|Home | About | Journals | Submit | Contact Us | Français|
Campylobacter jejuni is a leading cause of food-borne illness in the United States. Despite significant recent advances, its mechanisms of pathogenesis are poorly understood. A unique feature of this pathogen is that, with some exceptions, it lacks homologs of known virulence factors from other pathogens. Through a genetic screen, we have identified a C. jejuni homolog of the VirK family of virulence factors, which is essential for antimicrobial peptide resistance and mouse virulence.
Campylobacter jejuni is a leading cause of infectious diarrhea in industrialized and developing countries (2, 67). Although most often self-limiting, C. jejuni infections can also lead to severe disease and harmful sequelae, such as Guillain-Barré syndrome (4, 55). Despite the significant progress made during the past few years, the mechanisms of C. jejuni pathogenesis remain poorly understood. A number of potential virulence factors have been identified, and in some cases, their role in virulence and/or colonization has been demonstrated in animal models of infection. For example, motility has been shown to be crucial in order for C. jejuni to colonize or cause disease in several animal models of infection (1, 15, 30, 54). A variety of surface structures, such as adhesins (34, 40, 64) and polysaccharides (5, 6), and glycosylation systems (38, 74), which presumably modify some of these surface structures, have also been shown to be important for infection. Additional studies have revealed the importance of specific metabolic pathways in C. jejuni growth both in vitro and within animals (16, 25, 31, 60, 76). The ability of C. jejuni to invade and survive within nonphagocytic cells has also been proposed to be an important virulence determinant (21, 41, 57, 58, 68, 75, 80).
The available genome sequences of several C. jejuni strains have provided significant insight into C. jejuni physiology and metabolism (22, 32, 62, 63, 65). Remarkably, however, analysis of these C. jejuni genome sequences has revealed very few homologs of common virulence factors from other pathogens. A notable exception is the toxin CDT (cytolethal distending toxin), which is also encoded by several other important bacterial pathogens (36, 44, 45). In this paper we describe the identification of a transposon insertion mutant in C. jejuni 81-176, which results in increased susceptibility to antimicrobial peptides and a significant defect in the ability of the organism to cause disease in an animal model of infection. The insertion mutant was mapped to the CJJ81176_1087 open reading frame (Cj1069 in the C. jejuni NCT 11168 reference strain), which encodes a protein with very significant amino acid sequence similarity to the VirK (DUF535) family of virulence factors (13, 20, 56).
The C. jejuni strains used in this study are listed in Table S1 in the supplemental material and were derived from the wild-type strain 81-176 (9, 42). Routinely, C. jejuni was grown on Brucella broth agar or blood agar plates (Trypticase soy agar supplemented with 5% defibrinated horse blood [Becton Dickinson]) at 37°C under an atmosphere of 10% CO2. C. jejuni transformants were selected on plates supplemented with kanamycin (50 μg ml−1) or chloramphenicol (7.5 μg ml−1), as indicated. For liquid cultures, C. jejuni strains were grown in brain heart infusion (BHI).
For growth assays, C. jejuni cultures were adjusted to an optical density at 600 nm (OD600) of 0.1 to 0.2 in 3 ml of BHI and were grown under an atmosphere of 10% CO2 in a rotating wheel (50 rpm). Bacterial growth was monitored by measuring the OD600 with a spectrophotometer (Spectronic 20; Genesys). The plating efficiency of the cultures was monitored by plating appropriate dilutions on blood agar plates.
COS-7 (an African green monkey kidney fibroblast-like cell line) or T84 (a human colon carcinoma cell line) cells were obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum. All cell lines were grown under a 5% CO2 atmosphere.
The Tn552 insertion point was determined by nucleotide sequencing using primers (5′-CCAATTCTCGTTTTCATACC-3′ and 5′-GGATCAAGCCTGATTGGGAG-3′) complementary to the aphA cassette present in the transposable element. The transposable element was found to be inserted at nucleotide 260 of the CJJ81176_1087 open reading frame of C. jejuni.
DNA manipulations were performed according to standard protocols described elsewhere (69). Bacterial genomic and plasmid DNAs were isolated using the DNeasy tissue and QIAprep spin miniprep kits (Qiagen), respectively.
To construct a tagged version of C. jejuni 81-176 VirK, a FLAG epitope tag was fused to the C terminus of the CJJ81176_1087 gene. CJJ81176_1087 was PCR amplified with primers Cj1069/81-176_5_EcoRI (GGAATTCGCTTGATGGAGGACATATACTTT) and Cj1069/81-176_3_NcoI_FLAG (CATGCCATGGCGTCTTGAATATCCAATGTCTTAT) and was cloned in frame into a vector encoding three copies of the FLAG epitope. To move the tagged version of virK into the C. jejuni chromosome, the virK-3xFLAG gene was cloned downstream of the chloramphenicol acetyltransferase cassette (cat cassette) of the previously described plasmid pSB3021 (80). This plasmid allows the integration of a gene of interest into a C. jejuni CJJ81176_1539 chromosomal locus that encodes a component of a restriction/modification system and is not required for virulence. The resulting plasmid was introduced into both wild-type C. jejuni 81-176 and the virK mutant derivative by natural transformation, creating C. jejuni strains CB32 and CB35, respectively. Natural transformation in C. jejuni was carried out as described elsewhere (27, 78).
The ability of C. jejuni to invade cultured intestinal epithelial cells was evaluated by a standard gentamicin protection assay as described elsewhere, with minor modifications (59, 79). T84 or COS cells were split to about 70% confluence (~105 cells per well) in 24-well plates the day before an experiment. To prepare the bacterial inoculum, C. jejuni strains were first grown in BHI medium to early-logarithmic phase (OD600, ~0.5) and then adjusted to an OD600 of ~0.1 using Hanks balanced salt solution (HBSS) (Invitrogen). Cell monolayers were infected at a multiplicity of infection (MOI) of ~10, and the plates were spun down at 200 × g for 5 min to enhance bacterium-host cell contact and were then incubated for 2 h at 37°C under 5% CO2. Following incubation, the infected monolayers were washed three times with buffered saline containing 1% gelatin, and prewarmed Dulbecco's modified Eagle medium containing gentamicin (200 μg ml−1) was added for 3 h to kill extracellular bacteria. Cells were washed again and lysed with 0.1% sodium deoxycholic acid (Sigma) in phosphate-buffered saline (PBS) to release the intracellular bacteria. Alternatively, cells were lysed by mechanical disruption in water. The invasion ability was expressed as the percentage of the inoculum that survived treatment with gentamicin. Each assay was conducted in triplicate and was independently repeated at least three times.
To determine sensitivity to sodium deoxycholate under conditions resembling those of the invasion assay, C. jejuni strains were first grown in BHI medium to early-logarithmic phase (OD600, ~0.5) and then adjusted to a concentration of ~108 bacteria per ml in HBSS. One hundred microliters of bacterial cell suspension was transferred to microcentrifuge tubes containing 900 μl of either 0.1% sodium deoxycholate or HBSS. Samples were left at room temperature for 20 min, and the numbers of CFU were determined by plating on blood agar.
The surface hydrophobicity assay was carried out as previously described (33).
The bacterial culture to be tested was adjusted to an OD600 of ~0.1, and 2 μl was spotted onto soft agar (0.5%, wt/vol). Plates were incubated for ~24 h at 37°C; the swarming diameter of the tested strain was compared to that of the wild type and the nonmotile flaA (or motA) strain. In addition, bacterial cultures were examined for motility by light microscopy immediately before their utilization in the invasion assays.
The ability of C. jejuni to colonize mice was tested with a recently developed mouse infection model that has been described in detail elsewhere (80). Six- to 8-week-old myd88−/− nramp-1−/− male mice were infected orally by stomach gavage with 109 CFU or intraperitoneally with 107 CFU of the different C. jejuni strains. At the indicated time points, feces were collected into BHI broth, weighed, and plated onto blood agar plates containing Campylobacter-selective supplements (Oxoid SR0167E) to determine the number of CFU per gram of feces. To differentiate between C. jejuni strains, antibiotics (kanamycin at 50 μg ml−1 or chloramphenicol at 7.5 μg ml−1) were added to selective medium when appropriate. At the end of the experiment, mice were sacrificed; their organs were aseptically removed and homogenized in HBSS; and the bacterial loads in the different tissues were enumerated by plating on selective plates as described above. Statistical analysis of the results was carried out with the Wilcoxon matched-pair signed-rank test.
To test the resistance of the different C. jejuni strains to the bactericidal effects of serum, cultures were grown in BHI broth to early-logarithmic phase (OD600, ~0.5), recovered by centrifugation, and suspended in HBSS medium. About 5 × 107 bacteria were added to 500 μl of HBSS containing an indicated percentage of either normal or heat-inactivated (30 min at 60°C; stored at −20°C) rabbit serum. Samples were incubated at 37°C under a 5% CO2 atmosphere for the time indicated, and series of 10-fold dilutions in HBSS containing 10 mM EDTA were plated onto blood agar plates. Survival was calculated as the percentage of CFU obtained for serum-treated bacteria relative to the number of CFU from the heat-inactivated serum control.
To determine sensitivity to polymyxin B sulfate (Sigma) or Magainin-1 (AnaSpec, San Jose, CA), C. jejuni strains were first grown in BHI medium to early-logarithmic phase (OD600, ~0.5) and then adjusted to a concentration of ~106 bacteria per ml in PBS. Cell suspensions were treated with different concentrations of polymyxin B or Magainin-1 and were incubated for 45 min (for polymyxin B) or 20 min (for Magainin-1) at 37°C. The CFU after treatment was determined by plating on blood agar. Statistical analysis of the results was carried out with the Student t test.
Bacterial subcellular fractionation was carried out as described elsewhere (43), with some modifications. C. jejuni cultures were grown to an OD600 of ~0.3 to 0.5, and cells were harvested by centrifugation at 4,000 rpm for 15 min. Bacterial cells were resuspended in 1 ml of chilled PBS containing 0.1 mg of lysozyme/ml, 1 mM phenylmethylsulfonyl fluoride, and 10 mM EDTA; then they were incubated on ice for 1 h. C. jejuni cells were lysed by sonication (Branson Sonifier with a microtip duty of 50%; power 5) and clarified by centrifugation at 17,000 × g for 10 min (twice) to eliminate cell debris and unbroken cells. Samples were passed through a 0.22-μm-pore-size filter (Millex-GV) to remove unlysed bacteria. The supernatant was transferred to a fresh tube, and cell envelopes were collected by centrifugation at 100,000 × g for 1 h at 4°C. After centrifugation, the supernatant containing cytoplasmic and periplasmic soluble proteins was transferred to a fresh tube, and pellets were washed with PBS by ultracentrifugation (100,000 × g for 1 h at 4°C). The tagged protein was extracted from membrane pellets by suspending membrane fractions in either (i) PBS, (ii) 1 M NaCl and 100 mM Na2CO3 (pH 11.0), or (iii) 0.5% sodium lauroyl sarcosinate (Sarkosyl NL; Sigma) and 5 mM EDTA. After 1 h of incubation at 4°C, samples were subjected to high-speed centrifugation (100,000 × g for 1 h at 4°C), and the resulting pellet and supernatant fractions were analyzed for the presence of VirK by Western immunoblotting. Analysis of protein in culture supernatants was done as described elsewhere (17).
Intracellular and extracellular bacteria were enumerated by fluorescence microscopy as previously described (23). COS cells were first grown on poly l-lysine-treated 12-mm-diameter round glass coverslips to ~70% confluence and then infected with C. jejuni at an MOI of 50 in HBSS for 2 h. Cells were briefly washed with PBS, fixed in a 2% paraformaldehyde solution for 15 min at room temperature, and incubated in a 5% suspension of nonfat dry milk (Carnation) in PBS for 30 min. Fixed cells were incubated with rabbit anti-C. jejuni 81-176 serum in PBS containing 5% (wt/vol) milk for 30 min. Cells were washed three times with PBS and incubated with Alexa Fluor 488-conjugated goat anti-rabbit immunoglobulin G (1:1,000; Molecular Probes) for 30 min. The coverslips were washed with PBS and permeabilized with PBS containing 0.05% (wt/vol) saponin for 15 min. Cell were incubated again with rabbit anti-C. jejuni 81-176 serum, washed three times with PBS containing 0.05% (wt/vol) saponin, and incubated with Alexa Fluor 568-conjugated goat anti-rabbit immunoglobulin G (1:1,000; Molecular Probes) to stain intracellular bacteria. Cell nuclei were stained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI). The coverslips were mounted onto glass slides with ProLong Gold antifade reagent (Molecular Probes), and the numbers of intracellular bacteria per cell were quantified under a fluorescence microscope (Nikon Diaphot 300) with a 100× (numerical aperture, 1.4) oil immersion objective.
C. jejuni lipooligosaccharides (LOS) were prepared from whole-cell lysates as described previously (19). LOS were separated on 16% Tricine gels (70) and either silver stained as described elsewhere (61) or analyzed by Western immunoblotting using primary rabbit antibodies generated against heat- or formalin-killed C. jejuni strain 81-176 bacteria.
Membrane fractionation was carried out essentially as described elsewhere (77) using a continuous 29-to-50% sucrose gradient (wt/vol) and a gradient maker (Gradient Station; BioComp Instruments, Inc., Canada). Fractions that had the highest correspondence to the inner and outer membrane proteins were subjected to differential fluorescence 2-dimensional (2D) gel electrophoresis proteome profiling at the Keck Biotechnology Resource Laboratory (Yale University) using protein labeling with Cy3 and Cy5 N-hydroxysuccinimidyl ester dyes according to its standardized protocol (http://keck.med.yale.edu/dige).
During a screen to identify C. jejuni 81-176 Tn552-kan transposon insertion mutants defective for entry into and/or survival within epithelial cells (V. Novik and J. E. Galán, unpublished data), we isolated an insertion mutant that mapped to the CJJ81176_1087 open reading frame (Cj1069 in the reference strain NCTC 11168), which encodes a homolog of the VirK family of proteins (Fig. (Fig.1).1). VirK was first identified in Shigella flexneri as a factor required for the efficient intercellular dissemination of this pathogen (56). A homolog of VirK has also been identified in Salmonella enterica serovar Typhimurium as a virulence factor required for intracellular replication and mouse virulence (13, 20). We will therefore refer to Cjj81176_1087 as VirK. To verify that the entry defect was due to the inactivation of virK, we constructed a C. jejuni virK mutant derivative carrying a FLAG epitope-tagged wild-type copy of the inactivated gene elsewhere in the chromosome (see Materials and Methods). As shown in Fig. Fig.2A,2A, the invasion defect of the virK mutant was fully restored by the introduction of a functional copy of virK. All strains exhibited equal sensitivities to gentamicin (data not shown).
The invasion assay protocol used in these studies requires lysis of the mammalian cells by addition of 0.1% sodium deoxycholate. At this concentration, this detergent does not affect the viability of the wild-type bacteria. However, certain mutations are known to increase the sensitivity of C. jejuni to this compound (49). To rule out the possibility that the observed phenotype of the C. jejuni virK mutant strain was due to increased sensitivity to the detergent, this strain was treated with 0.1% sodium deoxycholate for 20 min, and its viability was compared to that of the equally treated wild-type parent strain. No difference in the sensitivity of these strains to sodium deoxycholate was detected (see Fig. S1 in the supplemental material), and equivalent results were obtained when cells were lysed mechanically in the presence of water during the invasion assay (data not shown). Therefore, the differences observed in the invasion assay reflect the intrinsic differences in these strains' abilities to invade and/or survive within cultured intestinal epithelial cells.
The gentamicin protection assay used in the screen cannot distinguish between a defect in bacterial entry and a loss of viability immediately after entry. To distinguish between these phenotypes, we examined the ability of the C. jejuni virK mutant to enter cultured epithelial cells by a microscopy-based assay. This assay uses a staining protocol that can distinguish between internalized and extracellular bacteria. Furthermore, this assay is independent of strain culturability after infection. Using this assay, we detected no significant defect in the ability of the C. jejuni virK mutant to enter epithelial cells (Fig. (Fig.2B).2B). We therefore conclude that the viability of C. jejuni virK decreases shortly after entry into epithelial cells.
Based on the secondary structural features and hydropathy profile of Shigella flexneri VirK, and the fact that it lacks a typical secretion signal, it has been hypothesized previously that VirK is localized in the bacterial cytoplasm (56). However, no studies have been conducted to investigate specifically the localization of any VirK family member. Although some studies have suggested that Shigella flexneri VirK and its homolog in S. Typhimurium may be involved in the remodeling of the bacterial envelope (12, 81), no studies have addressed how these proteins may exert their function. To gain insight into the function of C. jejuni VirK, we examined its subcellular localization. First, we constructed a FLAG epitope-tagged version of VirK and examined its ability to complement the virK mutation. The VirK-FLAG protein was able to complement the intracellular survival defect of this mutant strain, indicating that the presence of the FLAG epitope tag did not affect the function of VirK or, presumably, its subcellular localization (Fig. (Fig.2A).2A). We then carried out subcellular fractionation of C. jejuni and probed for the presence of VirK-FLAG in the different fractions by Western immunoblotting. VirK-FLAG was equally present in the cytoplasmic and membrane fractions (Fig. (Fig.3)3) but was absent from the culture supernatant (data not shown). Interestingly, in the cytoplasmic fraction, VirK migrated as a distinct doublet with an estimated difference of ~1.4 kDa (Fig. (Fig.3).3). We hypothesize that this difference is due to the presence of an alternative start codon in the virK gene, which would result in a longer version of this protein (see Fig. S2 in the supplemental material). The significance of the alternative start codon for virK or the extra amino acids is unknown. To probe the nature of the association of VirK with the membrane, the bacterial membrane fraction was subjected to different extraction procedures. Washing with PBS resulted in no extraction of VirK from the membrane, while treatment with high-salt and high-pH solutions resulted in only partial extraction (Fig. (Fig.3).3). Only treatment with 0.5% Sarkosyl completely solubilized VirK. Taken together, these results indicate that VirK is peripherally (although tightly) associated to the cytoplasmic side of the plasma membrane.
The finding that VirK is cytoplasmic or peripherally associated to the cytoplasmic side of the inner membrane suggested that it exerts its effect on intracellular survival indirectly, perhaps by modulating the composition of the bacterial envelope. To test this hypothesis, we compared the total protein profiles of bacterial envelopes isolated from wild-type C. jejuni 81-176 and its isogenic virK mutant strain by using differential fluorescence 2D gel electrophoresis proteome profiling. No significant differences in the protein profiles of the envelope fractions of these two bacterial strains were detected (see Fig. S3A and B in the supplemental material), indicating that the absence of VirK does not significantly alter the protein composition of the bacterial envelope. Consistent with this hypothesis, no difference in the surface hydrophobicity profiles of the wild type and the virK mutant derivative was detected by the salting-out method as described in Materials and Methods (data not shown).
We compared the ability of the C. jejuni virK mutant strain to grow in liquid culture with that of the wild type, both when grown separately and in mixed cultures. As shown in Fig. S4 in the supplemental material, the growth properties of C. jejuni virK were indistinguishable from those of the wild type or the complemented mutant. These results indicate that VirK is not required for C. jejuni growth in vitro.
Previous studies demonstrated that VirK plays an important role in systemic S. Typhimurium infection in mice (20, 47). We therefore tested whether the C. jejuni homolog played a similar role during infection by using a recently developed mouse colonization model (80). Equal numbers of the wild-type C. jejuni strain (or the complemented virK mutant strain) and the virK mutant strain were simultaneously administered orally or intraperitoneally to MyD88/Nramp1-deficient mice. The presence of different C. jejuni strains in the feces of infected animals was monitored over time as described in Materials and Methods. When bacteria were administered intraperitoneally, the virK mutant strain was cleared rapidly, while the wild type or the complemented strain was readily recovered from the feces (Fig. 4A and C). No CFU from the mutant strain was detected in the feces 2 weeks postinfection when the mutant strain was competed against the complemented strain (Fig. (Fig.4C).4C). When the virK mutant strain was competed against the wild type, only one mouse out of six shed a small number of CFU of the mutant 6 weeks postinfection, while the levels of the wild-type strain remained high (Fig. (Fig.4A).4A). Six weeks after infection, low levels of C. jejuni virK were recovered from tissues of only one infected animal in each group (Fig. 4B and D). However, those numbers were still significantly lower than those of the wild-type strain. When administered orally, the numbers of CFU of the virK mutant strain in the feces of infected animals were also lower than those of the complemented mutant strain (Fig. (Fig.5A).5A). However, the intestinal colonization defect was not as pronounced as when the mutant was administered intraperitoneally and did not reach statistical significance. The numbers of CFU of the virK mutant strain in systemic tissues after oral inoculation were also significantly lower than those of the complemented mutant strain (Fig. (Fig.5B).5B). In summary, our results demonstrate that C. jejuni VirK plays an important role in colonization and systemic infection in a mouse model of infection.
The resistance to complement-mediated bactericidal activity in serum is critical for a number of gram-negative pathogens to establish successful systemic infections in their hosts (3, 18, 37, 66, 71). Previous observations suggested that C. jejuni strains are susceptible to the killing action of complement (11), which may be related to the usually transient nature of C. jejuni-associated bacteremia (26, 46). However, some isolates can cause systemic infection in immunocompromised hosts, suggesting that these strains may have increased serum resistance (10, 35). Since the virK mutant showed a severe defect in its ability to colonize, spread, and survive in systemic tissues, we examined whether this mutant had increased sensitivity to serum. We compared the C. jejuni virK mutant with its complemented derivative for the ability to survive in the presence of 2% and 10% normal rabbit serum (see Materials and Methods for experimental details). While the percentage of bacteria that survived treatment with 2% serum for 60 min remained relatively high, the bacteria treated with 10% serum declined sharply in numbers after 30 min, and by 60 min both the mutant and complemented strains exhibited 20-fold reductions in CFU (see Fig. S5 in the supplemental material). These data are in agreement with the findings of a previous report, which tested the sensitivities of different C. jejuni isolates to 10% human serum (11), and they indicate that the absence of VirK does not result in increased serum sensitivity.
Antimicrobial peptides are an essential component of host innate immunity (14, 28). Resistance to killing by antimicrobial peptides is an essential virulence attribute of many pathogenic bacteria, and of enteropathogens in particular. It has been shown previously that the S. Typhimurium VirK protein is required for resistance to antimicrobial peptides (13, 20). Therefore we examined whether the C. jejuni VirK homolog would have a similar role. First, we tested the sensitivity of the C. jejuni VirK homolog to polymyxin B, since this antibiotic is known to bind to bacterial lipopolysaccharide and to mimic the action of antimicrobial peptides (7, 20, 24, 73). Wild-type C. jejuni, the virK mutant, and its complemented derivative were exposed in vitro to different concentrations of polymyxin B, and the numbers of surviving bacteria 45 min after treatment were determined by enumerating the CFU. As shown in Fig. Fig.6A,6A, the virK mutant strain was significantly more sensitive to polymyxin B treatment than the wild type strain.
We then tested the sensitivity of the virK strain to the antimicrobial peptide Magainin-1, which has been shown previously to be a good surrogate for testing overall sensitivity to antimicrobial peptides (72). Consistent with the polymyxin B resistance results, the C. jejuni virK mutant strain showed very significantly increased (>100-fold) sensitivity to this antimicrobial peptide relative to that of either the wild type or the complemented mutant (Fig. (Fig.6B).6B). In summary, these results showed that VirK is required for resistance to antimicrobial peptides, which may explain its important role in C. jejuni mouse colonization.
One of the surprising findings that emerged from the analyses of genome sequences from several C. jejuni isolates is the paucity of putative homologs of virulence factors from other pathogenic bacteria (22, 32, 62, 65). Here we have described the identification of a C. jejuni 81-176 gene, CJJ81176_1087, which encodes a homolog of the VirK family of virulence factors. Although little is known about the specific mechanisms by which the VirK homologs in various pathogens exert their function, it is intriguing that VirK family members are encoded by many pathogenic bacteria. For example, in addition to the VirK protein of Shigella spp., the first family member identified, homologs of this protein have been found in Salmonella enterica (also known as VirK) and Pasteurella multocida (LapB) (20, 29, 39, 53, 56).
We found that the C. jejuni virK mutant has a pronounced defect in its ability to colonize the intestine and deeper tissue of a mouse after either oral or systemic administration. This mutant strain also exhibited a reproducible defect in its ability to survive upon internalization into epithelial cells. However, we do not believe that the severe colonization defect observed in a mouse model of infection is a result of the intracellular survival defect of this strain. Neither do we believe that the observed reduction in the intracellular survival of the mutant strain is significant enough to suggest that this gene product is directly involved in mediating this process. Rather, we believe that the intracellular survival defect may be an indirect result of some changes in the composition of the C. jejuni envelope that may indirectly affect this phenotype. This is consistent with the observation that VirK is located within the bacterial cytoplasm, peripherally associated to the inner membrane, a location that makes this protein unlikely to mediate intracellular survival directly.
What could account for the profound defect in mouse colonization exhibited by the C. jejuni virK mutant strain? We hypothesize that the colonization defect is due largely to the increased susceptibility of the mutant strain to antimicrobial peptides. Indeed, antimicrobial peptides are one of the most important lines of defense against bacterial pathogens, and the virK mutant strain exhibited a >100-fold increase in susceptibility to at least one antimicrobial peptide in vitro. Consistent with this hypothesis, a virK mutant of S. Typhimurium, which is also defective in mouse colonization, exhibited significantly increased susceptibility to antimicrobial peptides (13, 20). This observation also suggests that these two protein homologs may exert their effect in a similar manner.
What could be the role of VirK in conferring resistance to antimicrobial peptides? The intracellular location of VirK suggests that it is unlikely to mediate antimicrobial peptide resistance directly. Rather, it is possible that VirK may modulate the composition of the bacterial envelope, which is the target of the antimicrobial peptide activity. We compared the protein profiles of the bacterial envelopes of wild-type C. jejuni and the isogenic virK mutant strain by differential fluorescence gel electrophoresis proteome profiling. However, despite the relative sensitivity of this technique, we found no significant differences between the envelope protein profiles of these strains. Although it is possible that differences could have escaped detection, we believe that this is unlikely. Rather, it is possible that VirK may be responsible for modulating the composition of other surface structures, such as the LOS or the capsule polysaccharide, or the lipid composition of the inner membrane itself. We did examine the LOS of both strains and found no gross differences (see Fig. S6 in the supplemental material), but it is possible that differences may have escaped detection by the rather insensitive approach we used in the analysis.
Our studies do not shed light on the actual mechanism by which VirK may exert its function. It should be noted that several VirK homologs are missannotated in the available databases as either putative transporters, ATPases, or ATP-binding cassette proteins. Furthermore, we have detected no predicted transmembrane domains in any of the many DUF535 homologs that we have analyzed, despite the fact that some of them are annotated as membrane proteins. Although we are not certain where the annotation errors come from, it is likely that they are due to automatic annotation, which has not been followed by careful curation. With the possible exception of a homolog in Pseudomonas entomophila, all DUF535 proteins, like VirK, are composed of a single domain (i.e., the DUF535 domain). The exception, a homolog in which the DUF535 domain is linked to a putative glycosyl transferase, is most likely a sequencing error, since in highly related bacteria, DUF535 and the glycosyl transferases are encoded by separate, immediately adjacent genes. Since it is unlikely that VirK has enzymatic activity, we hypothesize that it may modulate the activities of other bacterial proteins, perhaps by locating them to an appropriate site, quite possibly to the inner leaflet of the bacterial inner membrane, where VirK is primarily localized. In fact, it has been shown previously that the Shigella sp. VirK protein influences the ability of this bacterium to spread intracellularly (56). It is believed that VirK exerts this effect by regulating the stability and/or activity of IcsP, a protease required for the posttranslational regulation of IcsA (81). The latter is a Shigella surface protein that catalyzes actin nucleation at one of its poles (8, 48). C. jejuni 81-176 encodes a putative peptidase immediately upstream of CJJ81176_1087. This protease, CJJ81176_1086, belongs to the M50 family of metalloproteases, involved in regulated intramembrane proteolysis (51). Interestingly, a member of this family in Mycobacterium tuberculosis, Rv2860c, posttranslationally regulates a factor involved in modulating the cell envelope composition (50). Another member of this family, YaeL, degrades the virulence regulator TcpP in Vibrio cholerae under certain environmental conditions (52). Nevertheless, it is unknown whether C. jejuni VirK regulates the activity of the CJJ81176_1086 peptidase. Addressing the potential relationship between VirK and CJJ81176_1086 will require further experimentation, which is beyond the scope of this study.
In summary, we have characterized a C. jejuni protein that belongs to the VirK family of virulence factors and have determined that it plays an important role in virulence in a mouse model of infection. This protein, therefore, joins CDT as one of the rather limited number of virulence factors in C. jejuni for which homologs exist in other pathogenic bacteria.
We thank María Lara-Tejero for providing the Myd88−/− mice and Heather Scobie for useful discussion and critical review of the manuscript.
This work was supported by a grant from the Ellison Medical Foundation to J.E.G.
Editor: V. J. DiRita
Published ahead of print on 21 September 2009.
†Supplemental material for this article may be found at http://iai.asm.org/.