Campylobacter jejuni survival within human ... - Wiley Online Library

Molecular Microbiology (2011) 80(5), 1296–1312 䊏

doi:10.1111/j.1365-2958.2011.07645.x First published online 11 April 2011

Campylobacter jejuni survival within human epithelial cells is enhanced by the secreted protein CiaI mmi_7645 1296..1312

Daelynn R. Buelow, Jeffrey E. Christensen, Jason M. Neal-McKinney and Michael E. Konkel* School of Molecular Biosciences, College of Veterinary Medicine, Washington State University, Pullman, Washington, USA, 99164-7520.

Summary Although it is known that Campylobacter jejuni invade the cells that line the human intestinal tract, the bacterial proteins that enable this pathogen to survive within Campylobacter-containing vacuoles (CCV) have not been identified. Here, we describe the identification and characterization of a protein that we termed CiaI for Campylobacter invasion antigen involved in intracellular survival. We show that CiaI harbours an amino-terminal type III secretion sequence and is secreted from C. jejuni through the flagellar type III secretion system. In addition, the ciaI mutant was impaired in intracellular survival when compared with a wild-type strain, as judged by the gentamicin-protection assay. Fluorescence microscopy examination of epithelial cells infected with the C. jejuni ciaI mutant revealed that the CCV were more frequently co-localized with Cathepsin D (a lysosomal marker) than the CCV in cells infected with a C. jejuni wild-type strain. Ectopic expression of CiaI-GFP in epithelial cells yielded a punctate phenotype not observed with the other C. jejuni genes, and this phenotype was abolished by mutation of a dileucine motif located in the carboxy-terminus of the protein. Based on the data, we conclude that CiaI contributes to the ability of C. jejuni to survive within epithelial cells.

Introduction Campylobacter jejuni is recognized as one of the leading bacterial causes of gastroenteritis (De Melo et al., 1989; Konkel and Cieplak, 1992; Coker et al., 2002; RuizPalacios, 2007). The ability of C. jejuni to cause disease is a complex multifactorial process (Larson et al., 2008). Motility, adherence, invasion, protein secretion, intracelluAccepted 18 March, 2011. *For correspondence. E-mail konkel@ wsu.edu; Tel. (+1) 509 335 5039; Fax (+1) 509 335 4159.

© 2011 Blackwell Publishing Ltd

lar survival and toxin production are some of the factors that contribute to the pathogenicity of this bacterium. A serious complication of C. jejuni infection is the development of Guillain–Barré syndrome, an autoimmune disease affecting the peripheral nervous system. Although C. jejuni has been recognized as a pathogenic organism since the early 1970s, its virulence strategies and determinants are ill-defined. Bacterial motility is pivotal throughout C. jejuni infection. C. jejuni targets the jejunum/ileum of the small intestine where it resists the peristaltic forces and flushing of the small intestine by binding to extracellular matrix components via the CadF and FlpA adhesins (Konkel et al., 1997; 2005; 2010; Monteville et al., 2003). In response to the host environment, C. jejuni alters its gene expression profile resulting in the synthesis of the Campylobacter invasion antigens (Cia) (Konkel et al., 1999; 2001; RiveraAmill et al., 2001). The Cia proteins are exported from the bacterium’s flagellar type III secretion system (T3SS), and their delivery to the host cell promotes maximal cell invasion (Konkel et al., 2004). Internalization of the bacteria requires the stimulation of host cell signalling (i.e. Rho GTPase activation), as well as rearrangement of the host cell cytoskeleton (Monteville et al., 2003; KrauseGruszczynska et al., 2007). Following internalization, C. jejuni are able to survive and multiply within epithelial cells (De Melo et al., 1989; Konkel et al., 1992a). In the past decade, it has been found that C. jejuni relies on its flagella during the course of infection for both motility and secretion of the Cia proteins (Konkel et al., 1999; 2004; Rivera-Amill and Konkel, 1999). In vivo studies indicate that secretion of the Cia proteins contributes to the severity of C. jejuni-mediated enteritis. More specifically, piglets inoculated with a C. jejuni strain deficient in Cia secretion exhibited increased time of onset and reduced disease severity compared with piglets inoculated with a wild-type strain of C. jejuni (Konkel et al., 2001). Although the Cia proteins play a significant role in the development of C. jejuni-mediated enteritis, the majority of the secreted proteins have yet to be identified. Once in a human host, C. jejuni can invade the cells lining the intestinal tract. Although it is not understood how C. jejuni are able to survive and multiple within epithelial cells, progress has been made to characterize Campylobacter-containing vacuoles (CCV) (De Melo

C. jejuni survival within epithelial cells 1297

et al., 1989; Konkel et al., 1992a; Watson and Galan, 2008). Early studies suggested that the CCV diverged from the normal endocytic pathway within epithelial cells based on the observation that they became clustered around the cell nuclei (Konkel et al., 1992a). More recently, researchers have found that immediately following C. jejuni cell uptake, the CCV displays markers of early endosomes, such as early endosomal antigen-1 (EEA-1). The CCV then recruit the GTPases, Rab5 and Rab7, and lysosomalassociated membrane protein 1 (LAMP-1), sequentially. This recruitment of cellular markers is similar to the maturation of early endosomes to late endosomes. Furthermore, CCV were rarely seen to associate with the lysosomal marker Cathepsin B, indicating that the CCV diverge from the classical endocytic pathway. How C. jejuni alter the properties of the vacuole to prevent fusion with the lysosome is not known. We hypothesized that C. jejuni secretes one or more Cia proteins to modify the CCV and survive within the host cell. Here we identify the first C. jejuni secreted protein involved in intracellular survival. We show that Cj1450, which we termed Campylobacter invasion antigen involved in Intracellular survival (CiaI), is secreted via the flagellar T3SS and prevents the delivery of CCV to lysosomes.

Results Salient features of the Cj1450 ORF and the protein it encodes Culturing C. jejuni in the presence of deoxycholate (DOC) triggers the expression of the cia genes, thus priming the bacterium for host cell invasion (Malik-Kale et al., 2008). To identify new Cia proteins, we analysed genes upregulated by DOC with annotated functions that could be involved in C. jejuni pathogenesis. One DOC regulated gene identified was Cj1450, which was capable of encoding a protein with a calculated molecular mass of 21 385.94 Daltons. Cj1450 is annotated as a putative ATP/GTP binding protein (Parkhill et al., 2000), as it contains the ATP/GTP binding site motif-A (Residues 36–43, ANNEEGKT, PROSITE # PS00017, consensus [AG]-x(4)-G-K-[ST]). This motif is found in a variety of eukaryotic proteins, including the Ras family of GTP binding proteins. Although T3S effectors contain an N-terminal sequence that facilitates their secretion from the T3SS (Sory et al., 1995; Schesser et al., 1996), a consensus sequence that directs export of a protein from the T3SS has not been identified. Given that the expression of Cj1450 is induced by DOC and its deduced amino acid sequence contains a putative ATP/GTP binding motif, we first explored the possibility that Cj1450 harbours an amino-terminal export

sequence using a Yersinia enterocolitica phospholipase indicator agar (PLA) assay described previously (Christensen et al., 2009). Briefly, the first 108 nucleotides of Cj1450 were cloned into pCPS50, which contains the truncated phospholipase YplA reporter, and the recombinant vector was conjugated into a Y. enterocolitica yplAB mutant strain. A Y. enterocolitica yplAB mutant harbouring the Cj1450 : YplA fusion showed a zone of secretion on PLA plates comparable with that of the full-length YplA protein (both 1.7 mm), while the T3S sequence negative control CysM and the empty vector pCPS50 showed no zone of secretion (Fig. S1A). Additional experiments confirmed that the Cj1450 : YplA fusion protein was secreted from Y. enterocolitica (Fig. S1B–E). Based on the data, we concluded that Cj1450 harbours a T3S amino-terminal export sequence. Cj1450 is a Cia Campylobacter jejuni secretes the Cia proteins from its flagellum when cultured with epithelial cells or in medium supplemented with serum (Konkel et al., 1999; RiveraAmill and Konkel, 1999). To determine if Cj1450 is secreted from the Campylobacter flagellar T3SS, we generated a Cj1450::tetO mutant as well as a Cj1450::cat mutant. The phenotype displayed by the Cj1450 tetracycline and chloramphenicol resistant mutants were indistinguishable from each other in all assays performed. Thus, we show the result for the Cj1450::tetO mutant or the Cj1450::cat mutant for each experiment. The secretion of the Cia proteins from the C. jejuni wild-type strain and Cj1450 mutant was determined by [35S]-methionine labelling assays coupled with autoradiography. In contrast to the C. jejuni wild-type strain, a protein with a Mr of 21 kDa was absent in the secretion profile of the Cj1450::tetO mutant (Fig. 1A, compare lanes 1 and 3). To determine if the Cj1450 protein is indeed secreted from C. jejuni, the wild-type strain and Cj1450::tetO mutant were transformed with pRY111 containing a Cj1450 FLAG-tagged translational fusion (c. molecular mass of 22 398.92 Daltons). The immunoblots probed with the a-FLAG antibody revealed a 22 kDa reactive band in the secretion profiles of the wildtype strain and Cj1450 mutant both harbouring the Cj1450 FLAG-tagged pRY111vector (Fig. 1B, lanes 2 and 4). To ensure that the detection of Cj1450 FLAG-tagged protein in the supernatant was not due to bacterial cell lysis, supernatants were probed for CysM, a cytoplasmic protein. CysM was not detected in the supernatants (Fig. 2C); however, CysM was detected in the WCLs (Fig. 1F). To determine if the secretion of Cj1450 was dependent on the flagellum and the presence of serum, secretion of Cj1450 was assessed in the flgBC flagellar basal body mutant and in the absence of fetal bovine

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 80, 1296–1312

1298 D. R. Buelow, J. E. Christensen, J. M. Neal-McKinney and M. E. Konkel 䊏

Fig. 1. Secretion of Cj1450 and Cj1450 FLAG-tagged protein from the C. jejuni flagellar T3SS. Secretions of Campylobacter invasion antigens (Cia) were induced by 1% FBS after labelling the bacterial proteins with [35S]-methionine. Supernatants (Panels A, B, and C) and whole cell lysates (WCLs; Panels D, E and F) of the C. jejuni F38011 wild-type strain, Cj1450 mutant and flgBC basal body mutant were subjected to autoradiography and immunoblot analysis. A. Autoradiograph of supernatants. The Cj1450 protein (21.4 kDa) is indicated with an open arrow head (䉰) and Cj1450 FLAG-tagged protein (22.4 kDa) is indicated with a closed arrow head (䉳). B. Immunoblot of supernatants probed with FLAG antibody. C. Immunoblot of supernatants probed with CysM antibody. D. Autoradiograph of WCLs. E. Immunoblot of WCLs probed with FLAG antibody. F. Immunoblot of WCLs probed with CysM antibody.

serum (FBS) (Konkel et al., 1999; 2004; Rivera-Amill and Konkel, 1999). As expected, the flgBC mutant was secretion negative, as judged by both [35S]-methionine labelling assays coupled with autoradiography (Fig. 1A, lanes 5 and 6) and immunoblot analysis using a FLAG-specific antibody (Fig. 1B, lane 6). Importantly, the synthesis of the Cj1450 FLAG-tagged protein in the flgBC mutant was confirmed by immunoblot analysis of WCLs using the FLAG-specific antibody (Fig. 1E, lane 6). Consistent with previous work, the secretion of the Cia proteins including Cj1450 was found to be dependent on the addition of serum to the labelling medium, as the Cia proteins were absent in the supernatants from the C. jejuni wild-type strain when the labelling assay was performed in medium without FBS (Fig. 1A, lane 7). To ensure that the absence of the Cia proteins in the supernatants collected from C. jejuni labelled in medium without FBS was not due to reduced levels of metabolic activity, WCLs were prepared following the labelling assay. Autoradiography of the WCLs revealed that all of the bacteria used in the assay were metabolically active (Fig. 1D). Collectively, these

data demonstrate that Cj1450 is secreted from the C. jejuni flagellar export apparatus. Cj1450 promotes C. jejuni intracellular survival Binding, internalization and survival assays were performed with INT 407 cells using the C. jejuni wild-type strain and Cj1450::cat mutant. The binding and internalization data shown in Table 1 and the survival data shown in Fig. 2 are from the same experiment. We observed minor differences in the ability of the wild-type strain and Cj1450::cat mutant to bind to and invade the INT 407 cells (Table 1). However, there were significant differences in the survival of the wild-type strain and Cj1450 mutant in INT 407 cells at 24 and 48 h, as judged by immunofluorescence microscopy and the gentamicin-protection assay (Fig. 2). Consistent with a previous report (Konkel et al., 1992a), the wild-type strain showed a decline in the number of intracellular bacteria between 6 and 24 h, followed by an increase in the number of intracellular bacteria after 24 h. In contrast, there was a 4- and 7-fold



to that of the wild-type strain (not shown). Based on these data, we concluded that Cj1450 contributes to C. jejuni intracellular survival. We named the protein encoded by Cj1450 to be Campylobacter invasion antigen involved in Intracellular survival (CiaI).

CCV containing the ciaI mutant associate with both LAMP-1 and Cathepsin D

Fig. 2. Cj1450 promotes the intracellular survival of the C. jejuni wild-type strain. A C. jejuni Cj1450 mutant is reduced in survival within INT 407 cells when compared with a wild-type strain. A. INT 407 cells inoculated with the C. jejuni wild-type strain and Cj1450 mutant were analysed by indirect immunofluorescence 6, 24 and 48 h post infection. Nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole) and Campylobacter were labelled with Texas-Red. B. Intracellular survival of the C. jejuni wild-type strain (䉬) and Cj1450 mutant ( ). The number of bacteria recovered from the INT 407 cells was determined over time using a gentamicin survival assay. The assay was performed in quadruplicate and repeated three times. A representative experiment is shown. Error bars represent standard deviations. A greater number of bacteria were recovered from cells inoculated with the C. jejuni wild-type strain when compared with the Cj1450 mutant (P < 0.01) as determined by the unpaired Student’s t-tests (*).

reduction in the number of intracellular Cj1450::cat mutant bacteria when compared with the wild-type strain at 24 and 48 h respectively (Fig. 2, Panel B). The growth rate, motility and binding of the Cj1450::cat mutant was equal

We wanted to examine the localization pattern of CiaI when ectopically expressed as a CiaI-GFP fusion protein in human epithelial cells. We chose to use HeLa cells for these experiments because we had been able to obtain reasonable transfection efficiencies with this cell line. However, we first needed to make sure that C. jejuni could bind to and invade HeLa cells. Therefore, we performed binding and internalization assays with INT 407 and HeLa cells, and found that C. jejuni bound to both cell types with equally high efficiency (Fig. S2). This was of little surprise because American Type Culture Collection notes the INT 407 cell line was derived via HeLa contamination, and because it is well known that C. jejuni bind to and invade cell lines of human origin with equally high efficiencies (Konkel et al., 1992b). We then compared the trafficking of the C. jejuni wild-type strain and ciaI mutant in HeLa cells and performed the ectopic expression experiments (see the next section of the paper for the results of the ectopic expression studies). Lysosomal-associated membrane protein 1 and Cathepsin D (a lysosomal aspartyl protease) localization was assessed in HeLa cells infected with the C. jejuni wild-type strain and ciaI mutant at 18 and 24 h post infection. Examination of more than 50 CCV revealed that the wild-type strain and ciaI mutant CCV were labelled with LAMP-1 antibody (Fig. 3, Panels A and B). This finding suggested that CiaI was not involved in the acquisition of LAMP-1 to the CCV. We then assessed the CCV for localization with Cathepsin D. At 18 h, fewer than 5% of the CCV containing the wild-type or ciaI mutant were stained with the Cathepsin D antibody (Fig. 4, Panels A and B). However, by 48 h, greater than 80% of the CCV harbouring the ciaI mutant were associated with Cathepsin D whereas fewer than 20% of the wild-type CCV were associated with Cathepsin D. The differences observed between the wild-type and

Table 1. Adherence and invasion of the C. jejuni wild-type and Cj1450 mutant. Strain

Adherencea

Invasiona

I/Ab

Wild-type strainc Cj1450::cat mutantc

(6.7 ⫾ 1.1) ¥ 106 (6.6 ⫾ 0.4) ¥ 106

(2.0 ⫾ 0.6) ¥ 105 (1.2 ⫾ 0.2) ¥ 105

3.1% 1.8%

a. Number of viable bacteria. b. Percentage of internalized bacteria relative to adherent bacteria. c. The INT 407 cells were inoculated with 3 ¥ 107 cfu of the C. jejuni wild-type strain and 2.6 ¥ 107 cfu of the C. jejuni 1450 mutant.



Fig. 3. CCV containing the C. jejuni wild-type strain and ciaI mutant are associated with the LAMP-1 marker. HeLa cells were infected with the C. jejuni wild-type strain and ciaI mutant. At 18 and 24 h post inoculation of HeLa cells, C. jejuni and lysosomal markers were labelled with Texas-Red and FITC, respectively, and visualized. White circles indicate areas where both C. jejuni and the lysosomal marker LAMP-1 co-localized. A minimum of 50 CCV were analysed for co-localization with LAMP-1 at 18 and 24 h, and the values obtained are reported as a percentage in the merged image. A. HeLa cells infected with the C. jejuni wild-type strain. B. HeLa cells infected with the ciaI mutant.

ciaI mutant with respect to the staining of the CCV with Cathepsin D suggested that CiaI may play a role in retarding the CCV from fusing with lysosomes. A C-terminal dileucine motif mediates localization of ectopically expressed CiaI Bacterial effector proteins often localize to specific cellular structures and modulate host cell function (Brumell et al., 2003; Weber et al., 2006; de Felipe et al., 2008; Voth et al., 2009). As we observed that CiaI influences the survival of C. jejuni within epithelial cells, presumably by preventing the CCV from fusing with lysosomes, we hypothesized that ectopic expression of ciaI would result in localization to vacuoles. We reasoned that the examination of CiaI in isolation could provide insight not gained from comparing the C. jejuni wild-type strain and ciaI mutant based on the possibility that: (i) another protein may have a similar function as CiaI in C. jejuni and/or (ii) the high likelihood that other proteins contribute to C. jejuni survival within cells. For the ectopic expression studies, we cloned various genes or mutated genes (e.g. ciaI, ciaI gene variants, Cj0040 and Cj0787) into the

pEGFP-N1 vector. Noteworthy is that the expression of all C. jejuni genes was driven by the immediate early promoter of human cytomegalovirus (PCMV IE) contained within the vector. HeLa cells were transfected with 400 ng of the various vectors and analysed for distinct GFP patterning. HeLa cells transfected with a plasmid encoding the wildtype CiaI-GFP showed a punctate pattern (Fig. 5). In contrast, cells transfected with the empty vector, which contained only GFP, or the C. jejuni Cj0040 and Cj0787 genes, which are upregulated in the presence of DOC, showed diffuse fluorescence with no specific cellular localization (Fig. 5). Even though ciaI and the other C. jejuni genes were expressed in HeLa cells from PCMV IE, it appeared from the immunofluorescence microscopy examination of the HeLa cells that the transfection efficiency varied significantly among the constructs. For example, the transfection efficiency for the HeLa cells with the pEGFP-C1 vector without an insert was approximately 45%, whereas the transfection efficiencies for the pEGFP-N1 vector containing the Cj0040 and ciaI genes were 28% and 20% respectively. Regardless, these experiments demonstrated ectopic expression of ciaI in epithelial cells yields a unique phenotype.



Fig. 4. CCV containing the C. jejuni ciaI deletion mutant are associated with Cathepsin D at 24 h. HeLa cells were inoculated with the C. jejuni wild-type strain and ciaI mutant. At 18 and 24 h, C. jejuni and Cathepsin D were labelled with Texas-Red and FITC, respectively, and visualized. White circles indicate areas were both C. jejuni and Cathepsin D co-localized. A minimum of 50 CCV were analysed for co-localization with Cathepsin D at 18 and 24 h, and the values are reported as a percentage in the merged image. A greater percentage of CCV were co-localized with Cathepsin D in cells inoculated with the C. jejuni ciaI mutant than in cells inoculated with the C. jejuni wild-type strain (P < 0.05), as determined by the unpaired Student’s t-tests (*). A. HeLa cells infected with the C. jejuni wild-type strain. B. HeLa cells infected with the ciaI deletion mutant.

Fig. 5. Cj1450-GFP transfected cells show a distinct punctate phenotype. HeLa cells were transfected as outlined in ‘Experimental procedures’ with the vectors harbouring GFP fusions to CiaI (pEGFP-N1-CiaI-GFP), Cj0040 (pEGFP-N1-Cj0040-GFP), Cj0787 (pEGFP-N1-Cj0787-GFP) or with GFP (pEGFP-C1) alone. CiaI-GFP shows a distinct punctate phenotype compared with cells transfected with GFP, Cj0040-GFP and Cj0787-GFP, which all show diffuse GFP fluorescence.



Fig. 6. CiaI-GFP is associated loosely with LAMP-1. HeLa cells were transfected with CiaI-GFP (green) fusion vector. Cellular markers of the endocytic compartments were labelled with Rhodamine (red). Cellular staining of CiaI-GFP transfected cells with a-EEA1, a-Rab7, a-LAMP-1 and a-Cathepsin D were all captured in the same field of view. White offset boxes are enlarged areas of the small white boxes in LAMP-1 ( ) and Cathepsin D (䉬) merged images. For each experiment at least 1000 CiaI-GFP punctate spots were analysed for co-localization with cellular markers. A representative image is shown for each cellular marker.

To determine if CiaI is targeted to vesicles within the endocytic pathway, we investigated the localization of CiaI-GFP with EEA-1, Rab7, LAMP-1 and Cathepsin D by immunofluorescence microscopy. EEA-1 is a marker for early endosomes, while Rab7 and LAMP-1 are markers of late endosomal vesicles, and Cathepsin D is a lysosomal marker. By indirect immunofluorescence, we found 15.4% and 6.4% of CiaI-GFP vesicles were labelled with LAMP-1 and Rab7 respectively (Fig. 6). There was relatively little co-localization of CiaI-GFP with EEA-1 (1%) or Cathepsin

D (2.6%) (Fig. 6). These data suggest that CiaI is targeted to vesicles with characteristics of late endosomes. Because CiaI appeared to be targeted to vesicle structures, we examined whether this protein contained an endosomal targeting/retaining motif. A series of CiaI C-terminal truncations fused to GFP were generated (Fig. 7). Transfected cells containing CiaI(1–165) fused to GFP gave the characteristic punctate phenotype observed with the full-length fusion protein (Fig. 7A and B), whereas CiaI(1–145)-GFP and CiaI(1–100)-GFP lost



Fig. 7. The C-terminus of CiaI mediates a distinct punctate phenotype. HeLa cells transfected with different CiaI truncations fused to GFP were visualized by immunofluorescence for cellular phenotypes. HeLa cells transfected with the full-length CiaI-GFP (A) and CiaI(1–165)-GFP (B) show a distinct punctate phenotype, while CiaI(1–145)-GFP (C), CiaI(1–100)-GFP (D) and CiaI(L153,154A)-GFP (E) showed a diffuse phenotype with a greater concentration of protein around the nucleus compared with GFP alone (F).

the characteristic punctate phenotype (Fig. 7C and D). This finding revealed that the CiaI targeting motif was localized in the region of 145 to 165. In silico analysis revealed that CiaI contained a stretch of amino acids from position 149 to 154 (DSKKLL) that matched a dileucine motif consensus sequence [DE]XXXL[LI] (Letourneur and Klausner, 1992; Pond et al., 1995). The dileucine motif (DXXXLL signal) is recognized by Golgi-localizing, g-adaptin ear homology domain, ADP-ribosylation factor binding proteins (GGAs), and is responsible for transport from the trans-Golgi network to endosomes (Takatsu et al., 2001; Kelly et al., 2008; Braulke and Bonifacino, 2009). To test whether the dileucine motif at position 149 to 154 in CiaI is responsible for its endosomal retention, we mutated the two Leu at positions 153 and 154 to Ala. When CiaI(L153,154A)-GFP, a modified full-length CiaI without the dileucine motif, was transfected into HeLa

cells, the characteristic punctate phenotype of CiaI was abolished (Fig. 7E). Moreover, CiaI(L153,154A)-GFP transfected cells resembled the fusion constructs that retained less than the first 145 amino acids. Experiments were also performed to determine if the amount of GFP fusion protein in the population of cells correlated with punctate or non-punctate phenotypes (i.e. a greater amount of protein within the cell could be responsible for the appearance of the vesicles). Given the differences observed in the transfection efficiencies with the various vectors, it was expected that the HeLa cells transfected with the various constructs exhibited differences in the amount of the GFP fusion proteins (Fig. 8A). The greatest amount of GFP was observed in HeLa cells transfected with the GFP vector without an insert. In comparison, a reduction was observed in the amount of Cj0040-GFP and Cj0787-GFP. The least amount of GFP was detected in cells trans-



Fig. 8. Immunoblot analysis of transfected HeLa cells harbouring either GFP alone or different GFP fusion constructs. The upper panels are immunoblots of whole cell lysates (WCLs) probed with a-EGFP antibody. The lower panels are immunoblots of WCLs probed with a-actin antibody. A. WCLs of HeLa cells transfected with GFP alone (lane 1), CiaI-GFP (lane 2), Cj0040-GFP (lane 3), or Cj0787-GFP (lane 4). The predicted molecular mass of GFP is 26.9 kDa, CiaI-GFP is 48.3 kDa, and Cj0040-GFP is 39.6 kDa, while Cj0787-GFP is 36.6 kDa. B. WCLs of HeLa cells transfected with CiaI-GFP (lane 5) or CiaI(L153,154A)-GFP (lane 6). The predicted molecular weight of CiaI-GFP is 48.3 kDa and CiaI(L153,154A)-GFP is 48.2 kDa.

fected with the CiaI-GFP construct. However, the intensity of the GFP band in the CiaI-GFP expressing cells was more than in CiaI(L153,154A)-GFP expressing cells, yet the punctate phenotype was observed only in the CiaI-GFP expressing cells (Fig. 8B). Collectively, these data suggest that the dileucine motif found at position 149 to 154 in CiaI participates in targeting the protein to vesicles.

of the sequence upstream of the ciaI ORF revealed the sequence TAAA n17 CGATTT, which matches the published C. jejuni s28 promoter consensus sequence T[A/T]AA n17 CGAT[T/A]T (Carrillo et al., 2004; Wosten et al., 2008). In addition, inspection of the sequence upstream of the methionine initiation codon of dut revealed the sequence TATAAT, which perfectly matches the sequence of nucleotides found in the C. jejuni s70 promoter consensus sequence TA[T/A]AAT (Wosten et al., 2008). We then generated a mutation in the dut gene and performed RT-PCR analysis to analyse transcript levels of ciaI and dut. The glyA transcript was used as an internal control. A reduction was not observed in the amount of dut transcript in the Cj1450::tetO mutant when compared with the C. jejuni wild-type strain, as judged by RT-PCR analysis (Fig. S3B). However, a faint PCR band was amplified using the cDNA from the C. jejuni wild-type strain with a forward primer to ciaI and a reverse primer to dut (Fig. S3C). Thus, the possibility remained that the dut gene is regulated, in part, by the ciaI promoter. As a result of the inability to complement the C. jejuni ciaI mutant and/or prove that the ciaI and dut genes were transcribed independently, we examined HeLa cells transfected with a Dut-GFP expressing vector and the survival of the dut mutant in INT 407 cells. In contrast to CiaI-GFP, HeLa cells transfected with the Dut-GFP showed diffuse fluorescence (Fig. S4). Intracellular survival assays performed with the wild-type strain and dut mutant revealed no differences in survival within INT 407 cells at 24 and 48 h (Fig. S5). Collectively, the data indicate that the phenotype of the ciaI mutant was not due to a polar effect of the ciaI gene knockout on dut gene expression.

Discussion

Analysis of CiaI (Cj1450) and Dut (Cj1451) In preliminary studies, attempts were made to restore the survival of the ciaI mutant in INT 407 cells to that of the wild-type strain by complementation with a shuttle vector containing the CiaI FLAG-tagged protein. However, the attempts to functionally complement the ciaI mutant in trans with a plasmid harbouring the CiaI FLAG-tagged protein were not successful. Based on this result, a suicide vector harbouring the ciaI gene with a FLAG-tag was generated and the gene introduced into the chromosome of the C. jejuni ciaI mutant at a site downstream of the flaB gene. Even though the Cj1450 FLAG-tagged protein was readily detected in the WCL of the C. jejuni ciaI mutant, we could not detect the Cj1450 FLAG-tagged protein in the supernatant fluids (not shown). Thus, the possibility remained that the mutation in ciaI had a polar effect on Cj1451 (dut). To address this concern, we examined the genomic sequence of C. jejuni NCTC 11168 for potential promoter sequences (Fig. S3A). Inspection

CiaI is an effector protein involved in C. jejuni survival within epithelial cells Campylobacter jejuni secretes the Cia proteins via the flagellar export apparatus (Konkel and Cieplak, 1992; Konkel et al., 1999; 2004). Autoradiographs of the secreted proteins show that C. jejuni secrete a minimum of eight Cia proteins. Previous work has also revealed that culturing C. jejuni in the presence of the bile salt DOC triggers the expression of the cia genes, thus priming the bacterium for host cell invasion (Malik-Kale et al., 2008). Based on this information, we attempted to identify a new Cia protein by analysing the genes upregulated by DOC with annotated functions that could be involved in C. jejuni pathogenesis. One gene, which we have now termed ciaI, was identified that could encode a protein with a putative ATP/GTP binding site. In support of the hypothesis that CiaI is a secreted protein, we found that the CiaI protein harbours an amino-terminal T3S sequence (Fig. S1). We



also demonstrated that the native CiaI protein and FLAGtagged CiaI protein were secreted from C. jejuni in a flagellar-dependent manner (Fig. 1). More specifically, a 22.4 kDa band was detected in the supernatant from a C. jejuni wild-type strain and ciaI mutant harbouring the pRY111 vector containing the CiaI FLAG-tagged translational fusion protein, as judged by the [35S]-methionine radiolabelling assay. In addition, we were able to detect the product of the CiaI FLAG-tagged gene expressed from its native promoter in concentrated supernatants using a FLAG antibody (Fig. 1). Finally, a knockout in the ciaI gene resulted in the absence of one protein in the secretion profile when compared with a C. jejuni wild-type strain, and was impaired in its ability to survive within cells when compared with the wild-type strain. While we were able to demonstrate that the CiaI protein is exported from the ciaI mutant, our attempts to restore the survival of the ciaI mutant in INT 407 cells by complementation with the pRY111 vector containing the CiaI FLAG-tagged protein were not successful. The most likely reason for the inability to restore the host survival phenotype of the ciaI mutant to that of the wild-type strain was that the secretion of the CiaI protein was reduced in the complemented mutant versus a wild-type strain (Fig. 1). Thus, a survival assay was performed with the C. jejuni wild-type strain and dut mutant to address whether CiaI or Dut was responsible for the reduction in C. jejuni survival in INT 407 cells. Unlike the C. jejuni ciaI mutant, no difference was observed in the survival of the wild-type strain and dut mutant in INT 407 cells at 24 and 48 h (Fig. S5). In addition, the ectopic expression of ciaI in epithelial cells yielded a punctate phenotype that was not observed for the other C. jejuni genes examined including dut. Finally, a dileucine motif within CiaI was demonstrated to contribute to the punctate phenotype of this protein in cells. The dileucine motif consensus sequence [DE]XXXL[LI] has previously been reported to be responsible for the targeting of proteins to the endocytic pathway (Takatsu et al., 2001; Kelly et al., 2008; Braulke and Bonifacino, 2009). Collectively our findings suggest that the phenotype displayed by the ciaI mutant is solely due to the mutation in the ciaI gene. Based on the data, we have concluded that CiaI contributes to C. jejuni virulence. Potential role of CiaI Ectopic expression of CiaI-GFP results in a punctate phenotype, suggesting that CiaI localizes to specific vesicles within the host cell. This phenotype was not observed with other DOC responsive genes that were expressed ectopically within HeLa cells, or when dut, the gene immediately downstream of ciaI, was expressed ectopically within HeLa cells (Figs 5 and S3). Analysis of the deduced amino acid sequence of CiaI using the Eukaryotic Linear

Motif database (http://elm.eu.org/) (Gould et al., 2010) revealed a stretch of amino acids from position 149 to 154 (DSKKLL) that contained a dileucine motif [DE]XXXL[LI] (Letourneur and Klausner, 1992; Pond et al., 1995). Mutagenesis of the dileucine motif in tyrosinase, a tissuespecific resident protein of the melanosome, ablated late endosomal localization in HeLa cells (Calvo et al., 1999). In addition, site-directed mutagenesis of the chicken invariant chain (Ii) revealed that the dileucine motif is critical in endocytic targeting (Xu et al., 2008). Interestingly, the dileucine motif has been shown to be involved in binding to the cellular adaptor proteins, which are involved in protein sorting into endosomal vesicles (Coleman et al., 2006). When we mutated the two leucine residues to alanine residues within the dileucine motif of CiaI, it resulted in a loss of the punctate phenotype that was observed with the wild-type CiaI-GFP fusion protein (Fig. 7). Examination of the whole cell lysates of HeLa cells transfected with the pEGFP-N1-CiaI(L153,154A) construct revealed less GFP protein compared with the GFP, Cj0040-GFP, Cj0787-GFP and CiaI-GFP expressing cells (Fig. 8). Thus, we concluded that the dileucine motif, and not the level of GFP fusion protein within a cell, plays a role in targeting CiaI to specific vesicles. We hypothesize that cellular recognition of the dileucine motif directs CiaI to vesicles, allowing C. jejuni to subvert normal endocytic trafficking and avoid fusion of the CCV with lysosomes. CiaI may modify the CCV and prevent lysosomal fusion by three mechanisms: (i) selective exclusion, where a host cell protein involved in maturation is never acquired; (ii) retention of specific markers, therefore inhibiting the maturation of the vacuole and/or (iii) selective acquisition, where a specific marker not normally part of the pathway is acquired (Brumell and Scidmore, 2007). Regardless of the mechanism, CiaI appears to contribute to remodelling the CCV to become a unique organelle with altered trafficking. This hypothesis is supported by the observation that CiaIGFP transfected cells show proximal association with Rab7 and LAMP-1. Studies are currently being performed to further dissect the function of the CiaI protein. Perspective Dissecting the virulence mechanisms of C. jejuni is essential to understanding the infection process of this bacterium as it does not use conventional strategies exploited by more intensely studied pathogens. Few proteins, if any, have been implicated to promote the facultative survival of C. jejuni within epithelial cells. While the survival of other pathogens in epithelial cells has been shown to be impacted by reactive oxygen species, surprisingly, catalase (KatA) does not appear to contribute to Campylobacter’s intraepithelial cell survival (Day et al., 2000). The identification of CiaI, which is unique to C. jejuni,



is another example of how this pathogen has devised unique virulence strategies. The current work suggests that the CiaI protein, which is secreted from the flagellar T3SS, is involved in mediating cellular trafficking. Based on the finding that CiaI appears to contribute to the ability of C. jejuni to survive within epithelial cells (Fig. 2), which is regulated by s28 (data not shown) (Carrillo et al., 2004), we propose that both s54- and s28-dependent genes are expressed by C. jejuni within the CCV. Alternative s factors have been shown to contribute to the virulence traits of other pathogens. For example, sE (RpoE) helps Salmonella enterica serovar Typhimurium combat oxidative stress. Thus, inactivation of rpoE contributes to decreased growth and survival of S. enterica within host macrophages (Humphreys et al., 1999; Cano et al., 2001). While additional studies are needed to determine how C. jejuni co-ordinates the expression of the cia genes, the current evidence indicates that these virulence genes are regulated in a s28- and s54-dependent fashion (i.e. s54 for ciaC and s28 for ciaI ). In summary, CiaI is the first flagellar type III secreted protein to be identified that plays a role in C. jejuni survival within cells. Ectopic expression experiments coupled with mutational analysis supports the possibility that the C-terminal dileucine motif within CiaI contributes to the targeting of the protein to unique vesicles, where it could block the fusion of CCV with lysosomes (Fig. 9). Future studies will dissect the molecular details of the facultative intracellular lifestyle of Campylobacter.

Experimental procedures

ratories, Novato, CA, USA), 1% glutamine (Invitrogen) and 1% sodium pyruvate (Cellgro, Manassas, VA, USA). Cultures were maintained at 37°C in a humidified 5% CO2 incubator. Cells were grown to confluency and passaged using 10% trypsin (Sigma, St. Louis, MO, USA).

Generation of the Cj1450 and Cj1451 (dut) mutant The Cj1450 and dut mutant was generated using allelic exchange. Briefly, the Cj1450 gene, including both upstream and downstream fragments, were amplified from C. jejuni NCTC 11168 using Cj1450-F1-SacI and Cj1450-R2-XhoI and cloned into the suicide vector pBSK-Kan2. An internal portion of Cj1450 ORF was deleted using inverse PCR and the primers, Cj1450-R1-SacII and Cj1450-F2-SacII. The deleted portion was replaced with either a cat cassette or a tetO cassette (Taylor et al., 1987). The Cj1450::tetO mutant and Cj1450::cat mutant were both tested in the cell binding, invasion and survival assays. For the dut gene, upstream and downstream fragments were amplified from the C. jejuni F38011 strain using Dut-upF-SacI and Dut-dnR-XhoI and cloned into the pBSK-Kan2 vector. The internal region of dut ORF was deleted by inverse PCR using the primers, Dut-upR-SacII and Dut-dnF-SacII, and replaced with the cat gene. The resulting vectors, pBSK-Kan2-Cj1450::tetO, pBSK-Kan2-Cj1450::cat and pBSK-Kan2-dut::cat were electroporated, individually, into the C. jejuni F38011 wild-type strain. Transformants were selected for on MHB plates supplemented with Tet or Cam. TetR or CamR isolates were screened for Kan sensitivity, indicating a double crossover had occurred. Chromosomal disruption of Cj1450 with either tetO or cat and dut with the cat gene were confirmed via PCR using gene-specific primers.

Construction of a Cj1450 shuttle vector

Bacterial strains, plasmids and tissue cultures The strains, plasmids and tissue cultures used in this study are described in Table S1. Primers used in this study are listed in Table S2. All strains were constructed using standard genetic techniques.

Media, antibiotics and growth conditions All Campylobacter strains were grown in Mueller–Hinton (MH) broth or on MH agar plates containing 5% citrated bovine blood (MHB) at 37°C in a microaerobic environment, unless indicated otherwise. Where indicated, MHB plates were supplemented with chloramphenicol (Cam; 8 mg ml-1), tetracycline (Tet; 2 mg ml-1), kanamycin (Kan; 50 mg ml-1) or cefoperazone (Cef; 30 mg ml-1). Strains were passaged on fresh MHB plates every 48 h. E. coli strains were grown in either Luria–Bertani (LB) broth or on LB plates, supplemented with either ampicillin (Amp; 100 mg ml-1) or Kan (50 mg ml-1). Y. enterocolitica strains were incubated at 26°C in LB broth or plates supplemented with Tet (10 mg ml-1). INT 407 (ATCC #CCL-6) and HeLa (#CCL-2) cells were grown in Minimal Essential Medium (MEM; Invitrogen, Carlsbad, CA, USA) supplemented with 10% FBS (Hyclone Labo-

We generated the pRY111-Cj1450 FLAG-tagged and pRY111-Cj1450 (with no FLAG-tag) complementation vectors using the following approach. To generate the construct that produced Cj1450 with the C-terminal FLAG, a fragment harbouring the Cj1450 gene and native promoter was PCR amplified using the primers Cj1450-EcoRI-F and Cj1450-BglII-R. A DNA fragment containing the FLAG tag was amplified from vector FlaA-FLAG (J. Neal-McKinney and M.E. Konkel, unpublished, contains flaA with in-frame FLAG tag in pRY111 MCS) using the primers FlaA-FLAG-F and FlaA-FLAG-R. The amplicon containing Cj1450 was digested with BglII, and the amplicon containing the FLAG tag was digested with BamHI. The two fragments were then ligated, and the ligation product used as template for PCR using the primers Cj1450-EcoRI-F and FLAG-TAA-SacI-R, to generate an amplicon containing the native promoter and ORF of Cj1450 fused in frame with the FLAG tag and a TAA stop codon. The PCR product was then cloned into pRY111 using EcoRI and SstI sites to generate the pRY111-Cj1450 FLAG-tagged vector. We also generated a pRY111-Cj1450 complementation vector without the FLAG-tag using primers Cj1450-EcoRI-F and Cj1450-CompBamHI-R to amplify Cj1450 with its native promoter, which was then cloned into pRY111 using BamHI and EcoRI. The resultant complementation vectors were sequence confirmed



Fig. 9. Model of C. jejuni binding, internalization and intracellular survival. C. jejuni binds to fibronectin via the CadF and FlpA adhesins, causing clustering of the a5b1 integrins. The clustering of integrins, as well as undefined secreted proteins (red circles), stimulates host cell signalling through Rho GTPase activation. C. jejuni is then internalized. Once internalized, C. jejuni are contained in CCV. With the aid of the secreted proteins, CCV acquire different cellular markers, EEA-1 ( ), Rab7 (䉲) and LAMP-1 ( ), but not Cathepsin D (–), as they are trafficked through the cell via a pathway that is distinct from the classical endocytic pathway, ultimately residing near the Golgi apparatus. © 2011 Blackwell Publishing Ltd, Molecular Microbiology, 80, 1296–1312


and electroporated into the C. jejuni F38011 wild-type strain and Cj1450::tetO mutant. Transformants were selected on MH blood agar containing Cam 8 mg ml-1 and the presence of the vectors was confirmed by PCR.

Generation of Cj1450-FLAG chromosomal insertion vector The Cj1450-FLAG gene was inserted into the chromosome of the C. jejuni Cj1450 mutant via allelic exchange. Two PCR products were amplified from C. jejuni F38011 chromosomal DNA using the primers FlaInsert-Up-SacI-F and FlaInsert-UpXhoI-SacII-R for an upstream homologous region containing Cj1337, and FlaInsert-XhoI-NotI-BamHI-Dn-F and FlaInsertDn-KpnI-R for a downstream homologous region containing Cj1338. They were subsequently cloned into pBSK-Kan2 using SacI and XhoI for the upstream region and XhoI and KpnI for the downstream region, creating a multiple cloning site between the two homologous regions that contained SacII, XhoI, BamHI and NotI. The cat gene from pRY111 was amplified using the primers CAT-BglII-F and CAT-BglII-R, digested with BglII, and cloned into the BamHI site. The Cj1450-FLAG gene was amplified from the pRY111-Cj1450 FLAG-tagged vector using the primers M13R-SacII and M13F-SacII, and cloned into the SacII site to create the Cj1450-FLAG chromosomal insertion vector. The suicide vector was electroporated into the C. jejuni Cj1450 mutant as previously described. Transformants were plated on MHB Cam media, and Camr colonies were selected and screened for sensitivity to Kan. Synthesis of Cj1450-FLAG was confirmed by immunoblot using an a-FLAG antibody (not shown).

Analysis of Cj1450 operon RNA was extracted from the C. jejuni wild-type strain, Cj1450::tetO mutant and dut::cat mutant using Ambion RiboPure-Bacteria (Applied Biosystems, Carlsbad, CA, USA) as outlined by the supplier. DNase treatment was carried out following RNA extraction using the RQ1 RNasefree DNase (Promega, Madison, WI, USA) followed by a phenol/chloroform extraction. cDNA was generated using ThermoScript RT-PCR System (Invitrogen) according to manufacturer’s methods and 2 mg of RNA. As a control to access the amount of chromosomal DNA present samples were also run in the absence of ThermoScript RT. Real-time RT-PCR amplification of 0.5 ml of cDNA was then performed with Cj1450, dut and glyA (an internal control) using Power SYBRGreen PCR master mix (Applied Biosystems) according to manufacturer’s protocols. The Cj1450-forwardRTPCR and Cj1450-reverse-RTPCR primers were used for analysis of Cj1450. The Cj1450-R2-Xho1 and Dut-RTForward primers were used for analysis of dut. The GlyA RT-F and GlyA RT-R primers were used for analysis of glyA. The real-time RT-PCR analysis was performed using a 7500 Fast thermocycler with the following PCR parameters: 50°C for 2 min, 95°C for 10 min and then 40 cycles of denaturation at 95°C for 15 s and annealing at 55°C for 1 min. Ct values were determined using 7500 software v2 (Applied Biosystems). When comparing these same RNA sample in the absence of reverse transcriptase, DNA contamination was not detected (data not shown).

The primers used for real-time RT-PCR analysis of Cj1450 and dut were also used to amplify products from the cDNA samples using Taq (Invitrogen) in a standard PCR reaction, to determine if Cj1450 and dut were co-transcribed. Internal portions of Cj1450 and dut were amplified using the primer pairs Cj1450-forward-RTPCR and Cj1450-reverse-RTPCR, and Cj1450-R2-Xho1 and Dut-RT-Forward respectively. Cj1450-forward-RTPCR and Cj1450-R2-Xho1 were used to amplify transcripts from cDNA samples treated with and without RT.

PLA assay and analysis The phospholipase assay was performed as previously described (Christensen et al., 2009). Briefly, Y. enterocolitica strains were shaken overnight at 26°C in LB broth. Secretion was induced by spotting the culture on TYE PLA medium and incubation at 26°C. Each isolate along with the positive control, Y. enterocolitica strain expressing wild-type YplA, was tested for secretion at least three times from at least two independent PLA plate assays to ensure reproducible results. All plates were scanned at 24 h at 300 dpi resolution to create a digital archive of the secretion results. The secretion zone widths were measured manually from digital images using select tools in Adobe Photoshop CS2 version 9.0.2 (Adobe Systems Incorporated, USA).

Determination of YplA fusion protein secretion by immunoblot Immunoblots for detection of YplA fusion proteins were performed as previously described (Christensen et al., 2009). Briefly Y. enterocolitica strains were incubated overnight in LB broth with shaking at 26°C. Secretion was induced by inoculation of TYE broth with Y. enterocolitica cultures and incubation with shaking at 26°C for 4–6 h. The OD540 of all cultures was determined to be 0.5 for the 0 h time-point of the secretion assay. After 2 h of shaking at 26°C, OD540 were determined for normalization of whole cell lysate samples, and 1 ml of each supernatant was harvested. Proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were electrophoretically transferred to polyvinylidene fluoride membranes (PVDF; Immobilon P; Millipore Corp., Bedford, MA, USA) for immunoblot analysis. Immunoblot development was done by chemiluminescence (Western Lightning, PerkinElmer Life Sciences) and film exposure (Biomax MR film, Kodak). Detection of YplA fusion proteins was performed a minimum of two times on samples collected from different experiments.

C. jejuni secretion assay Secretion assays and metabolic labelling with [35S]methionine of different C. jejuni strains were performed as described elsewhere (Konkel and Cieplak, 1992). Briefly, strains were grown overnight on MH plates supplemented with 0.1% DOC, and resuspended in MEM containing [35S]-methionine in the presence or absence of 1% dialysed FBS. Cultures were then incubated at 37°C for 3 h in



microaerobic conditions. Following incubation, the WCLs and supernatants were collected. Supernatants were concentrated tenfold with an HCl-acetone precipitation. Samples were separated via SDS-PAGE (12% or 15% gels). Gels were soaked in Amplify (Amersham Biosciences), dried and exposed to film (BioMax MR Film, Kodak). After a 4 day exposure, the autoradiograph was developed. Secretion assays were preformed in triplicate.

FBS MEM containing 250 mg ml-1 of gentamicin was added for 3 h to kill any bacteria that had escaped from the epithelial cells. The epithelial cells were then rinsed with PBS, lysed, and the number of intracellular bacteria determined for the 48 h time-point. The binding, internalization and intracellular bacteria assays were performed in quadruplicate and repeated three times. Results were analysed for significance using the unpaired Student’s t-test (Bender et al., 2008).

Immunoblots

Immunofluorescence assays of C. jejuni-infected cells

Immunoblots were performed on WCLs or supernatants from secretion assays. The proteins were subjected to SDS-PAGE and transferred to PVDF membranes (Towbin et al., 1979). The membrane was probed using a-CysM (1:1000) (Christensen et al., 2009), a-FlaA/FlaB (1:1000) (Neal-McKinney et al., 2010) or a-FLAG (1:1000; Sigma) as the primary antibody. A secondary a-rabbit IgG conjugated to peroxidase (Sigma) was used at a dilution of 1:5000. Proteins were visualized by chemiluminescence (Western Lightning, PerkinElmer Life Sciences). Immunoblots were performed a minimum of two times on samples collected from different experiments. Band intensities were quantified with the ImageJ Analysis program (National Institutes of Health; http:// rsbweb.nih.gov/ij) and standardized to protein loads, as judged by Coomassie Brilliant Blue R-250 stained SDSpolyacrylamide gels. Results were analysed for significance using an unpaired Student’s t-test (Bender et al., 2008).

Indirect immunofluorescence was performed on monolayers that had been infected with C. jejuni. Briefly, the epithelial cells were seeded at 7.5 ¥ 104 on sterile coverslips (13 mm) in 24-well plates and incubated the same as in the internalization and intracellular survival assays. At 18 and 24 h for the intracellular survival assay, or after 18 h of transfection, cells were fixed for 5 min in cold 4% paraformaldehyde, washed three times with 0.1% BSA PBS, permeabilized with 0.1% saponin PBS for 20 min, and washed three times in PBS. Monolayers were then stained with 1° antibody in 0.1% saponin PBS for 45 min unless stated otherwise, washed three times in PBS, stained with 2° antibody in 0.1% saponin PBS for 45 min and washed three times with PBS. Coverslips were then mounted with Vectashield (Vector laboratories) containing DAPI (4′,6-diamidino-2-phenylindole) and visualized using the Nikon Eclipse TE2000 inverted epifluoresence microscope. Images were captured using the imaging software MetaMorph version 5, and processed in Adobe Photoshop CS3 Extended. Merged images were generated by overlaying three individual channel images (Red, Green and Blue) into a single three-colour image. Primary antibodies used were a-EEA1 (1:50; Transduction Laboratories, Lexington, KY, USA), a-Rab7 (1:15; Santa Cruz Biotechnology, Santa Cruz, CA, USA), a-LAMP1 (1:50; Developmental Studies Hybridoma Bank), a-Cathepsin D (1:50; Santa Cruz Biotechnology), and a-Campylobacter whole cell lysate cocktail for four different strains (1:1000). Secondary antibodies used were Rhodamine conjugated a-mouse and a-goat (1:250; Santa Cruz Biotechnology), FITC conjugated a-mouse and goat (1:250; Santa Cruz Biotechnology) and Texas Red conjugated a-rabbit (1:8000; Jackson ImmunoResearch, West Grove, PA, USA). A minimum of 50 CCV were analysed with each cellular marker.

Binding, internalization and intracellular survival assays Binding, internalization and intracellular survival assays were carried out as previously described with minor modifications (Konkel et al., 1992a). Semi-confluent monolayers of INT 407 or HeLa cells (1.5 ¥ 105 cells ml-1) were infected at an MOI of 300 (5 ¥ 107 cfu). Bacteria and monolayers were centrifuged at 600 g for 5 min to enhance bacteria–cell contact. The infected monolayers were then incubated at 37°C in a humidified 5% CO2 incubator. To determine the number of adherent bacteria after 30 min, the epithelial cells were rinsed with PBS and lysed with a solution of 0.1% Triton X-100. The cell lysates were serially diluted and plated on MHB plates to determine the number of bound bacteria. For internalization, the C. jejuni-infected cell monolayers were incubated for 3 h, at which time they were rinsed with MEM containing 1% FBS and incubated with 1% FBS MEM containing 250 mg ml-1 of gentamicin for 3 h to kill extracellular bacteria. The number of internalized bacteria was quantified as described for the bound bacteria. To determine intracellular survival at 24 and 48 h, two separate plates of C. jejuni-infected cells were rinsed with 1% FBS MEM after a 3 h incubation, and then incubated in with 1% FBS MEM containing 10 mg ml-1 of gentamicin for an additional 18 h. Following this incubation period, one plate was rinsed with PBS and lysed with a solution of 0.1% Triton X-100. Serial dilutions were plated on MHB plates to determine the number of intracellular bacteria 24 h post inoculation. The second plate was rinsed with 1% FBS MEM and the media was replaced with fresh media without antibiotic. At 45 h post inoculation, the C. jejuniinfected cells were rinsed twice with 1% FBS MEM, and 1%

Construction of pEGFP-N1 vectors All the pEGFP-N1 vectors were constructed in a similar fashion. Each gene of interest was PCR amplified from C. jejuni F38011 using a forward primer that annealed six nucleotides downstream of the methionine ATG initiation codon and a reverse primer. The forward primer contained a 5′ XhoI restriction site followed by a Kozak sequence (ACCATGGGC), and the reverse primer contained a 5′ SacII restriction site, which annealed to the end of the ORF minus the termination codon. The PCR products were digested and cloned into pEGFP-N1. The primers used to construct the different pEGFP-N1 vectors are listed in Table S2. CiaI-GFP variants were constructed using the Cj1450-FGFP-Cterm forward primer and a specific reverse primer that



would result in the truncation. CiaI(L153,154A)-GFP was constructed by performing inverse PCR on pEGFP-N1-CiaI using the primers CiaI-invA-Dileu and CiaI-invB-Dileu.

HeLa cell transfections, quantification of CiaI localization with cell markers, and immunoblot analysis HeLa cells were transfected with the pEGFP-N1 using Effectene (Qiagen, Valencia, CA, USA) according to the manufacturer’s protocol with a few variations. Cells were seeded at 5 ¥ 104 on sterile cover slips (13 mm) in 24-well plates and incubated for 24 h in a 5% CO2 incubator. Cells were then washed with MEM, and DNA-enhancer mixture (400 ng DNA, 1.6 ml enhancer, 5 ml Effectene, 56.4 ml EC buffer, 750 ml MEM) was added and incubated for 18 h. CiaI-GFP transfected cells were fixed, labelled with cell markers, mounted and imaged according to the protocol outlined for the immunofluorescence assays performed on the C. jejuni-infected cells. For each cellular marker, approximately 1000 CiaI-GFP punctate spots were analysed for co-localization. The number of CiaI-GFP punctate spots co-localized with each cellular marker reported as a percentage. Transfected HeLa cells were lysed at 4°C for 20 min in 100 ml of lysis buffer [10 mM Hepes, 10% glycerol, 50 mM sodium fluoride, 150 mM sodium chloride, 1% Triton X-100, Protease cocktail (Sigma, P2714) and 1 mM sodium orthovanadate]. Cell slurries were then concentrated eightfold by HCl-acetone precipitation and resuspended in water. A standard BCA was performed to determine protein concentration. The protein samples were separated in SDS-15% polyacrylamide gels (5 mg of protein per well), and transferred to PVDF (Towbin et al., 1979). Membranes were incubated overnight at 4°C with a-EGFP (1:1000, Clontech, Mountain View, CA, USA) and a-actin (1:200, Santa Cruz Biotechnology). The immunoblot was then labelled with either a secondary a-mouse (1:1000, Sigma) or a-goat (1:1000, Sigma) IgG conjugated to peroxidase and developed using chemiluminescence (Western Lightning, PerkinElmer Life Sciences). The experiments and immunoblots were repeated twice.

Acknowledgements We thank S.A. Pacheco for constructing the Cj1450 mutant. The LAMP-1 monoclonal antibody developed by J.T. August and J.E.K. Hildreth was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology, Iowa City. We also thank C.L. Larson, D.R. Samuelson and T. Eucker for assistance with the experimental assays and for critical review of the manuscript. This study was supported by a grant awarded to M.E. Konkel from the National Institute of Health, Department of Health and Human Services, under contract number NO1-Al-30055, and, in part, from funds provided by the School of Molecular Biosciences.

References Bender, F., Groher, M., Khamene, A., Wein, W., Heibel, T.H., and Navab, N. (2008) 3D dynamic roadmapping for abdominal catheterizations. Med Image Comput Comput Assist Interv 11: 668–675.

Braulke, T., and Bonifacino, J.S. (2009) Sorting of lysosomal proteins. Biochim Biophys Acta 1793: 605–614. Brumell, J.H., and Scidmore, M.A. (2007) Manipulation of rab GTPase function by intracellular bacterial pathogens. Microbiol Mol Biol Rev 71: 636–652. Brumell, J.H., Kujat-Choy, S., Brown, N.F., Vallance, B.A., Knodler, L.A., and Finlay, B.B. (2003) SopD2 is a novel type III secreted effector of Salmonella typhimurium that targets late endocytic compartments upon delivery into host cells. Traffic 4: 36–48. Calvo, P.A., Frank, D.W., Bieler, B.M., Berson, J.F., and Marks, M.S. (1999) A cytoplasmic sequence in human tyrosinase defines a second class of di-leucine-based sorting signals for late endosomal and lysosomal delivery. J Biol Chem 274: 12780–12789. Cano, D.A., Martinez-Moya, M., Pucciarelli, M.G., Groisman, E.A., Casadesus, J., and Garcia-Del Portillo, F. (2001) Salmonella enterica serovar Typhimurium response involved in attenuation of pathogen intracellular proliferation. Infect Immun 69: 6463–6474. Carrillo, C.D., Taboada, E., Nash, J.H., Lanthier, P., Kelly, J., Lau, P.C., et al. (2004) Genome-wide expression analyses of Campylobacter jejuni NCTC11168 reveals coordinate regulation of motility and virulence by flhA. J Biol Chem 279: 20327–20338. Christensen, J.E., Pacheco, S.A., and Konkel, M.E. (2009) Identification of a Campylobacter jejuni-secreted protein required for maximal invasion of host cells. Mol Microbiol 73: 650–662. Coker, A.O., Isokpehi, R.D., Thomas, B.N., Amisu, K.O., and Obi, C.L. (2002) Human campylobacteriosis in developing countries. Emerg Infect Dis 8: 237–244. Coleman, S.H., Madrid, R., Van Damme, N., Mitchell, R.S., Bouchet, J., Servant, C., et al. (2006) Modulation of cellular protein trafficking by human immunodeficiency virus type 1 Nef: role of the acidic residue in the ExxxLL motif. J Virol 80: 1837–1849. Day, W.A., Jr, Sajecki, J.L., Pitts, T.M., and Joens, L.A. (2000) Role of catalase in Campylobacter jejuni intracellular survival. Infect Immun 68: 6337–6345. De Melo, M.A., Gabbiani, G., and Pechere, J.C. (1989) Cellular events and intracellular survival of Campylobacter jejuni during infection of HEp-2 cells. Infect Immun 57: 2214–2222. de Felipe, K.S., Glover, R.T., Charpentier, X., Anderson, O.R., Reyes, M., Pericone, C.D., and Shuman, H.A. (2008) Legionella eukaryotic-like type IV substrates interfere with organelle trafficking. PLoS Pathog 4: e1000117. Gould, C.M., Diella, F., Via, A., Puntervoll, P., Gemund, C., Chabanis-Davidson, S., et al. (2010) ELM: the status of the 2010 eukaryotic linear motif resource. Nucleic Acids Res 38: D167–D180. Humphreys, S., Stevenson, A., Bacon, A., Weinhardt, A.B., and Roberts, M. (1999) The alternative sigma factor, sigmaE, is critically important for the virulence of Salmonella typhimurium. Infect Immun 67: 1560–1568. Kelly, B.T., McCoy, A.J., Spate, K., Miller, S.E., Evans, P.R., Honing, S., and Owen, D.J. (2008) A structural explanation for the binding of endocytic dileucine motifs by the AP2 complex. Nature 456: 976–979. Konkel, M.E., and Cieplak, W., Jr (1992) Altered synthetic



response of Campylobacter jejuni to cocultivation with human epithelial cells is associated with enhanced internalization. Infect Immun 60: 4945–4949. Konkel, M.E., Hayes, S.F., Joens, L.A., and Cieplak, W., Jr (1992a) Characteristics of the internalization and intracellular survival of Campylobacter jejuni in human epithelial cell cultures. Microb Pathog 13: 357–370. Konkel, M.E., Corwin, M.D., Joens, L.A., and Cieplak, W. (1992b) Factors that influence the interaction of Campylobacter jejuni with cultured mammalian cells. J Med Microbiol 37: 30–37. Konkel, M.E., Garvis, S.G., Tipton, S.L., Anderson, D.E., Jr, and Cieplak, W., Jr (1997) Identification and molecular cloning of a gene encoding a fibronectin-binding protein (CadF) from Campylobacter jejuni. Mol Microbiol 24: 953– 963. Konkel, M.E., Kim, B.J., Rivera-Amill, V., and Garvis, S.G. (1999) Bacterial secreted proteins are required for the internaliztion of Campylobacter jejuni into cultured mammalian cells. Mol Microbiol 32: 691–701. Konkel, M.E., Monteville, M.R., Rivera-Amill, V., and Joens, L.A. (2001) The pathogenesis of Campylobacter jejunimediated enteritis. Curr Issues Intest Microbiol 2: 55– 71. Konkel, M.E., Klena, J.D., Rivera-Amill, V., Monteville, M.R., Biswas, D., Raphael, B., and Mickelson, J. (2004) Secretion of virulence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. J Bacteriol 186: 3296–3303. Konkel, M.E., Christensen, J.E., Keech, A.M., Monteville, M.R., Klena, J.D., and Garvis, S.G. (2005) Identification of a fibronectin-binding domain within the Campylobacter jejuni CadF protein. Mol Microbiol 57: 1022–1035. Konkel, M.E., Larson, C.L., and Flanagan, R.C. (2010) Campylobacter jejuni FlpA binds fibronectin and is required for maximal host cell adherence. J Bacteriol 192: 68– 76. Krause-Gruszczynska, M., Rohde, M., Hartig, R., Genth, H., Schmidt, G., Keo, T., et al. (2007) Role of the small Rho GTPases Rac1 and Cdc42 in host cell invasion of Campylobacter jejuni. Cell Microbiol 9: 2431–2444. Larson, C.L., Christensen, J.E., Pacheco, S.A., Minnich, S.A., and Konkel, M. (2008) Campylobacter jejuni secretes proteins via the flagellar type III secretion system that contribute to host cell invasion and gastroenteritis. In Campylobacter. Nachamkin, I., Szymanski, C.M., and Blaser, M.J. (eds). Washington, DC: ASM Press, pp. 315– 332. Letourneur, F., and Klausner, R.D. (1992) A novel di-leucine motif and a tyrosine-based motif independently mediate lysosomal targeting and endocytosis of CD3 chains. Cell 69: 1143–1157. Malik-Kale, P., Parker, C.T., and Konkel, M.E. (2008) Culture of Campylobacter jejuni with sodium deoxycholate induces virulence gene expression. J Bacteriol 190: 2286–2297. Monteville, M.R., Yoon, J.E., and Konkel, M.E. (2003) Maximal adherence and invasion of INT 407 cells by Campylobacter jejuni requires the CadF outer-membrane protein and microfilament reorganization. Microbiology 149: 153–165.

Neal-McKinney, J.M., Christensen, J.E., and Konkel, M.E. (2010) Amino-terminal residues dictate the export efficiency of the Campylobacter jejuni filament proteins via the flagellum. Mol Microbiol 76: 918–931. Parkhill, J., Wren, B.W., Mungall, K., Ketley, J.M., Churcher, C., Basham, D., et al. (2000) The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403: 665–668. Pond, L., Kuhn, L.A., Teyton, L., Schutze, M.P., Tainer, J.A., Jackson, M.R., and Peterson, P.A. (1995) A role for acidic residues in di-leucine motif-based targeting to the endocytic pathway. J Biol Chem 270: 19989–19997. Rivera-Amill, V., and Konkel, M.E. (1999) Secretion of Campylobacter jejuni Cia proteins is contact dependent. Adv Exp Med Biol 473: 225–229. Rivera-Amill, V., Kim, B.J., Seshu, J., and Konkel, M.E. (2001) Secretion of the virulence-associated Campylobacter invasion antigens from Campylobacter jejuni requires a stimulatory signal. J Infect Dis 183: 1607– 1616. Ruiz-Palacios, G.M. (2007) The health burden of Campylobacter infection and the impact of antimicrobial resistance: playing chicken. Clin Infect Dis 44: 701–703. Schesser, K., Frithz-Lindsten, E., and Wolf-Watz, H. (1996) Delineation and mutational analysis of the Yersinia pseudotuberculosis YopE domains which mediate translocation across bacterial and eukaryotic cellular membranes. J Bacteriol 178: 7227–7233. Sory, M.P., Boland, A., Lambermont, I., and Cornelis, G.R. (1995) Identification of the YopE and YopH domains required for secretion and internalization into the cytosol of macrophages, using the cyaA gene fusion approach. Proc Natl Acad Sci USA 92: 11998–12002. Takatsu, H., Katoh, Y., Shiba, Y., and Nakayama, K. (2001) Golgi-localizing, gamma-adaptin ear homology domain, ADP-ribosylation factor-binding (GGA) proteins interact with acidic dileucine sequences within the cytoplasmic domains of sorting receptors through their Vps27p/Hrs/ STAM (VHS) domains. J Biol Chem 276: 28541– 28545. Taylor, D.E., Hiratsuka, K., Ray, H., and Manavathu, E.K. (1987) Characterization and expression of a cloned tetracycline resistance determinant from Campylobacter jejuni plasmid pUA466. J Bacteriol 169: 2984–2989. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76: 4350–4354. Voth, D.E., Howe, D., Beare, P.A., Vogel, J.P., Unsworth, N., Samuel, J.E., and Heinzen, R.A. (2009) The Coxiella burnetii ankyrin repeat domain-containing protein family is heterogeneous, with C-terminal truncations that influence Dot/Icm-mediated secretion. J Bacteriol 191: 4232– 4242. Watson, R.O., and Galan, J.E. (2008) Campylobacter jejuni survives within epithelial cells by avoiding delivery to lysosomes. PLoS Pathog 4: e14. Weber, S.S., Ragaz, C., Reus, K., Nyfeler, Y., and Hilbi, H. (2006) Legionella pneumophila exploits PI(4)P to anchor secreted effector proteins to the replicative vacuole. PLoS Pathog 2: e46.



Wosten, M.M., Mourik, A.V., and van Putten, J.P. (2008) Regulation of genes in Campylobacter jejuni. In Campylobacter, 3rd edn. Nachamkin, I., Szymanski, C.M., and Blaser, M.J. (eds). Washington, DC: ASM Press, pp. 611– 624. Xu, F.Z., Ye, H., Wang, J.J., and Yu, W.Y. (2008) The effect of site-directed mutagenesis of the ambient amino acids of leucine-based sorting motifs on the localization of chicken invariant chain. Poult Sci 87: 1980–1986.

Supporting information Additional supporting information may be found in the online version of this article. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
