Protective Immunity and Competitive Exclusion in Development of Effective Intervention Products for Poultry

Final report summary: Campylobacter jejuni remains the leading cause of bacterial enteritis in England, with over 40,000 cases reported in 2003. The handling and/or consumption of contaminated poultry meat is believed to be a major source of infection. Broiler flocks within the UK are commonly colonised and faecal contamination routinely occurs during processing. At present a two-pronged approach to intervention on the farm has been largely adopted by research workers. This approach comprises (i) the identification of sources of poultry infection and improved biosecurity and (ii) the manipulation of the avian gut environment to exclude the organism by mechanisms including vaccination and competitive exclusion (CE) agents.

<P>
In this project further insight into the molecular basis for campylobacter colonisation of the avian gut has been obtained using a number of post-genomic approaches such as proteomics. These approaches included the first reported uses of subtractive hybridisation and transcript arrays in campylobacter research. This work has now been published (Ahmed et al., 2002; Gaynor et al., 2004). Many of the genes identified by these approaches have roles in metabolism and/or stress survival. A number of the genes identified as having potential roles in colonisation have been subjected to mutagenesis. These mutants, and others of genes identified by ourselves or collaborators using more traditional approaches from similarities with genes in other bacteria, have been tested in a standard, internationally-recognised oral 1-day-old chick colonisation model developed at the VLA. <P>This model has now been used to investigate the individual colonisation-related properties of over 30 genes. In this study the following genes have been shown to have a role in optimal colonisation or persistance: ggt, dtpT, cmeB, Cj415, and Cj1223. Further characterisation of these genes has been undertaken or is in progress. This information may contribute to the development of novel intervention strategies. <P>A major output of this work was the fortuitous discovery of the original clinical isolate (11168O) of the genome sequence strain NCTC11168 in the culture collection bequeathed to the VLA by Dr Martin Skirrow. Unlike the NCTC strain, 11168O has a classical phenotype (spiral shaped, hypermotile and invasive) and is an excellent coloniser of chickens. This isolate now provides an opportunity to fully exploit the genome sequence without the requirement of prior manipulation as in the NCTC strain.
<P>Because the current measures undertaken to enhance biosecurity do not predictably delivery campylobacter-negative flocks, complementary procedures are urgently required to reduce the bacterial load entering the human food chain. Prophylatic approaches, such as vaccination, competitive exclusion, bacteriophages and bactericins, all benefit from the window of opportunitty afforded by the lag phase. Nevertheless, the nature of this important event remains unclear. Preliminary experimental colonisation studies, using chicks derived from commercial campylobacter-negative laying hens, clearly indicate that at least some of this protection is due to maternal immunity. This suggests that the lag phase may be extendable beyond the current 2-3 weeks but further studies using an experimental campylobacter-negative parent flock would be required. <P>
Work on vaccine development has continued from previous studies. During these studies saponin was identified as a avian mucosal adjuvant using a model antigen approach. Using this adjuvant, for the first time efficacy of an oral unconjugated subunit vaccine, has been demonstrated indicating that such vaccination can deliver a 2 log 10 reduction in colonisation. The effective antigen was acid/glycine extract, which contains high proportions of flagellin. Interestingly no antibodies could be detected directed against this antigen in the immunised chickens. Further work is now needed to identify the mechanism of this immunity, demonstrate the cross-protection of such responses and improve the efficacy of the vaccine. Interesting, preliminary evidence suggests that C. jejuni antigens, present in the acid/glycine extract, are specifically cytotoxic for avian, but not human, immune effector cells. This effect may account for differences in disease presentation between humans and birds and may also have consequences for vaccine development using this antigen. <P>Several other vaccination strategies were investigated including in ovo, passive and DNA vaccination. In ovo vaccination was unprotective. Although passive vaccination has a prophylatic effect, consistent with the maternal immunity observed in the lag phase studies, this effect was not therapeutic. Preliminary results from DNA vaccination are being submitted in support of a patent application.
<P>The use of CE agents in chickens has also be further investigated. Previous studies have indicated that the prophylatic use of homologous CE agents can exclude subsequent challenge by other campylobacters. Such an approach would require the identification of strains not only with competitive ability but without the the capacity to cause disease in humans. Unfortunately markers of pathogenicity remain unknown, however, molecular epidemiological studies have identified strains apparently unable to survive the environmental stresses of poultry processing. Therefore, it is unlikely that such strains would be able to cause disease in humans. In order to provide a preliminary screen of the CE ability of such strains, an in vitro model of competition was established based on the method of Barrow and Page (2000). Our investigations confirm that, in general, the CE ability indicated by this in vitro model was consistent with in vivo results. This will enable a reduction in animal use for such studies in the future. Stress-sensitive strains, identified by this screening model, were then tested in vivo. At least one stress-sensitive strain was identified able to totally exclude challenge levels of a good coloniser that normally generated maximal colonisation. In order to progress this approach further, better understanding of the molecular basis and stability of campylobacter survival is required so that appropriate homologous exclusion strains can be selected. This homologous competitive exclusion agent approach was selected because previously published reports indicated that heterologous agents were ineffective for campylobacters. However, based on a recent publication by Hakkinen and Scheitz, the ability of Broilact to exclude a good coloniser was investigated. Total exclusion was achieved in birds challenged within 3 days but declined by 7 days and was lost by 26 days indicating that this protective prophylatic effect was unsustainable in the avian gut. Nevertheless, Broilact , but not Avigard , also demonstracted a significant therapeutic effect, decreasing the colonisation levels by up to 2 x 108 log in colonised birds at the age of slaughter. Unfortunately defined chicken commensal, Lactobacillus strain 85, which has been shown to be effective against other enteric pathogens had no such efficacy against campylobacter.
<P>Overall the results of this project have substantially advanced our understanding of the mechanisms of campylobacter colonisation in the chicken and our opportunities to intervene at the farm and during processing. Clearly vaccination and homologous competitive exclusion strategies can now be developed to extend the lag phase or reduce colonisation levels prophylatically within the flock on-farm while the commercial heterologous competitive exclusion agent, Broilact , could be used immediately as a therapeutic just prior to slaughter. Combinations of these, and other approaches, could reduce the campylobacter load entering the human food chain to below that determined by QRA models to be a risk to human health.

<P>