from equine platelets in response to stimuli including bacterial lipopolysaccharide (LPS) has been documented, neither ROS production, nor the effects of activated platelets on equine neutrophil ROS production, have been reported. This study first sought evidence that activated equine platelets inhibit bacterial growth. Platelet superoxide production in response to stimuli including Escherichia coli-derived LPS and lipoteichoic acid (LTA) from Staphylococcus aureus was then determined. The ability of LPS and LTA to up-regulate platelet P-selectin expression and induce platelet-neutrophil aggregate formation was investigated and the effect of co-incubating activated platelets with neutrophils on superoxide production measured. Growth of E. coli was inhibited in a time-dependent manner, and to a similar extent, by addition of platelet rich plasma (PRP) or platelet poor plasma (PPP) obtained by centrifugation of PRP. Activation of platelets in PRP by addition of thrombin led to a significant increase in the inhibitory action between 0.5 and 2h. Although phorbol myristate acetate (PMA) caused superoxide production by equine platelets in a protein kinase C-dependent manner, thrombin, platelet activating factor (PAF), LPS, LTA and formyl-methionyl-leucyl phenylalanine (FMLP) were without effect. LPS and LTA did induce platelet activation, measured as an increase in P-selectin expression (% positive cells: 17±3 (un-stimulated); 63±6 (1μg/ml LPS); 64±6 (1μg/ml LTA); n=5) but not platelet superoxide production or heterotypic aggregate formation. Co-incubation of activated platelets with neutrophils did not increase neutrophil superoxide production. This study has demonstrated for the first time that when activated, equine platelets, like those of other species, are capable of releasing ROS that could assist in bacterial killing. However, the findings suggest that neither superoxide production by platelets nor enhancement of production by neutrophils is likely to play a significant role. Nevertheless, as has been reported in man, equine PPP and PRP did inhibit E. coli growth in vitro, and addition of thrombin significantly increased the inhibitory effect of PRP. This suggests that products released from activated platelets could contribute to antimicrobial activity in the horse. The factors in equine plasma and released by activated platelets that are responsible for inhibiting bacterial growth have yet to be determined.

Colonization, Persistence, and Tissue Tropism of Escherichia coli O26 in Conventionally Reared Weaned Lambs

ABSTRACT Escherichia coli O26 is recognized as an emerging pathogen associated with disease in both ruminants and humans. Compared to those of E. coli O157:H7, the shedding pattern and location of E. coli O26 in the gastrointestinal tract (GIT) of ruminants are poorly understood. In the studies reported here, an stx-negative E. coli O26 strain of ovine origin was inoculated orally into 6-week-old lambs and the shedding pattern of the O26 strain was monitored by serial bacteriological examination of feces. The location of colonization in the GIT was examined at necropsy at two time points. The numbers of O26 organisms excreted in feces declined from approximately 107 to 104 CFU per gram of feces by day 7 and continued at this level for a further 3 weeks. Beyond day 30, excretion was from few animals, intermittent, and just above the detection limit. By day 38, all fecal samples were negative, but at necropsy, O26 organisms were recovered from the upper GIT, specifically the ileum. However, no attaching-effacing (AE) lesions were observed. To identify the location of E. coli O26 within the GIT early after inoculation, two lambs were examined postmortem, 4 days postinoculation. High numbers of O26 organisms were recovered from all GIT sites examined, and ∼109 CFU were recovered from 1 gram of ileal tissue from one animal. Despite high numbers of O26 organisms, AE lesions were identified on the mucosa of the ascending colon of only one animal. These data indicate that E. coli O26 readily colonizes 6-week-old lambs, but the sparseness of AE lesions suggests that O26 is well adapted to this host, and mechanisms other than those dependent upon intimin may play a role in persistence.

Prevalence of attaching and effacing Escherichia coli serogroup O103 in orphan lambs on an open farm in eastern England

multi plex PCR targeting the Shiga toxin 1 (stx1), Shiga toxin 2 (stx2) and intimin (eae) genes as described previously by La Ragione and others (2002). For isolates identified as eae positive, the specific intimin subtype was determined by PCR as described by Aktan and others (2004), and the serogroup was established by the E coli serotyping section at the Veterinary Laboratories Agency. Of a total of 530 isolates confirmed to be E coli isolates, 61 (11·5 per cent) carried the eae gene. No eae-positive isolate possessed either stx1 or stx2 genes, and of the remaining 469 isolates, two were stx1-positive and two were stx1 and stx2positive. Twenty-one eae-positive isolates were serogrouped; in order of frequency, there were 17 O103K+ strains, two O8: K+ strains, one O118:K+ strain and one Rough strain. All of the eae-positive isolates identified as belonging to serogroup O103:K+ possessed the eaeε (epsilon) intimin type. No other serogroup recognised as a human pathotype was identified. In order to investigate the possible relationships between the E coli O103:K+ isolates obtained during the study, pulsed-field gel electrophoresis (PFGE) fingerprints of 10 isolates selected at random were determined. Each isolate was analysed by PFGE as previously described by Liebana and others (2001). There was no difference in the Xba1 (Promega) PFGE patterns for these isolates. The clinical significance of E coli O103:K+ for the orphan lambs is unknown, although from a veterinary public health perspective it has potential zoonotic significance. A recent survey of the prevalence of non-O157 E coli on Scottish farms revealed a mean prevalence of 2·7 per cent for serogroup O103 (Pearce and others 2006) in cattle. Those authors stated that the carriage of stx was rare in the serogroup O103 isolates found in their study, and the absence of stx concurs with the present findings. The study of Pearce and others (2006) involved surveillance of 6086 samples from 338 farms across Scotland. While no equivalent survey data exist for England, the biased sample studied here and previous small surveys (Liebana and others 2001, Aktan and others 2004) indicate that E coli serogroup O103 may also be prevalent in sheep. The present study showed that the E coli O103 isolates shared an identical PFGE profile, indicating clonality. Without more detailed investigation no conclusions can be drawn about the source of the organism, but it can be speculated that either a common clonal line is widespread and infected lambs arrived on-farm, or the clone is resident on-farm and the orphan lambs are readily colonised upon exposure on the farm. What is clear is that the orphan lambs studied were colonised with and shedding very high numbers of this organism, as proven by direct detection by plating without any selection. The environmental burden arising from this level of shedding would be of greater veterinary public health concern if the isolates possessed Shiga toxins. A previous small survey conducted on animals at slaughter in the UK revealed that E coli serotype O103 lacking Shiga toxins comprised 0·6 per cent of the total number of isolates tested, and 8 per cent of the various attaching and effacing E coli detected in cattle, sheep and pigs (Aktan and others 2004). It has been suggested that E coli O103 lacking Shiga toxins from rabbits and E coli O103 possessing Shiga toxins from human beings have a common origin but different mechanisms of adaptation to their hosts (Mariani-Kurkdjian and others 1993). More recent data suggest that, while there is genetic diversity within serogroup O103 isolates, probably mediated through acquisition of different virulence genes and DNA rearrangements, certain subtypes (for example, multilocus sequence typing [MLST] profile II) from human beings and animals were very similar except for the presence or absence of the Shiga toxins (Beutin and others 2005). The authors also demonstrated that serogroup O103 isolates possessing the eaeε intimin type, as found in the present study, belong to the MLST profiles I and III, which elaborate Shiga Veterinary Record (2007) 161, 386-388

Escherichia coli O115 forms fewer attaching and effacing lesions in the ovine colon in the presence of E. coli O157:H7.

Escherichia coli O115 has been isolated from healthy sheep and was shown to be associated with attaching-effacing (AE) lesions in the large intestine. Following previous observations of interactions between E. coli O157 and O26, the aim of the present study was to assess what influence an O115 AE E. coli (AEEC) would have on E. coli O157 colonisation in vitro and in vivo. We report that E. coli O115- and O157-associated AE lesions were observed on HEp-2 cells and on the mucosa of ligated ovine spiral colon. In single strain inoculum, E. coli O115 associated intimately with HEp-2 cells and the spiral colon in greater numbers than E. coli O157:H7. However, in mixed inoculum studies, the number of E. coli O115 AE lesions was significantly reduced suggesting negative interference by E. coli O157. Use of the ligated colon model in the present work has allowed in vitro observations to be extended and confirmed whilst using a minimum of experimental animals. The findings support a hypothesis that some AEEC can inhibit adhesion of other AEEC in vivo. The mechanisms involved may prove to be of utility in the control of AE pathovars.

PKC isoenzymes in equine platelets and stimulus induced activation.

Protein kinase C (PKC) is an important regulator of platelet activation and different isoenzymes can play positive and negative regulatory roles. The PKC isoenzymes expressed in equine platelets have not been documented but pharmacological inhibition has suggested a role for PKC delta (δ) in modulating responsiveness to platelet activating factor (PAF) (Brooks et al., 2009). Here the PKC isoenzyme profile in equine platelets has been characterised and PKCδ activation by PAF investigated. Platelet lysates were probed by Western blotting using a panel of antibodies against individual PKC isoenzymes. PKCδ and eight other isoenzymes were identified, namely classical PKCs alpha (α), beta (β), (both βI and βII) and gamma (γ), the novel PKCs epsilon (ɛ), eta (η) and theta (θ) and atypical PKC zeta (ζ). Having shown PKCδ to be present, a method was developed to measure PAF-induced isoenzyme translocation by preparing cytosolic and membrane fractions from digitonin permeabilised platelets. Phorbol 12-myristate 13-acetate (PMA) was shown to cause translocation of PKCδ to the membrane within 5s. PAF also caused PKCδ translocation although the response occurred more slowly; a significant, 7.6 ± 1.2 fold, increase in band density compared to unstimulated platelets was observed at 15 min; p=0.036, n=3. These data support a role for PKCδ in regulating PAF-induced functional responses in equine platelets.

Interaction between attaching-effacing Escherichia coli O26:K60 and O157:H7 in young lambs

Recent surveys have shown that Escherichia coli O26 is prevalent in ruminants compared with E. coli O157. These serogroups share common colonisation factors and we hypothesised that prior colonisation by E. coli O26 may show reduced colonisation by E. coli O157. To test this hypothesis, strains of E. coli O26:K60 and O157:H7 were tested in competitive in vitro and in vivo studies. Using an established 6-week-old lamb model, an experimental group of lambs was dosed orally with E. coli O26:K60 and then E. coli O157:H7 four days later. The faecal shedding of O26:K60 and O157:H7 organisms from this experimental group was compared with that from animals dosed with either O26:K60 alone or O157:H7 alone. Shedding data indicated that counts for O157:H7 were unaffected by the competition from O26:K60, whereas the O26:K60 counts were lower when competing with O157:H7.