Development of a PCR test based on a gene region associated with the pathogenicity of Pasteurella multocida serotype B:2, the causal agent of haemorrhagic septicaemia in Asia.

We have developed a PCR assay to detect Pasteurella multocida serotype B:2, the causal agent of Haemorrhagic Septicaemia (HS) in Asia. Nucleotide sequence determination of a 16S rRNA-23S rRNA PCR product unique to B:2 strains was shown to share amino acid sequence homology with a bacteriophage Mu protein. Primers designed from this sequence when tested against a panel of isolates recovered from a wide geographical area and representing a large range of bacterial genera and species, were found to specifically amplify DNA from P. multocida, serotype B:2. Southern hybridisation confirmed the presence of this sequence in only the B:2 serotype of P. multocida, suggesting an association between bacterial virulence and the presence of bacteriophage genes in the bacterial genome. The results of this study demonstrate the potential application of PCR to the diagnosis of HS in cattle and buffalo in Asia. Application of PCR to support diagnosis of HS will greatly improve accuracy, laboratory response time, and will facilitate rational deployment of resources for controlling this disease.

The specificity of antibody in chickens immunised to reduce intestinal colonisation with Campylobacter jejuni.

Poultry consumption has been identified as a major risk factor for human infection with Campylobacter jejuni in developed countries. C. jejuni is present in the gastrointestinal tract of broiler chickens at the time of slaughter, and faecal contamination of carcases during processing results in significant campylobacter loads on carcases. One approach to reducing the level of carcase contamination with C. jejuni is to control campylobacter infection in broiler chickens. To this end, the study described here investigated the specificity of antibody in serum and intestinal secretions of chickens that had been immunised with campylobacter antigens and then challenged with viable bacteria. The immunodominant antigens in the serum of birds that showed a 2-log reduction in caecal colonisation with C. jejuni included flagellin protein (61-63 Kd) and three additional antigens of 67, 73.5 and 77.5 Kd. Only flagellin and the 67 Kd antigen were recognised by IgG antibody in gastrointestinal secretions of the same birds. Antibody from chickens immunised with purified native flagellin protein recognised flagellin protein and the 67 Kd antigen in Western blots probed with serum, but only the flagellin proteins (61-63 Kd) in Westerns probed with gastrointestinal secretions. Analysis of the specificity of the response to flagellin protein using recombinant clones that expressed regions of the flagellin gene suggests that epitopes in each region of the flagellin protein were immunogenic. Of the immunodominant antigens, only flagellin appeared to be surface-exposed on viable C. jejuni, although conformational epitopes of flagellin appeared to be sensitive to the method of antigen purification. The results of this study suggest that flagellin and possibly the 67 Kd antigen may be valuable for immunological control of intestinal infection with C. jejuni in chickens, but that further work is required to purify these as vaccine candidates by using methods that preserve conformational epitopes.

Genotypic Diversity among Campylobacter jejuni Isolates in a Commercial Broiler Flock.

Analysis of nucleic acid polymorphism in the flagellin genes of Campylobacter jejuni was used to investigate genetic diversity among Campylobacter spp. in a commercial broiler flock. Three hundred single colonies of C. jejuni were isolated from fecal samples collected weekly for 3 weeks immediately before slaughter. Both the flaA and flaB genes were amplified by PCR, and the PCR product was digested with the restriction enzyme AluI. The fragments generated were then analyzed by agarose gel electrophoresis. Among the 300 recovered isolates, five different restriction fragment length polymorphism profiles were observed. Three of these profiles were dominant during the course of the study, and the other two profiles were detected at low frequency. Analysis of genetic variation in C. jejuni over the course of an experimental infection lasting 7 weeks indicated that there was no obvious drift in the flagellin gene type. These findings demonstrate that a range of bacterial genotypes can constitute the bacterial population within a commercial poultry flock, with the most likely sources of these types being multiple environmental exposure and/or genetic drift within the population. This degree of diversity must be considered in epidemiological analyses which utilize genetic typing methods that investigate Campylobacter contamination of any food source, including poultry, to ensure that the total gene pool for C. jejuni is evaluated.

Immunisation of chickens to reduce intestinal colonisation with Campylobacter jejuni.

1. Systemic and intestinal antibody titres were measured in chickens following subcutaneous, intraperitoneal (i.p.), oral (p.o.) and combined i.p./p.o. administration of antigen, in soluble, emulsified or microparticulate form. Antigens tested included keyhole limpet haemocyanin (KLH), killed Campylobacter jejuni whole cells and purified campylobacter flagellin protein. 2. The effect of immunisation with purified flagellin protein or with killed C. jejuni whole cells in reducing intestinal colonisation was assessed. The ability of newlyhatched chicks to respond to immunisation was limited, possibly because of the immaturity of the immune system rather than maternal suppression of an immune response. Only 5 to 13 birds that were first immunised when 1-d-old with KLH showed a systemic response, even after 4 immunisations, whereas 10 of 11 birds that were first immunised at 24 d-old responded systemically. 3. In an immunisation and challenge experiment, birds that were immunised twice intraperitoneally, at 16 and 29 d-old, with killed C. jejuni whole cells, had fewer C. jejuni, in the caecal contents than unimmunised control birds. This reduction in intestinal colonisation, to less than 2% of bacterial numbers in control birds, was associated with an increase in specific IgG in intestinal secretions. There was no significant increase in specific IgA or IgM in intestinal secretions following immunisation and challenge. 4. These results indicate that immunisation can reduce the level of intestinal infection with C. jejuni. The protection may be enhanced by developing improved methods of immunisation that stimulate production of increased titres of specific antibody in intestinal secretions, particularly specific IgA antibody.

In ovo oral vaccination with Campylobacter jejuni establishes early development of intestinal immunity in chickens.

1. Chick embryos were orally immunised at day 16 of incubation by injection of heat-killed Campylobacter jejuni organisms into the amniotic fluid. The response to vaccination was observed at 5 d after hatching or, in some birds which received a postnatal oral booster vaccination, at 7 d after hatching, and the response was observed at 14 d of age. 2. The titres of antibody in serum, bile and intestinal scrapings, the distribution of immunoglobulin-containing cells in the spleen, duodenum and ileum and the expression on peripheral blood leukocytes (PBL) of the T cell surface markers CD3, CD4 and CD8 were determined. 3. Whereas low titres of anti-flagellin antibody were detected in serum, bile and intestinal scrapings of unimmunised birds, high titres were observed in immunised birds. 4. An increase in antibody of all isotypes was detectable in serum but the elevation in IgA antibody in intestinal scrapings and bile was particularly striking. This response was reflected in a dramatic increase in immunoglobulin-containing cells, detected by fluorescent histology, particularly those associated with IgA and IgM isotypes in the spleen and intestine of immunised birds. 5. Secondary oral boosting after hatching resulted in a depression in serum anti-flagellin antibody in immunised birds compared to pre-boosting titres (although still significantly higher than in non-immunised controls) but an increase in IgA antibody in intestinal scrapings and bile. The number of immunoglobulin-containing cells was also increased after boosting. 6. Neither immunisation regimen caused a significant change in the numbers of circulating CD3, CD4 or CD8 T cells. 7. These results indicate that in ovo oral immunisation with C. jejuni antigens stimulates the precocious development of immunity in chicks.

Controlling microbial contamination on beef and lamb meat during processing.

The microbiological quality of carcases, meat and environmental surfaces was evaluated in commercial boning rooms processing beef and lamb. There was considerable variation in the level of microbial contamination on both carcases and meat, with counts ranging from less than 20 to 10(8)/cm2 on carcases and to 2 x 10(7)/cm2 on meat. The level of microbial contamination on meat was influenced by the level of carcase contamination at boning and by the boning process itself. Carcase contamination was the major determinant of microbiological quality, as more than 70% of carcase had microbial counts greater than 10(3)/cm2. Cutting boards were a major source for microbial dissemination during boning, particularly when carcase counts were less than 10(3)/cm2. If carcases were heavily contaminated, the contamination of processing surfaces was irrelevant in determining microbial loads on meat. Where carcase contamination was at low to moderate levels, the contribution of the boning process to the contamination on meat assumed increased significance. Under these conditions, improved sanitation of cutting surfaces in the boning room resulted in a significant reduction in microbial contamination on the surface of meat. These results can form the basis for ensuring that improvements made in carcase management before boning, to improve microbiological quality, will be preserved through attention to cutting board hygiene during boning.

Immunisation of mares to control endometritis caused by Streptococcus zooepidemicus.

Normal mares were immunised by the intramuscular and intrauterine administration of an antigen with adjuvant and they and unimmunised control mares were later challenged by the intrauterine instillation of pathogenic Streptococcus zooepidemicus; the response of all the mares was monitored clinically and bacteriologically for seven days. Significantly fewer S zooepidemicus were present in cervical swabs taken from the immunised mares than from the control mares (P < 0.01) and the degree of inflammation in the genital tract of the immunised mares was also significantly less (P < 0.001). This protective effect of immunisation was associated with the specific IgG response in the serum, and an IgG and IgA response in the uterine secretions. These results are the first demonstration that a previous immunisation with a suitable antigen can reduce an infection of the reproductive tract of mares.

Bovine polymorphonuclear leukocyte killing of Tritrichomonas foetus.

The role of bovine antibody and complement in bovine neutrophil-mediated killing of Tritrichomonas foetus was investigated. No neutrophil-mediated trichomonacidal activity was detected when Hanks' balanced salt solution, a widely utilized and weakly buffered medium, was used. This lack of neutrophil activity was evident even in the presence of specific bovine antibody and bovine complement. Moreover, the pH of the weakly buffered Hanks' balanced salt solution was observed to fall from pH 7.0 to 5.8 in 4 h at 37 degrees C in the presence of T. foetus. The pH of 5.8 inhibited the bactericidal activity of bovine neutrophils for Staphylococcus epidermidis by 53.2% and may have contributed to the lack of neutrophil-mediated trichomonacidal activity in the weakly buffered salt solution. However, T. foetus was susceptible to bovine neutrophil-mediated destruction when a HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid)-buffered Hanks' balanced salt solution was used (21.8% killing by neutrophils alone). Neither specific bovine immune serum nor purified immune bovine immunoglobulin G2 alone enhanced bovine neutrophil-mediated killing. When complement-sensitized trichomonads were incubated with bovine neutrophils, killing of T. foetus was observed, a result which represented the additive effects of each treatment. Significant (P < 0.05) killing of trichomonads was observed when antibody- and complement-opsonized trichomonads were exposed to bovine neutrophils (> 70% parasite destruction), an effect which reflected the additive nature of each treatment.

Specific antibody to Haemophilus somnus in the bovine uterus following intramuscular immunization.

Sources of anti-Haemophilus somnus antibody in bovine uterine secretions following intramuscular immunization and subsequent intrauterine inoculation of killed H. somnus were investigated. Holstein cattle (n = 21) were immunized with a 270-kDa outer membrane protein from H. somnus (omp-270) by intramuscular injection. At estrus, the cattle were given an intrauterine inoculum of a heat-killed suspension of a homologous strain of H. somnus containing omp-270 (n = 7), a heterologous strain of H. somnus lacking omp-270 (n = 7), or phosphate-buffered saline (n = 7). Uterine secretions were sampled by saline lavage immediately prior to inoculation and at 6, 24, 48, 72, 96, and 120 h after inoculation. Immunoglobulin G subclass I (IgG1) and IgG2 antibody specific for omp-270 were detectable in estrous uterine secretions of all systemically immunized cattle from which an adequate sample was obtained. IgM antibody specific for omp-270 was detected in serum following immunization but was not consistently detected in the uterine secretions of any animal. IgA antibody specific for omp-270 was not detectable in either serum or uterine secretions following immunization or intrauterine inoculation. Ratios of antibody to immunoglobulin and ratios of immunoglobulin to albumin in serum and uterine secretions indicated that about half the IgG1 and essentially all the IgG2 in secretions originated in the serum. Relative titers of IgG1 and IgG2 omp-270-specific antibodies in the uterine lumen and serum gave no evidence for selective transport of either subclass from serum into local secretions. Neither heterologous nor homologous intrauterine inocula detectably altered the serum contribution to antibody in uterine secretions within the sampling period. On the basis of these results, development of a systemic IgG2 antibody response may provide the basis for local immunological protection in the bovine reproductive tract.

Immunoglobulin binding by Tritrichomonas foetus.

A better method for diagnosis of bovine trichomoniasis is needed because culture is slow and somewhat lacking in sensitivity. Immunodiagnosis of Tritrichomonas foetus infection usually involves detection of antigen-antibody reactions with an anti-immunoglobulin conjugate. However, nonspecific immunoglobulin (Ig), bound to the surface of T. foetus, would also be detected by an anti-Ig conjugate and thus would interfere with the specificity of the immunoassay. The goals of this study were to define the binding of bovine immunoglobulins to T. foetus. To determine whether nonimmune binding of Ig to T. foetus occurs, we immunized rabbits with organisms that had been grown in medium containing normal bovine serum and vigorously washed three times with phosphate-buffered saline. The rabbit antiserum had similar titers to T. foetus and to normal bovine serum by enzyme-linked immunosorbent assay (ELISA). Furthermore, two bovine serum proteins were immunoprecipitated by the rabbit antiserum in an immunoelectrophoretogram. One of the serum proteins had a distribution characteristic of IgG2. The rabbit antiserum was then shown to react with purified bovine IgG and IgM by ELISA. Reactivity to IgG was greater. To identify the IgG subisotypes bound and to confirm nonimmune binding of Ig, we grew T. foetus in agammaglobulinemic fetal calf serum and reacted it with IgG1, IgG2, and IgM specific for dinitrophenol (DNP) in ELISA. The binding of IgG2 was greatest, that of IgG1 was next, and that of IgM was least. Little competitive inhibition by DNP was detected, indicating that binding of DNP-specific antibodies was predominantly nonimmune rather than antigen-specific Ig binding. We also demonstrated that T. foetus grew well in medium containing agammaglobulinemic fetal calf serum or serum made agammaglobulinemic by ammonium sulfate precipitation of Igs. This may overcome the problem of low specificity in diagnostic assays as a result of antigen with Ig bound by nonimmune mechanisms.

Neutrophil migration into the bovine uterine lumen following intrauterine inoculation with killed Haemophilus somnus.

Polymorphonuclear neutrophils (PMN) in bovine uterine flushings following intrauterine deposition of killed bacteria were measured and the effect of immune status on the influx of PMN into the uterine lumen during oestrus was determined. Holstein heifers were immunized with a 270-kDa outer-membrane protein (omp-270) from Haemophilus somnus. During oestrus, immunized heifers (n = 21) received an intrauterine inoculum of either a heat-killed suspension of a homologous strain of H. somnus containing omp-270 (n = 7), a heterologous strain of H. somnus lacking omp-270 (n = 7), or phosphate-buffered saline (n = 7). Five additional heifers were inseminated with extended bovine semen. Uterine contents were collected in saline lavage immediately before inoculation (t0) and at 6, 24, 48, 72, 96, and 120 h after inoculation. The semen-inoculated heifers were lavaged only at t120. All groups experienced PMN infiltration which peaked 6 h after inoculation and tended to decline thereafter. Differences were not observed between treatment groups, indicating that neither bacterial inoculation nor immune status was as important in eliciting PMN effusion as the flushing procedure itself.

Characterization of Tritrichomonas foetus antigens by use of monoclonal antibodies.

The specificity for and function of monoclonal antibodies against Tritrichomonas foetus were characterized. Four monoclonal antibodies generated by immunization of mice with live T. foetus were selected on the basis of enzyme-linked immunosorbent assay reactions. The approximate molecular masses of the predominant proteins were determined by Western blotting (immunoblotting). Monoclonal antibody TF3.8 recognized a predominant band at approximately 155 kilodaltons, whereas TF3.2 reacted with several bands. Monoclonal antibodies TF1.17 and TF1.15 recognized broad bands between 45 and 75 kilodaltons. The first two antibodies (TF3.8 and TF3.2) did not react with the surface of T. foetus, as determined by live-cell immunofluorescence, agglutination, and immobilization, whereas two other monoclonal antibodies (TF1.17 and TF1.15) did react with surface epitopes, as determined by these criteria. The latter two monoclonal antibodies also mediated complement-dependent killing of T. foetus and prevented of adherence of organisms to bovine vaginal epithelial cells. One antibody, TF1.15, also killed in the absence of complement. Since these functions are in vitro correlates of protection, the antigens recognized by these monoclonal antibodies may induce protective immunity.

Antibody enhances killing of Tritrichomonas foetus by the alternative bovine complement pathway.

The role of bovine antibody and complement in host defense against Tritrichomonas foetus was measured by using an assay of trichomonad viability based on protozoal uptake of tritiated adenine. Moderate killing was measured in the absence of antibody only with high concentrations of complement-preserved hypogammaglobulinemic bovine serum. However, very low concentrations of hyperimmune serum promoted significant enhancement (P less than 0.05) of killing by complement. Heat inactivation of complement (56 degrees C for 30 min) eliminated antibody-dependent and -independent killing. Similarly, depletion of bovine factor B in serum by heat treatment (50 degrees C for 45 min) abolished antibody-dependent and -independent killing. However, selective inactivation of the classical complement pathway with magnesium ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid did not affect antibody-dependent or -independent killing by complement. These findings demonstrate antibody enhancement of complement-mediated killing of T. foetus by the alternative pathway of bovine complement.

Suspected ciliary dysfunction in Chinese Shar Pei pups with pneumonia.

Chronic pneumonia was investigated in a litter of young Chinese Shar Pei in which 4 of 6 dogs were affected. Serum immunoglobulin concentrations (IgA, IgG, IgM) determined by radial immunodiffusion varied over time, but were not consistently lower in affected dogs, compared with control dogs. Two dogs that died had hydrocephalus and lymphoid depletion, in addition to severe broncho-pneumonia. Evaluation of ciliary ultrastructure in 2 affected dogs revealed random orientation of adjacent respiratory tract or oviductal cilia and a greater number of microtubular disarrangements, compared with control dogs. In vivo tracheal mucociliary clearance of 99mtechnetium macroaggregated albumin was absent in 1 dog examined. The ciliary abnormalities were suspected to have resulted in an inefficient mucociliary transport system predisposing to the development of pneumonia. Further evaluation of 1 Chinese Shar Pei revealed lymphocyte mitogenesis results that were not consistently less than those of a control dog, normal total hemolytic complement values, and normal blood neutrophil chemotaxis.

Adherence of Tritrichomonas foetus to bovine vaginal epithelial cells.

Adherence of Tritrichomonas foetus to bovine vaginal epithelial cells (VECs) in vitro was investigated with fresh washed bovine VECs and log-phase cultures of T. foetus. Observation under phase-contrast microscopy showed that T. foetus usually adhered first by the posterior flagellum and later by the body. Significantly more keratinized squamous epithelial cells were detected with attached parasites than nonkeratinized round epithelial cells. The optimal pH range for attachment was 6.0 to 7.5, with peak attachment at pH 6.5 for squamous VECs. Surface-reactive bovine antiserum to T. foetus prevented adherence to bovine squamous VECs. Inhibition of adherence occurred at nonagglutinating, nonimmobilizing serum dilutions. Antiserum fractions enriched for immunoglobulin G1 inhibited adherence, but fractions enriched for immunoglobulin G2 did not. The inhibitory antiserum was specific for several medium- to high-molecular-weight membrane antigens as detected in Western blots (immunoblots). The ability of surface-reactive antibodies to prevent adherence and to agglutinate and immobilize T. foetus indicates that they may be protective.

Isotypic antibody responses in cattle infected with Haemophilus somnus.

Bovine antibody responses to Haemophilus somnus were compared on the basis of clinical and bacteriological findings. Serum IgG1 and IgM antibody titres were significantly increased in clinically normal cattle that were bacteriologically positive for H somnus from the nasal or vaginal mucosae compared with clinically normal, negative cows. IgG2 titres did not differ significantly between these two groups. However, IgG2 antibody was significantly higher in animals with H somnus disease (pneumonia or abortion) than in clinically normal cattle (whether bacteriologically positive or negative), while IgG1 and IgM titres did not differ between diseased and bacteriologically positive, clinically normal cattle. These antibody trends were duplicated in experimental H somnus abortion or pneumonia, with the greatest response occurring within the IgG2 subclass. Cattle vaccinated systemically with killed whole H somnus produced a predominant IgG2 response with minimal IgG1 and IgM responses. These results demonstrate that IgG2 antibody is consistently elevated in H somnus disease, and suggest that this response may be useful in discriminating diseased from asymptomatic cattle.

Immunoglobulin-binding activity among pathogenic and carrier isolates of Haemophilus somnus.

Nonimmune binding of immunoglobulin to whole bacteria was quantitated for North American isolates of Haemophilus somnus recovered from cattle with pneumonia, reproductive failure (abortion), or thromboembolic meningoencephalitis or from the vagina or prepuce of carrier cattle. Quantitative binding activity covered a wide range, with most pathogenic and carrier isolates demonstrating significant immunoglobulin-Fc binding. Isolates for which Fc binding was not detectable were recovered only from the prepuces of asymptomatic bulls. Expression of Fc-binding activity correlated with the presence of the 41,000-molecular-weight protein (41K protein) and 270K protein. Isolates that lacked Fc-binding activity did not possess 41K or 270K protein. A 33K protein was detected in isolates that lacked Fc-binding activity but not in isolates that bound Fc.

Comparison of IgG Fc receptors from clinical isolates of Streptococcus zooepidemicus.

A receptor binding the Fc region of equine immunoglobulin G (IgG) has been isolated from a heat-extracted preparation of a clinical isolate of Streptococcus zooepidemicus. This Fc receptor has a Mr of 45 x 10(3) and was occasionally seen as an apparent trimer of Mr 130 x 10(3). Antibodies prepared in horses against the receptor could be adsorbed to and eluted from whole live bacteria, confirming the surface location of this protein. Another 11 isolates of S. zooepidemicus from horses with pneumonia, abscesses or endometritis were tested for Fc-receptor activity. Although the Mr of the Fc receptors varied among isolates, their antigenicity was conserved. Thus, the Fc receptor is an attractive candidate for application in the diagnosis of, or protection against, infections with S. zooepidemicus.

Non-immune immunoglobulin binding by "Haemophilus somnus".

In-vitro culture of Haemophilus somnus in liquid or solid media supplemented with bovine blood or serum resulted in non-immune binding of immunoglobulin (Ig) by the organism. This binding was independent of the antigen-combining site of the Ig molecule, since binding of an IgG preparation specific for the hapten dinitrophenol was unaffected by the presence of the homologous antigen. Quantitative comparison of the binding of Ig fragments Fab and Fc demonstrated that the non-immune binding occurred in the Fc region of bovine IgG. The isotypes of Ig that became bound to H. somnus included both bovine IgG subclasses (IgG1 and IgG2), which were bound equally, and bovine IgM.

Isolation and characterization of Fc receptors from Haemophilus somnus.

Receptors that bind the Fc region of bovine immunoglobulin (Ig) have been isolated from the culture supernatant of Haemophilus somnus by chromatography on a Sepharose 4B column. One receptor with a relative molecular weight of 41,000 weakly binds both bovine IgG subclasses, IgA and IgM, while three high molecular weight receptors (350,000, 270,000, and 120,000) strongly bind bovine IgG2, IgA, and IgM. All four Fc receptors are antigenically related and the 41,000 receptor appears to be a subunit of the high molecular weight receptors. In addition to bovine Ig, the purified 270,000 Fc receptor strongly binds horse IgG, rabbit IgG, pig IgG, cat IgG, dog IgG, and sheep IgG. The receptor also reacts weakly with mouse, rat, chicken, human, and guinea pig IgG and does not bind goat IgG. Fc receptors from 19 H. somnus isolates were compared. Variations in the molecular weight of the 41,000 protein were demonstrated among preputial isolates from asymptomatic carriers, but all other isolates appeared to have identically migrating proteins.