Campylobacter phase variation and its impact on immunity and vaccine development.

Lead Research Organisation: University of Leicester
Department Name: Genetics


Campylobacter is the commonest cause of bacterial food poisoning in the UK. One of the most frequent sources of infection for humans comes from eating undercooked poultry or cross contamination by uncooked meat during food preparation. The main problem is that Campylobacter can readily colonise poultry to very high levels without causing disease while ingestion of small amounts of bacteria by humans leads to severe gastroenteritis. These infections can result in complications as the bacterium interacts with the human immune system to cause irritable bowl syndrome and Guillain-Barré syndrome, which is an effect on the nerves of patients causing paralysis. Control of Campylobacter in poultry is predicted to significantly reduce human cases of infection. One route to achieve this is by vaccination. Vaccines induce antibodies to specific surface components of bacteria. These antibodies bind to the bacteria cells and interfere with the ability of these cells to colonise the bacterial hosts. Vaccines for Campylobacter are available but have a variable efficacy in preventing the bacterium from colonising chickens. One reason this approach is a problem is because Campylobacter naturally induces the production of antibodies against its surface components but these antibodies do not prevent colonisation or clear the bacteria from birds. Given that a range of antibody responses are generated, it is surprising that these adaptive immune responses do not clear the bacteria. This suggests that the bacteria must have a way of avoiding antibody-mediated control. We believe that Campylobacter can avoid antibodies by switching 'off' the expression of key surface molecules so that they are not recognised by the chicken's antibodies. These switches can occur at high frequencies in a process known as phase variation, which is widespread in bacterial species and is usually associated with surface components. In Campylobacter, phase variation is mediated by repeat sequences in the DNA of the bacteria which undergo a higher than normal frequency of mutation during replication of the bacteria and result in genes being switched 'on' or 'off'. If an antibody acts against the product of one of these phase variable genes, this generates selection for a change in the component by phase variation facilitating persistence of the Campylobacter in the birds. This process is reversible allowing the bacteria to recover the function of the gene when the antibodies dissipate. There is only a limited amount of data on how this process works when Campylobacter infects chickens. This project will identify the frequency of these switches and how the production of antibody impacts on selection of phase variants by the immune system. Specifically, we will investigate whether phase variation is involved in avoidance of antibodies during natural infections and following vaccination. We will assess some novel vaccine candidates for which we have evidence that phase variation may interfere with the effectiveness of induced antibodies and will modify some genes so they cannot switch 'off'. This data will not only be important in understanding how this pathogen survives in poultry but also in humans where the surface components play a major role in disease. The data generated in this project will aid in the development of effective vaccines to protect both animals and man against infections by this bacterium.

Technical Summary

Phase variation (PV) of surface determinants influences both commensal and pathogenic behaviour in many bacterial species. PV in C. jejuni is mediated by mutations in polyC/G tracts. Each genome of C. jejuni contains 16 or more phase variable genes whose products modify capsule, lipo-oligosaccharide and flagella. These surface structures are major targets for host adaptive immunity. We have shown that some phase-variable genes are involved in persistence of C. jejuni in the avian GI-tract suggesting that understanding PV could improve the efficacy of current attempts to limit colonisation. .Reduction of C. jejuni in the poultry food chain would reduce the very high incidence of C. jejuni-associated gasteroenteritis. Colonisation of chickens by C. jejuni induces adaptive immune responses however there is limited evidence for this response promoting clearance. Our published and unpublished data indicates that PV of surface components, including glycosylation of flagella, facilitates colonisation and spread in flocks. There is therefore a strong rationale for investigating the role of PV in immune avoidance and host colonisation. We have already tested the methods required to study PV in vivo and in vitro and shown by analysis of input and output populations from chickens that we can accurately monitor PV in vivo. We will assess PV of 29 genes and the immune responses against C. jejuni following inoculation of chickens with strain 11168. Inoculation of naïve and immunised chickens will permit a thorough assessment of the impact of immune responses on colonisation and selection of phase variants. In silico models will be used to determine whether changes in proportions of phase variants are due to drift, bottlenecks or selection. Outcomes will include an improved understanding of the impact of PV on the course of infections and the variability of vaccine responses. Long-term benefits will comprise identification of novel vaccine strategies.

Planned Impact

Campylobacter jejuni is responsible for the majority of cases of bacterial food-borne gastroenteritis. These infections lead to a small but significant number of severe neurological complications. Plans for engagement of specific potential beneficiaries are highlighted below. Outreach to other beneficiaries will be pursued through publication of research in peer-reviewed journals, presentations at international and national meetings, involvement in local developmental networks and web-accessible resources. Development of vaccines to reduce colonisation of chickens by C. jejuni is an important aim of both policy-holders and commercial companies with interests in the farming sector. Pfizer is a company with a significant interest in C. jejuni and public health. This proposal could impact on development of a vaccine by identifying whether bacterial surface structures are subject to immune selection, by defining specific vaccine targets and by generating attenuated strains, with lower levels of colonisation, for use as live vaccines. An efficacious Campylobacter vaccine could have an immense effect on the burden of food-borne disease, and hence on public health, but would also be highly lucrative for the company involved in production of the vaccine. This research will identify potential vaccine targets/strategies which would then require 3-6 years for further testing and development. Jones, Bayliss and Barrow have experience of working with companies interested in vaccine development (Sanofi Pasteur, Intervet, Pfizer and Lohmann Animal Health) and so will be able to arrange for commercial exploitation of any relevant findings. Intervention strategies to prevent cases of campylobacteriosis in humans are being explored by governmental agencies. The identification of the risk factors associated with different strains and isolates of C. jejuni would inform development of these strategies. Research in this proposal will define whether phase variation contributes to the poor efficacy observed in many vaccine studies and allow rational modification of vaccine strategies to improve their efficacy. Full-implementation of a targeted strategy of this type could drive forward effective vaccine-based control for both animal and human use. The research in this proposal will provide a significant resource for informing lay people about the contributions of genetic mechanisms to the ability of bacteria to cause disease. The results of this research will provide basic material for lectures and tutorial discussions within the higher education establishments in which all the applicants are based. In addition there is the potential for communicating simpler findings to a more general lay audience. This latter aim will be achieved in collaboration with GENIE (, which is a Centre for Excellence in Teaching of Genetics based in the same department as Bayliss. This proposal will generate appropriate visual aids and interactive computer models to communicate concepts such as how rapid generation of genetic variants enables bacterial pathogens to adapt to changes in their environment which can be communicated to the wider educational community. All the applicants have significant teaching experience and Jones is involved in widening participation day-course delivery and has recently completed an MA in Higher education and has carried out projects using computer based delivery of interactive teaching videos and reviewed for the Wikivet project. The University of Nottingham is pursuing an open access policy for teaching materials. Bayliss has delivered presentations for the Royal Society Schools weeks.
Description Our research concerns how hypermutable sequences enable Campylobacter jejuni, a causative agent of foodborne gastroenteritis, to adapt to and survive in the gastrointestinal tracts of chickens.

Key Findings
A key methodological output of this proposal is development of a robust high-throughput method (28-locus CJ11168 PV analysis assay) for detection of changes in the hypermutable sequences of Campylobacter jejuni and translation of these mutations into changes in expression of specific genes (referred to as phase-variable genes). This methodology has utility for studying phase-variable genes in other Campylobacters and in other organisms where hypermutable sequences control expression from phase-variable genes.

A key resource output is application of the 28-locus CJ11168 PV analysis assay to a range of biological samples providing large-scale, portable data sets of utility to a wide range of academic researchers and commercial companies. Biological samples sets include in vitro passage with and without bottlenecks, passage in commercial chickens for up to 52 days and passage in vaccinated chickens.

Another key resource output is a mathematical model for analysis of whether alterations in phase-variable genes are due to mutation alone or to a combination of mutation and selection. This model enables estimation of whether selection is acting on the expression state of a single gene or a combination of genes (i.e. a phasotype). This model has been published in the Bulletin of Mathematical Biology and the publication is associated with this award under the publication tab.

The key biological outputs are detection of temporal trends during persistence of Campylobacter jejuni in chickens with changes in a small number of phase-variable genes, demonstration of the relative stability of phase-variable genes during the first two weeks following colonisation of naive chickens and detection of interactions between phase-variable genes (e.g. switching to the same or opposing expression states). Another finding is that the highest level of variation is in phase-variable genes encoding flagella modification genes, indicating that phase variation of parts of the structure of this surface determinant may contribute to immune evasion. Data analysis is on-going and further biological outputs are likely to be detected.

Completion of Objectives
Objective 1. Reporter constructs were constructed and switching rate data collected for three constructs in two phase-variable loci. An assay was developed for high-throughput detection of phase variation in 28 loci of C. jejuni strain NCTC11168. Two large-scale in vitro experiments were performed and data sets collected. These data sets indicated that selection for specific combinations of expression states can occur during passage on agarose plates. These data sets also indicate that single-cell bottlenecks can generate major changes in population structure.

Objective 2. A 52-day colonisation study in chickens was performed. Bacterial isolates (~3,000) and serum samples were collected. Bacterial isolates were analysed for 28 phase-variable loci. A gene-by-gene data analysis was performed and has identified specific effects of host persistence as well as the robustness of the experimental design. A range of statistical tests have been applied to analysis of these data sets and a rapid, new test for distinguishing the effects of mutation alone from mutation plus selection/bottlenecks has been developed. The serum samples from this experiment have been analysed for reactivity against whole cell lysates and recombinant flagellin. Reactivity has been found to increase as a function of colonisation time and there is a delayed reactivity to glycosylated flagellin as compared to non-glycosylated flagellin.

Objective 3. A vaccination study was performed whereby chickens were inoculated with either a whole cell vaccine or purified lipooligosaccharide. Vaccinated birds and non-vaccinated control birds were challenged with Campylobacter jejuni and samples (isolates and serum) collected. Bacterial isolates from this study have all been subject to analysis with the 28-locus CJ11168 PV analysis assay. Analysis of the data sets is on-going.

Objective 4. Campylobacter jejuni isolates with modified repeat tracts have been constructed and are available for chicken colonisation studies. Colonisation studies with these isolates were not performed due to the time constraints imposed by the large numbers of samples analysed in Objectives 2 and 3.
Exploitation Route Analysis of the phase variable genes may identify novel vaccine candidates that can be exploited by companies interested in developing a C. jejuni vaccine. Our data sets will also indicate if phase variation will interfere with the effectiveness of vaccines and phage therapy. The 28-locus PV assay and variants thereof will have utility for investigating how phase-variable genes contribute to the ability of C. jejuni to colonise and spread between chickens. These data sets will also provide a base-line for investigations of whether phase-variable switches contribute to the ability of C. jejuni to cause disease (i.e. is a different pattern of switches detected in a case of disease as compared to the pattern as this bacterial colonises and persists in chickens).
Sectors Agriculture, Food and Drink,Healthcare

Description Methodologies have been developed for analysing the impact of phase variation on adaptation of Campylobacter jejuni to different selective pressures. We have obtained a Pathfinder grant to explore the potential uses of this technology for diagnostic and vaccine development purposes. Our methodologies have been applied to sample sets of C. jejuni strains representative of long-term persistence in chickens. These experiments indicate that the highest variability is in the genes involved in the modification of the flagella. Altogether we have generated methodologies for exploring how phase variation will impact on intervention strategies (i.e. vaccination, phage therapy) and on biological outputs (e.g. persistence in the food chain, bird-to-bird transmission) with preliminary evidence indicative of flagella modification being a major source of immune evasion and/or host adaptation. We have recently applied this technology to samples provided by VaxAlta who are investigating the functionality of Campylobacter vaccines for reducing colonisation levels in chickens. This data set has been published in two articles, one showing that the phase-variable genes do not impact on vaccine escape of a novel glycoconjugate vaccine and the other showing that bird-to-bird variation was caused by sever bottlenecks at some point between inoculation and establishment of the infection. This methodology has also been utilised to underpin a successful application to the BBSRC to investigate whether phase-variable genes impact on how Campylobacter jejuni invades from the gastrointestinal tracts of birds into muscle and liver tissues.
First Year Of Impact 2015
Sector Agriculture, Food and Drink
Impact Types Economic

Description Host-pathogen interactions important in the movement of Campylobacter jejuni from the broiler chicken gut to edible tissues (CampAttack)
Amount £335,577 (GBP)
Funding ID BB/R016623/1 
Organisation Biotechnology and Biological Sciences Research Council (BBSRC) 
Sector Public
Country United Kingdom
Start 08/2018 
End 07/2021
Description REF Impact Fund
Amount £8,155 (GBP)
Organisation University of Leicester 
Sector Academic/University
Country United Kingdom
Start 03/2016 
End 07/2016
Title GeneScan for C. jejuni Phase Variable Genes 
Description Multiplex PCR and GeneScan assays using fluorescent primers have been developed for 28 of the phase variable loci of C. jejuni strain NCTC11168. 
Type Of Material Technology assay or reagent 
Year Produced 2014 
Provided To Others? Yes  
Impact We have used this method to analyse samples provided by Prof Lone Brondsted's group at the University of Copenhagen. The method has been used to analyse C. jejuni populations subject to selection by bacteriophages. 
Title Reporter constructs 
Description Reporter constructs involving a fusion between lacZ and four C. jejuni genes (0044, 0045, 1318 and 0685) have been constructed. 
Type Of Material Model of mechanisms or symptoms - in vitro 
Provided To Others? No  
Impact We are currently analysing the data to extend our understanding of the mutation rates of Campylobacter genes. 
Description Analysis of phase variable genes of Campylobacter jejuni strain 81-176 
Organisation VaxAlta Inc.
Country Canada 
Sector Private 
PI Contribution Developed an assay for a specific strain of Campylobacter jejuni and tested whether changes in phase-variable genes contributed to the ability of this strain to avoid vaccine elicited immune responses.
Collaborator Contribution Provision of samples.
Impact One publication in Applied and Environmental Microbiology and another manuscript submitted.
Start Year 2016
Title Mutation with Selection Model of Phasotypes (combinatorial Phase-Variable expression states) 
Description This model combines mathematical equations to estimate the effects of mutation and selection on the distribution of ON-OFF expression states of combinations of phase-variable genes for an inoculum and output population. Using a limited amount of experimental inputs and previously determined estimates of mutation rate, this model can determine if the output population could have been derived from the input population by mutation alone or by a combination of mutation and selection. The model is accessed through the Shiny web site and is under pinned by a series of R scripts derived in respect of mathematical inputs. 
Type Of Technology Webtool/Application 
Year Produced 2017 
Impact a publication describing this model is found at the DOI below. Title and authors are:-Mutation and Selection in Bacteria: Modelling and Calibration. C. D. Bayliss, C. Fallaize, R. Howitt & M. V. Tretyakov 
Description Article in Food Safety Magazine 
Form Of Engagement Activity A magazine, newsletter or online publication
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact This article entitled "Campylobacter jejuni: Understanding the New Food Superbug" described the research undertaken at the University of Leicester by both my group and that of Prof Julian Ketley. The article describes how our research my impact on control strategies for dealing with prevention of food-borne gastroenteritis by Campylobacters.
Year(s) Of Engagement Activity 2016
Description Discovery Day 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Supporters
Results and Impact The Discovery Day was organised in collaboration with the Meningitis Research Foundation. The aim was to inform some of their supporters (which included patients or parents of affected individuals) about our research on meningococcal disease and efforts to combat this disease. The Day consisted of a series of short talks aimed at a general audience, questions and answer sessions and an activity (isolation of DNA from bananas).
Year(s) Of Engagement Activity 2016
Description Outreach Event in Tianjin, China 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact We utilised an outreach activity centre on understanding how phase variation can give rise to phenotypic variation and alter the ability of bacteria to cause disease. This activity was part of an outreach event that took place within the Tianjin Natural Museum in Tianjin China and took place on the 12th November 2016. There was a large audience for this event consisting of members of the general public and school parties. The helpers at the event were Chinese foriegn language students who also benefited from involvement in this activity.
Year(s) Of Engagement Activity 2016
Description Simulating Hypervariable Genome Sequences 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Schools
Results and Impact We have generated an educational tool to explain how mutations in specific regions of bacterial genomes contribute to the adaptability of bacterial populations. We have focussed on the bacterium - Campylobacter jejuni - describing the essential features of this species, how it is spread and how it cause disease in humans. The models explains how this bacterial species adapts to environmental changes via changes in hypermutable sequences. The model is interactive enabling the user to set their own parameters and to see how their settings influence the structure of a bacterial population.

This model has been tested with groups of university students and on school groups with a mixed age range of young people/

no actual impacts realised to date
Year(s) Of Engagement Activity 2013,2015,2016
Description Website for Schools 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Schools
Results and Impact We have written a number of web pages and mounted this information onto a website suitable for engagement by pupils in secondary schools.
Year(s) Of Engagement Activity 2016