|Home | About | Journals | Submit | Contact Us | Français|
Campylobacter jejuni is the leading cause of bacterial food borne illness. While helical cell shape is considered important for C. jejuni pathogenesis, this bacterium is capable of adopting other morphologies. To better understand how helical-shaped C. jejuni maintain their shape and thus any associated colonisation, pathogenicity or other advantage, it is first important to identify the genes and proteins involved. So far, two peptidoglycan modifying enzymes Pgp1 and Pgp2 have been shown to be required for C. jejuni helical cell shape. We performed a visual screen of ~2000 transposon mutants of C. jejuni for cell shape mutants. Whole genome sequence data of the mutants with altered cell shape, directed mutants, wild type stocks and isolated helical and rod-shaped ‘wild type’ C. jejuni, identified a number of different mutations in pgp1 and pgp2, which result in a change in helical to rod bacterial cell shape. We also identified an isolate with a loss of curvature. In this study, we have identified the genomic change in this isolate, and found that targeted deletion of the gene with the change resulted in bacteria with loss of curvature. Helical cell shape was restored by supplying the gene in trans. We examined the effect of loss of the gene on bacterial motility, adhesion and invasion of tissue culture cells and chicken colonisation, as well as the effect on the muropeptide profile of the peptidoglycan sacculus. Our work identifies another factor involved in helical cell shape.
Infection by Campylobacter spp, especially Campylobacter jejuni, is considered to be the most prevalent cause of bacterial diarrhoeal disease worldwide . The bacterium is found in the gastrointestinal tract of healthy animals, especially chickens, destined for human consumption. The helical shape of C. jejuni is believed to be important for the bacteria to colonise chickens and during infection, to move through the mucus layer of the gastrointestinal tract and to ‘corkscrew’ into the cells of a human (or other animal) host.
There is limited understanding of how C. jejuni adopts a helical morphology. One study identified a mutation in flhB that affected flagella formation and apparently correlated with C. jejuni becoming rod-shaped , but mutations at other sites in the same flagellar gene resulted in bacteria that remained helical. A mutant in cj1564 (transducer-like protein 3, Tlrp3) has many altered phenotypic characteristics including loss of curvature, but the mechanism for the change in shape is not clear . Occasionally, laboratory strains of C. jejuni lose cell curvature and become rod shaped . C. jejuni can also undergo a transition from helical cells to rod shaped or coccoid forms in older cultures, and under conditions of stress. It is not clear whether C. jejuni can move back and forth between different conformational states during growth. The only genes known to be involved in determination of the helical cell shape of C. jejuni are pgp1 and pgp2 , , , and their protein products are peptidoglycan (PG) peptidases that are important for PG modification , .
The bacterial cell wall is important for providing both rigidity and shape to cells and is composed of layers of PG, or murein, which forms the murein sacculus . In Gram-negative bacteria, such as C. jejuni, the murein sacculus is very thin and lies in the periplasm between the inner and outer membranes. PG is a web of glycan polymers joined by peptide side chains, which are either directly crosslinked or joined by short peptide bridges. The peptide side chains are synthesised at the inner membrane as pentapeptides and may be cleaved into shorter fragments by a number of peptidases. Peptidases may be endopeptidases or carboxypeptidases depending on whether they cleave an internal or C-terminal amino acid, respectively. Peptidases are also classified by whether they hydrolyse the bond between two d-amino acids (DD) or between a l-amino acid and a d-amino acid (LD or DL). The number and length of peptides attached to the glycan backbone provide unique muropeptide profiles for each bacterium. The PG modification pathway in bacteria is known to contain a wide array of carboxy- and endopeptidases responsible for cleaving monomeric, dimeric and trimeric peptides .
To date, only two carboxypeptidases involved in cleaving monomeric peptides have been identified in C. jejuni, Pgp1  and Pgp2 . Pgp2 is an ld-carboxypeptidase, which cleaves disaccharide tetrapeptides into tripeptides . Pgp1 is a dl-carboxypeptidase, which cleaves disaccharide tripeptides into dipeptides . Pgp1 activity is metal-dependent and requires the activity of Pgp2 to provide the tripeptide substrate . When either of the pgp1 or pgp2 genes is mutated in the laboratory the muropeptide profile radically changes and helical cell shape cannot be maintained , . Loss of pgp1 causes a decrease in dipeptides and tetrapeptides and an increase in tripeptides . Loss of pgp2 causes a decrease in dipeptides and tripeptides and an increase in tetrapeptides . Furthermore, overexpression of pgp1 in C. jejuni results in a kinked rod morphology, and muropeptide analysis of the pgp1 overexpressing strain demonstrates a decrease in tripeptides and an increase in dipeptides . Combined, these findings suggest that even subtle changes to proportions of peptides in the PG can affect C. jejuni cell shape.
Pgp2 orthologs are present in a wide range of bacteria that display helical, rod, vibroid (curved rod) or coccoid cell shapes . In contrast, Pgp1 is most highly conserved in helical and vibroid species of the Epsilon- and Delta-proteobacteria . The Pgp1 ortholog in H. pylori, Csd4, has also been characterised as a necessary determinant of cell shape in this helical pathogen. A defined csd4 mutant in H. pylori generates a rod-shaped strain that exhibits a similar muropeptide profile to Δpgp1 in C. jejuni , . The conserved nature of Pgp1 in particular supports the hypothesis that this protein is fundamental to cell curvature and helical cell shape.
While it is known that peptidases can be redundant , , single and double knockouts of Pgp1 and Pgp2 do not demonstrate any change to levels of peptide crosslinking , , suggesting that there remain unidentified PG peptidases in C. jejuni. Thus, further identification and characterisation of the enzymes involved in PG synthesis and modification systems and how these enzymes are localised and regulated is required before we can fully understand how helical shape is generated in C. jejuni.
We recently performed a visual screen of 1933 transposon (Tn) mutants of C. jejuni for changes in cell morphology . Whole genome sequence (WGS) data of the Tn mutants with altered cell shape, directed mutants, wild type (WT) stocks and isolated helical and rod-shaped ‘WT’ C. jejuni, identified a number of different genetic mutations in pgp1 and pgp2, which result in a change in helical to rod bacterial cell shape . In addition, we identified an isolate with a loss of curvature. In this study, we report the genome change leading to the loss of curvature and initial characterisation of the gene.
C. jejuni strains were routinely cultured on Mueller Hinton (MH) agar (Oxoid) supplemented with 5% defibrinated horse blood (Thermo Scientific) and 5 μg/ml trimethoprim (Tp). Defined mutants and complemented strains were selected on 10 μg/ml chloramphenicol (Cm) or 50 μg/ml kanamycin (Km), as appropriate. C. jejuni cultures were grown in standard microaerophilic conditions (5% CO2, 5% H2, 85% N2, 5% O2) at 42 °C, unless otherwise indicated. Electrocompetent Escherichia coli and C. jejuni used in cloning were prepared and transformed as previously described . Bacterial strains and plasmids used in this study are detailed in Table 1.
Sanger sequencing was performed by Source BioScience LifeSciences. WGS was performed at the Wellcome Trust Sanger Institute. Isolates were sequenced as multiplex libraries with 100 or 150 base paired-end reads using next-generation Illumina HiSeq® or MiSeq® sequencing technology, respectively. De novo draft assemblies were created using Velvet v1.2.08 or v1.2.10  and sequencing reads were mapped to the reference genome using SMALT v.0.6.4 and v.0.7.4 . SNPs and INDELs were called using SAMtools mpileup .
Standard methods were used for molecular cloning . Chromosomal and plasmid DNA purification, DNA modification and ligations were performed using commercial kits according to the manufacturers' instructions (QIAGEN, Thermo Scientific, New England Biolabs). DNA concentration was measured using a Nanodrop ND-1000 spectrophotometer (Thermo Scientific). PCR primers were purchased from Sigma (Sigma-Genosys). Thermal cycling was performed in a Gene Amp® PCR System 9700 (PE Applied Biosystems) or T100™ Thermal Cycler (Bio-Rad). Thermal cycling conditions were 96 °C for 2 min, then 30 cycles at 96 °C for 1 min, 55–60 °C for 1 min and 72 °C for 30 s/kb, and finally an extension at 72 °C for 5 min.
Targeted gene deletions of CJJ81176_1105 and CJM1_1064 were performed by exchanging the gene with a chloramphenicol acetyl-transferase (cat) cassette from pRY111 . The cat cassette was amplified with primers containing PstI (dare010) or SacI (dare011) restriction endonuclease (RE) target sites. Flanking regions of CJJ81176_1105 and CJM1_10643 were amplified using upstream and downstream primers (dare_1001 to 4) containing RE sites matched to the cat cassette primers. PCR-amplified fragments were ligated to pUC19 prior to transformation into E. coli. Purified plasmid DNA was used to naturally transform C. jejuni. The correct genomic rearrangement was confirmed by PCR and sequencing using the primers dare_ck1 and ck2, respectively. Primers used in this study are listed in Table 2.
Complementation of CJJ81176_1105 and CJM1_1064 in the targeted deletion strains was performed by amplifying CJM1_1064 from DNA isolated from WT helical isolates of strain M1 using primers darec_F and darec_R. The PCR product was ligated into the Campylobacter shuttle vector pCE107/70 (KmR)  and transformed into electrocompetent mutants. A novel genetic complementation system that we have developed  was used to complement CJJ81176_1105 in the strain 81-176, since the transformation of C. jejuni 81–176 with pCE107/70 was unsuccessful after repeated attempts. DNA isolated from 81 to 176 WT strain served as the template for the amplification of CJJ81176_1105 coding sequence using primers darec_F and darec_R. The PCR product was cloned into pSV009 (KmR) using the BamHI and PstI restriction sites. The resulting plasmid, pSV009-pgp3c (KmR), was confirmed by PCR and sequencing. Following which, the CJJ81176_1105 complementation region was amplified by PCR from pSV009-CJJ81176_1105c using the primers pSV009_GCampl_FW1/RV1 and subsequently introduced into C. jejuni 81–176 by electroporation. Primers used in this study are listed in Table 2.
PG purification and digestion protocols were adapted from those described in Glauner , Li et al.  and Frirdich et al. . HPLC of purified and muramidase-digested C. jejuni PG was performed in the same manner and using the same instrumentation as described in Christie et al. .
The motility of C. jejuni was quantified using motility agar made with 0.4%, 0.6%, 0.8% and 1.0% (w/v) select agar (Sigma) in MH broth. Motility agar was used to fill 6-well plates (7 ml of agar per well) 20 min prior to use. C. jejuni isolates were transferred via pipette tip from 12 h lawn growth (on MH agar plates) into each well of the motility agar. For each strain to be tested, three replicate 6-well plates were incubated for each motility agar concentration. Motility was measured as the diameter of the halo of motility after 12 h incubation.
Caco-2 cell lines were purchased from the ATCC (CC-L244, HTB-37). Cells were grown using DMEM (Gibco) supplemented with 10% FBS and 1% non-essential amino acids. Cells were routinely grown in 75 cm2 tissue culture flasks and incubated at 37 °C with 5% CO2 in a humidified atmosphere.
Caco-2 cells were seeded at 5 × 104 cells on 24 well plates (Greiner) until confluency was observed. Caco-2 cells were infected with different C. jejuni strains at a multiplicity of infection (MOI) of 100. To assay adherence/invasion, infected cells were incubated at 37 °C with 5% CO2 in a humidified atmosphere for 2 h. At this point, non-adherent bacteria were removed, subjected to 10-fold serial dilutions and plated on BHI blood agar plates with 5 μg/ml trimethoprim. Wells were washed three times with PBS, and cells were lysed with 0.1% Triton-X-100 in PBS for 15 min. Lysed cells were subjected to 10-fold serial dilutions and plated on BHI blood agar plates with 5 μg/ml trimethoprim. To determine the number of internalised bacteria, infected Caco-2 cells were incubated at 37 °C with 5% CO2 in a humidified atmosphere. After 2 h, the media overlaying the infected cells was changed to complete DMEM containing 250 μg/ml gentamycin sulphate and infected cells were incubated at 37 °C with 5% CO2 in a humidified atmosphere for a further 2 h. Cells were then washed three times with PBS and lysed with 0.1% Triton X-100 in PBS for 10 min. Serial dilutions of the cell lysates were carried out and plated on BHI blood agar plates with 5 μg/ml trimethoprim. Dilutions of mutant C. jejuni strains were plated on BHI blood agar plates containing 10 μg/ml chloramphenicol, whereas genetically complemented strains were recovered on MH blood agar plates supplemented with 10 μg/ml chloramphenicol and 50 μg/ml kanamycin. All plates were incubated for 48 h under microaerophilic conditions at 42 °C before colony counting took place. For both total association and invasion experiments, the percentage of C. jejuni interacting with Caco-2 cells was calculated as a percentage of the non-adherent fraction, to account for various strain survival within DMEM (n = 3).
All work was conducted in accordance with UK legislation governing experimental animals under project licence 40/3652 and was approved by the University of Liverpool ethical review process prior to the award of the licence.
One-day-old Ross 308 broiler chicks were obtained from a commercial hatchery. Chicks were housed in the University of Liverpool, High Biosecurity Poultry unit. Chicks were maintained in floor pens at UK legislation recommended stocking levels allowing a floor space of 2,000 cm2 per bird and were given ad libitum access to water and a pelleted laboratory grade vegetable protein-based diet (SDS, Witham, Essex, UK). Chicks were housed in separate groups at a temperature of 30 °C, which was reduced to 20 °C at 3 weeks of age. Prior to experimental infection, all birds were confirmed as Campylobacter-free by taking cloacal swabs, which were streaked onto selective blood-free agar (mCCDA) supplemented with Campylobacter Enrichment Supplement (SV59; Mast group, Bootle, Merseyside, UK) and grown for 48 h in microaerobic conditions at 41.5 °C. All microbiological media were purchased from Lab M (Heywood, Lancashire, UK).
At 21 days of age birds were infected with 2 × 109 CFU of either C. jejuni M1 or the CJM1_1064 mutant. At 5 days post infection (p.i.), chickens were killed by cervical dislocation. At necroscopy the ceca were removed aseptically and the cecal contents plated onto mCCDA Campylobacter selective agar plates for enumeration as previously described .
When isolating helical and rod bacteria from WT C. jejuni strains based on colony morphology as described in Esson et al. , we noticed a colony, which was not quite as grey and flat as the typical rod colony morphology but still distinct from the helical colony morphology, and was composed of bacteria with a loss of curvature (‘kinked rod’ cell morphology, 81176_KR). This cell morphology contrasted with the helical morphology of WT 81–176 and was confirmed by scanning electron microscopy (SEM) (Fig. 1). Based on a qualitative assessment of our SEM analyses, there appears to be a difference in the degree of curvature between shorter (presumably younger cells) and longer (presumably older) cells of the ‘kinked rod’ bacteria.
The 81–176 isolate with a loss of curvature (81176_KR) was analysed by WGS and was a change in the number of bases in a documented phase variable region (PVR), and two unique point mutations.
Phase variation (PV) enables genetic and phenotypic variation in a number of bacteria, including C. jejuni , , . Regions of the bacterial genome that are prone to these reversible mutations are called PVR. In C. jejuni the PVRs are typically homopolymeric tracts (HTs) that are highly susceptible to slipped-strand mispairings, which alter the length of the tracts and generate frameshift mutations during DNA replication and repair , . In this way, PVRs are able to randomly switch genes ‘on’ and ‘off’ and stochastically regulate gene expression . The unique PV pattern was in PVR3 and demonstrated a mostly ‘on’ length, which contrasted with the mostly ‘off’ lengths of other 81–176 isolates. PVR3 in strain 81-176 is 118 bp upstream of CJJ81176_0590, encoding a putative uncharacterised protein. This PVR correlates to PVR2 in strain M1, which demonstrated ‘on’ lengths in helical M1 Tn mutants and WT isolates (data not shown). For this reason, as well as the absence of a CJJ81176_0590 ortholog in strain NCTC11168 , we hypothesised that the altered polyG tract upstream of CJJ81176_0590 was not responsible for the loss of curvature of isolate 81176_KR.
One of the unique point mutations in 81176_KR was a non-synonymous SNP (G > A) in rpiB at base location 860819 (CP000538.1). This SNP was predicted to cause a single glutamic acid to lysine amino acid change. The protein product of rpiB, ribose 5-phosphate isomerase B, is involved in carbohydrate metabolism  and our searches did not demonstrate any link between this enzyme and bacterial cell shape. Therefore, we hypothesised that a single amino acid change to RpiB was not responsible for the observed loss of curvature in 81176_KR.
The other point mutation detected in 81176_KR was an INDEL in CJJ81176_1105, a predicted LytM peptidase-encoding gene. This single guanine deletion (2G > G) at base location 1022254 (CP000538.1) was predicted to cause a truncation at residue 65 of the 300 amino acid protein product. BLAST analysis of CJJ81176_1105 revealed that this gene is highly conserved (>50% coverage and >70% identity) in helical Campylobacter spp. (data not shown). We investigated the presence and allelic variances of CJJ81176_1105 in 859 genomes of C. jejuni and C. coli. The genomes were from a wide range of isolates: 192 from clinical, agricultural and wild bird sources , 319 from multiple stages of poultry processing, including farms, abattoirs and retail chicken meat  and 348 from clinical cases . Analysis of these genomes revealed that CJJ81176_1105 is conserved (data not shown), suggesting this gene is core to both Campylobacter species.
A detailed comparison of the translated sequence of CJJ81176_1105 from four laboratory C. jejuni strains (M1, 81116, 81–176 and NCTC11168) demonstrated identical amino acid sequences at all except four residues (Fig. S1). The protein product of CJJ81176_1105 contains prefoldin, coiled-coil and peptidase domains. Prefoldin is a coiled-coil-containing molecular chaperone that assists in the proper folding of polypeptide products . In eukaryotes, prefoldin is responsible for the folding and localisation of the cytoskeleton components actin and tubulin . The peptidase domain is conserved within the Peptidase M23 (LytM) family, which is composed of zinc-dependent endopeptidases often involved in cell division, elongation and shape determination .
Further analysis revealed CJJ81176_1105 to be orthologous to csd1 (cell shape determinant 1) in H. pylori . In the helical pathogen H. pylori, a targeted deletion of csd1 results in a curved rod morphology, which is fully complemented when csd1 is supplied elsewhere on the chromosome . Sycuro et al.  compared the Csd1 protein product from H. pylori to the crystallised LytM endopeptidase from Staphylococcus aureus, which demonstrated conserved LytM active site residues in Csd1. Although the csd1 ortholog in C. jejuni was identified by Sycuro et al., the gene was unable to complement the Δcsd1 phenotype in H. pylori . Another group also investigated the role of the csd1 ortholog in C. jejuni morphology but results from these preliminary studies were reported as inconclusive (unpublished work mentioned in Frirdich et al. ).
Due to the sequence similarity between csd1 and CJJ81176_1105 and the similar ‘intermediate’ morphologies of the H. pylori Δcsd1 strain and 81176_KR, we hypothesised that the frameshift mutation in CJJ81176_1105 was responsible for loss of curvature morphology of 81176_KR. Moreover, due to its predicted endopeptidase function , , we hypothesised that the CJJ81176_1105 protein product might be involved in the same PG modification cascade as Pgp1 and Pgp2.
To test whether the mutation in CJJ81176_1105 was responsible for the loss of curvature of 81176_KR, we constructed targeted deletions, and complemented strains, on different C. jejuni WT backgrounds (CJJ81176_1105 for strain 81-176 and CJM1_1064 for strain M1). The defined mutants displayed loss of curvature morphologies, which were restored to helical morphologies by complementation (Fig. 2). From these data, we conclude that CJJ81176_1105, and its homolog in other strains, is necessary for a fully helical morphology in C. jejuni. We next compared physiological characteristics of the CJJ81176_1105 and CJM1_1064 mutants and complemented strains with helical-shaped and rod-shaped WT isolates.
We tested the motility of WT helical and rod isolates, against the CJJ81176_1105 and CJM1_1064 mutants and complemented strains, across a range of motility agar concentrations (Fig. 3). The results showed that migration through increasing agar concentrations (decreasing porosity) was significantly reduced in the WT-rod (INDEL in pgp1) and CJJ81176_1105 and CJM1_1064 mutants compared to the helical isolates (WT-helical and complemented strains). Our work demonstrates that at 0.4 and 0.6% (w/v) agar, the motility of the CJJ81176_1105 and CJM1_1064 mutants is slightly greater (although not statistically significant) from the WT-rod isolates. At 0.8% (w/v) agar the motility of the WT-rod (INDEL in pgp1) and CJJ81176_1105 and CJM1_1064 mutants were comparable, and significantly reduced from the helical isolates. All the isolates were effectively non-motile through 1.0% (w/v) agar, i.e. all isolates measured 1 mm in diameter, roughly equivalent to the original pipette stab.
Next, the ability of the CJJ81176_1105 and CJM1_1064 mutants to adhere to, and invade, Caco-2 cells was measured (Fig. 4). The WT-rod and CJJ81176_1105 and CJM1_1064 mutants displayed statistically significant reductions in adhesion and invasion compared to the WT-helical and complemented strains. For both C. jejuni backgrounds, M1 and 81–176, the adherence and invasion of the mutant was slightly greater (although not statistically significant) from the WT-rod isolate (INDEL in pgp1).
Chicken colonisation experiments were performed using C. jejuni strain M1 WT-helical isolate, a natural poultry isolate which is an efficient coloniser of chickens , and the CJM1_1064 mutant. Chickens were inoculated with 2 × 109 CFU of either strain. At 5 days post infection (p.i.), chickens were killed and the cecal contents plated onto mCCDA Campylobacter selective agar plates as previously described . The viable counts per gram of cecal contents revealed that there was no difference in the colonisation of the WT or the CJM1_1064 mutant (Fig. 5).
Muropeptide analysis via high-performance liquid chromatography (HPLC) and mass spectrometry (MS) was used to compare the PG sacculi of WT, CJJ81176_1105 and CJM1_1064 mutant and complemented C. jejuni strains. Muropeptide profiles of mutanolysin-digested PG sacculi isolated from CJJ81176_1105 and CJM1_1064 deletion strains appeared virtually identical to muropeptide profiles derived from the parental strains (Fig. 6). Hence in contrast to Pgp1 and Pgp2, CJJ81176_1105 (and CJM1_1064) activity in modelling the PG sacculus appears to be below the threshold for detection via the muropeptide analysis technique, and definitive identification of the hydrolytic bond specificity against PG will require further attention.
There is no effective vaccine against C. jejuni and preventative measures aimed at reducing environmental contamination have so far proved ineffective. There is a need for alternative strategies to reduce campylobacteriosis. Most of the Campylobacteraceae are helical and it appears that the helical shape of C. jejuni is important for its ability to colonise its hosts and cause disease. To address this hypothesis, it is essential to know how helical shape is determined in C. jejuni, both genetically and biochemically, but we currently have limited understanding of this. Loss of helical cell shape through interference may hold therapeutic potential by reducing this pathogen's virulence or ability to colonise animals.
This work identifies CJJ81176_1105 as a novel cell shape determinant in C. jejuni. CJJ81176_1105 was identified by the isolation and WGS analysis of a bacterium with loss of curvature within our laboratory WT C. jejuni 81-176 stock. This isolate was found to contain a nonsense mutation in CJJ81176_1105. Targeted deletions of the gene in both the C. jejuni 81–176 and M1 backgrounds reproduced the loss of curvature morphology of the original isolate. This intermediate cell shape was rescued by supplying the gene in trans, which confirmed that it is necessary for a fully-helical C. jejuni morphology. The homology of CJJ81176_1105 to the endopeptidase Csd1 in H. pylori suggests that it may also be involved in this muropeptide cascade.
The intermediate nature of the loss of curvature morphology implies that there exists a hierarchy to helical cell shape maintenance within C. jejuni. Based on the evidence demonstrating the importance of endo- and carboxypeptidases in helical cell shape , , , , this hierarchy is likely a product of PG peptide lengths and the degree of crosslinking that promotes the cell wall to twist.
As a possible explanation for the differences between the helical, loss of curvature and rod forms of C. jejuni, we hypothesise that shorter peptides within the cell wall are localised to the inside of the helix. As such, while the loss of di- and tripeptides in pgp1 and pgp2 mutants prevents the maintenance of any curvature ,  the predicted reduction of tetra- and pentapeptide substrates in the CJJ81176_1105 mutant may merely ration the distribution of di- and tripeptide products throughout the PG, lessening the tension of the helix. To address this hypothesis, future work will require an investigation into the distribution of PG peptides throughout the cell wall in situ.
This work was funded by The Wellcome Trust through a PhD training studentship awarded to DE, and was supported by an Isaac Newton Trust/Wellcome Trust ISSF/University of Cambridge joint research grant awarded to AJG. SG was funded by the Biotechnology and Biological Sciences Research Council grant BB/K004514/1. AEM, NRT and JP were supported by the Wellcome Trust grant number 098051. SKS was funded by Biotechnology and Biological Sciences Research Council grant BB/I02464X/1, Medical Research Council grant MR/L015080/1 and Wellcome Trust grant 088786/C/09/Z. GM was supported by a National Institute for Social Care and Health Research Fellowship (HF-14-13). The authors have no conflicting financial interests. The funders had no role in the study design, data collection and interpretation, or the decision to submit the work for publication.
Scanning Electron Microscopy was kindly performed by Dr Jeremy Skepper (Cambridge Advanced Imaging Centre). Dr Gemma Chaloner and Dr Lizeth Lacharme-Lora assisted during the chicken colonisation experiments.
Appendix ASupplementary data related to this article can be found at http://dx.doi.org/10.1016/j.micpath.2017.01.042.
The following is the supplementary data related to this article: