|Home | About | Journals | Submit | Contact Us | Français|
Vibrio vulnificus is a Gram-negative bacterium that causes a fatal septicemia. One of its virulence factors is a membrane-bound lipoprotein, IlpA, which can induce cytokine production in human immune cells. In the present study, the role of IlpA as an adhesion molecule was investigated. An ilpA-deleted V. vulnificus mutant showed significantly decreased adherence to INT-407 human intestinal epithelial cells, which in turn resulted in reduced cytotoxicity. The ΔilpA mutant recovered the adherence ability of the wild type by complementation in trans with the intact ilpA gene. In addition, pretreatment of V. vulnificus with anti-IlpA polyclonal antibodies resulted in a significant reduction of bacterial adherence. To localize the domain of IlpA required for cytoadherence, three truncated recombinant IlpA polypeptides were constructed and tested for the ability to adhere to human cells by a ligand-binding immunoblot assay and fluorescence microscopy. The polypeptide containing the carboxy (C)-terminal hydrophilic domain exhibited direct binding to INT-407 cells. Therefore, the C-terminal domain of IlpA allows this protein to be an adhesion molecule of V. vulnificus.
Vibrio vulnificus is a Gram-negative pathogenic bacterium that is encapsulated, motile, and invasive. This pathogen is frequently associated with primary septicemia following the consumption of contaminated shellfish. Over 50% of patients with V. vulnificus-induced septicemia die due to multiorgan failure as a result of rapidly progressive shock syndrome (3, 19).
Diverse virulence factors have been proposed for V. vulnificus. Based on the attenuated mouse lethality of V. vulnificus mutants deficient in capsular polysaccharide (CPS) or exopolysaccharides (EPS), these molecules have been shown to be important for V. vulnificus pathogenesis (18, 38). Type IV pilin was confirmed to be involved in the virulence of V. vulnificus via genetic deletion of its structural gene, pilA (26, 27). In addition, motility was also discovered to be a critical virulence determinant of V. vulnificus (21). Secreted proteins, such as a cytolytic hemolysin (37) and an elastase (24), have been proposed to cause damage to host tissues, but in vivo studies using a knockout V. vulnificus mutant of vvhA (cytolysis gene) or vvpE (elastase gene) did not convincingly demonstrate that these proteins are key virulence factors causing lethality in mice or lysis of human cells (11, 30, 36).
The initial stage of microbial infection of host cells is mediated by interactions of the surface proteins of the pathogen with the connective tissues or epithelial cells of the host to facilitate bacterial adherence and/or to elicit signal transduction within host cells (5). For example, the most abundant outer membrane protein (OMP) of V. vulnificus, OmpU, has been identified to be involved in interactions with host extracellular matrix proteins, such as fibronectin (7). Knockout mutagenesis of the ompU gene results in the loss of cytoadherence as well as in decreased bacterial toxicity toward epithelial cells and mice. Membrane-bound IlpA lipoprotein was also identified to interact with host cells via stimulating the production of proinflammatory cytokines in human monocytes via Toll-like receptor 2 (TLR2). Lipid moieties attached to the N-terminal end of the processed IlpA protein are required for cytokine production (8). Subsequent investigation of V. vulnificus IlpA in this study showed that this protein is also important in bacterial adherence to human cell lines. Thus, we further identified the region of IlpA required for adherence to human cells.
The strains and plasmids used in this study are listed in Table Table1.1. Escherichia coli was grown at 37°C in Luria-Bertani (LB) broth (1% [wt/vol] tryptone, 0.5% [wt/vol] yeast extract, 1% [wt/vol] NaCl, pH 7.5) supplemented with ampicillin at 100 μg/ml for the maintenance of plasmids. Various strains of V. vulnificus were cultured in LBS broth (1% [wt/vol] tryptone, 0.5% [wt/vol] yeast extract, 2% [wt/vol] NaCl, pH 7.5) at 30°C with tetracycline (2 μg/ml) when needed. All medium components were purchased from Difco, and the chemicals and antibiotics were obtained from Sigma.
A plasmid, pilpA[1-169], was constructed to express a mutated recombinant IlpA (rIlpA) polypeptide in which 99 amino acids at the C terminus were missing. The ilpA′ DNA fragment of 729 bp, which encodes amino acid residues from the N terminus to the 169th amino acid of IlpA, was amplified from the genomic DNA of wild-type V. vulnificus by a PCR using the primers ilpA-comF (5′-GGTTGGATCCATTGGTGAGCT-3′; the underlined sequence denotes a BamHI restriction site) and ilpAΔ170-CR (5′-CCCAAGCTTTCAGTCACGAACGGTCGCAAGTAG-3′; the underlined sequence indicates a HindIII restriction site), digested with BamHI and HindIII, and then cloned into the broad-host-range plasmid pLAFR5. The resultant plasmid, pilpA[1-169], was transformed into E. coli SM10 λpir and then transferred into the V. vulnificus ilpA mutant by conjugation. The exconjugants were selected on thiosulfate-citrate-bile-sucrose agar containing tetracycline (2 μg/ml). Expression of the truncated polypeptide in the ΔilpA mutant was examined by Western blotting using anti-IlpA polyclonal antibodies.
Adherence assays were performed with the INT-407 cell line, which is derived from human intestinal epithelial cells (ATCC CCL-6), in 24-well culture plates. Each well of the culture plates was seeded with 2 × 105 INT-407 cells and incubated overnight at 37°C in the presence of 5% CO2. After removal of the medium and two washes with phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10.1 mM Na2HPO4, and 2 mM KH2PO4, pH 7.3), 1 ml of serum-free Dulbecco's modified Eagle medium (DMEM; Gibco-BRL) was added to the INT-407 cells. Various V. vulnificus strains (ATCC 29307 harboring pLAFR5, YS101 carrying pLAFR5, YS101 carrying pLAFR-ilpA , and YS101 carrying pilpA[1-169]) were grown at 30°C in LBS broth with 2 μg/ml tetracycline. At an optical density at 600 nm (OD600) of 0.7, 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) was added to the bacterial cultures, and they were incubated for another 3 h at 30°C before being used for cytoadherence assays. Cell monolayers were inoculated in triplicate with 50 μl of the diluted bacteria to give a multiplicity of infection (MOI) of 10 and were incubated at 37°C in 5% CO2 for 30 min. While V. vulnificus shows similar growth rates and maximum yields for growth at 30 and 37°C, INT-407 cells exhibit optimal growth at 37°C. Thus, the assays including both V. vulnificus and INT-407 cells were performed at 37°C. The monolayer was washed five times with prewarmed PBS to remove nonadherent bacteria. Following the last wash, the INT-407 cells were lysed with 0.1% Triton X-100 for 15 min. The bacteria were recovered from these cells with PBS, serially diluted 10-fold, and then plated on LBS agar.
Prior to cytoadherence assays, 1 × 108 cells of V. vulnificus were preincubated with anti-IlpA antibodies (8) for 1 h at various concentrations ranging from 10 to 30 μg/ml. Preimmune rat serum was used as a control immunoglobulin (IgG) to monitor the nonspecific effect of serum on the binding of V. vulnificus to INT-407 cells.
The cytotoxicity of V. vulnificus toward INT-407 cells was measured using a CytoTox 96 nonradioactive cytotoxicity assay kit (Promega). This cytotoxicity kit measures the lactate dehydrogenase (LDH) activity released into the culture medium by lysed INT-407 cells. To measure the total LDH activity of the cell lines used in the assays, Triton X-100 was added to a final concentration of 1.0% (vol/vol) to lyse the host cells. Various V. vulnificus strains (ATCC 29307 harboring pLAFR5, YS101 carrying pLAFR5, YS101 carrying pLAFR-ilpA , and YS101 carrying pilpA[1-169]) were grown at 30°C in LBS broth with 2 μg/ml tetracycline. At an OD600 of 0.7, the expression of IlpA protein was induced with 1 mM IPTG for 3 h. Monolayers of INT-407 cells were then inoculated in triplicate with 50 μl of the diluted bacteria to give an MOI of 10 or 50 and were incubated at 37°C in 5% CO2 for 1 to 3 h. The released LDH activity was then determined by colorimetric assay according to the manufacturer's instructions and is presented as the percentage of LDH activity relative to the total LDH activity of the cells lysed by a treatment with 1.0% Triton X-100.
Human peripheral blood mononuclear cells (PBMCs) were prepared by density gradient centrifugation, using Ficoll (Amersham Biosciences), and were resuspended in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine. Prepared PBMCs (1 × 105 cells per well) were cultured in 96-well plates and then incubated for 18 h with the lysates of various V. vulnificus strains, concanavalin A (ConA), or bovine serum albumin (BSA) at a concentration of 10 μg/ml. The levels of gamma interferon (IFN-γ) in the cell-free supernatants were determined using enzyme-linked immunosorbent assays (BD Biosciences). The concentration of IFN-γ was determined based on the standard curve for commercially available IFN-γ (BD Biosciences).
The full-length IlpA protein was expressed in E. coli by use of an overexpression plasmid, pET-IlpA (8). A DNA fragment containing the open reading frame (ORF) of IlpA was dissected into three parts (Fig. (Fig.1).1). The 5′ region of ilpA (273 bp) was amplified from the genomic DNA of V. vulnificus ATCC 29307 by a PCR using the primers rIlpA#1FB (5′-CGACGGATCCATGAAATTTAGCCTTAAAGGTT-3′; the underlined region indicates a BamHI restriction site) and rIlpA#1RH (5′-CGCACTCGAGGTAGCCACGGTCTTCC-3′; the underlined region denotes a restriction site for HindIII), and the 3′ region of ilpA (303 bp) was amplified with another set of primers, rIlpA#3FB (5′-CGACGGATCCTCTATCAACCTAACTCCGG-3′; the underlined region indicates a restriction site for BamHI) and rIlpA#3RH (5′-CCTGCTCGAGTTTCAAATGTGACGACTGCC-3′; the underlined region indicates a HindIII restriction site). The central part of the ilpA gene, which codes for amino acid residues constituting the major hydrophobic regions of IlpA, was amplified as a 333-bp DNA fragment, using the following set of primers: rIlpA#2FB (5′-ATGCGGATCCGCGGCATTGGATGACGG-3′; the underlined region indicates a BamHI site) and rIlpA#2RH (5′-CCTGCTCGAGCGAACGGTCGCAAGTAGG-3′; the underlined region denotes a HindIII site). The resultant ilpA DNA fragments were cloned into pQE30 (Qiagen) by using restriction sites for BamHI and HindIII to generate a series of overexpression plasmids for the truncated rIlpA polypeptides (pQE-IlpA#1, pQE-IlpA#2, and pQE-IlpA#3). These rIlpA polypeptides were expressed in a histidine-tagged form in E. coli JM109 cells to which 1 mM IPTG had been added and were purified using nickel-nitrilotriacetic acid affinity chromatography as described by the manufacturer (Qiagen).
Each well of a 24-well culture plate was seeded with 2 × 105 INT-407 cells and incubated overnight at 37°C in the presence of 5% CO2. After removal of the medium and two washes with DMEM, the cells were incubated in 1 ml of serum-free DMEM containing 10 μg of the rIlpA protein (8) as described by Jin et al. (12). After five washes with DMEM, the cells were disrupted with 70 μl of lysis buffer (10 mM Tris-HCl, pH 8.4, and 0.8% SDS) and boiled for 10 min. To examine whether rIlpA was stably bound to INT-407 cell monolayers, a portion of the lysate (200 μg of protein per well) was subjected to SDS-PAGE and then transferred to a polyvinylidene difluoride (PVDF) membrane. The membranes were blocked with 5% nonfat dry milk in TBST (Tris-buffered saline with 0.1% Tween 20) for 1 h at room temperature, treated with anti-IlpA antibodies (8) at 4°C overnight, and subsequently incubated with alkaline phosphatase (AP)-conjugated anti-rat IgG (Sigma-Aldrich). To determine the domain of IlpA responsible for binding to host cells, the ligand-binding immunoblot assays were performed with the three truncated rIlpA polypeptides tagged with histidine residues, and then the bound polypeptide was detected using anti-poly-His monoclonal antibodies (Sigma). Immunoreactive bands were visualized by using a nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP) system (Promega).
To monitor the protein levels of INT-407 cells after each treatment, the INT-407 cell lysates used for the above ligand-binding immunoblot assays were subjected to SDS-PAGE and then transferred to a PVDF membrane. The membranes were blocked with 5% nonfat dry milk in TBST for 1 h at room temperature, treated with anti-human Cu/Zn superoxide dismutase (Cu/Zn SOD) antibodies (Millipore) at 4°C overnight, subsequently incubated with AP-conjugated anti-rabbit IgG (Sigma-Aldrich), and then visualized using a kit containing NBT and BCIP (Promega).
Intact rIlpA and three truncated rIlpA polypeptides, rIlpA#1, rIlpA#2, and rIlpA#3, were conjugated with a fluorescent probe by use of an Alexa Fluor 555 microscale protein labeling kit (A30007; Molecular Probes). Briefly, 100 μg of recombinant polypeptide was incubated with Alexa Fluor 555-succinimidyl ester, and then unincorporated probe was removed by spin filters supplied with the kit. The fluorescence of the proteins at 280 nm and at 555 nm was measured to ensure appropriate labeling of the recombinant polypeptides as directed by the manufacturer. INT-407 cells were seeded and grown on glass coverslips for 24 h. Cells were blocked with 1% BSA for 30 min in DMEM, followed by a 1 h-treatment with 10 μg/ml Alexa Fluor 555-labeled polypeptides at 37°C. Cells were washed with PBS three times and fixed in 3.7% paraformaldehyde-PBS at room temperature for 30 min. After being washed three times with PBS, the coverslips were then mounted on glass slides by use of a mounting medium that included an antifading agent and sodium azide (fluorescence mounting medium; Dako) and were then observed under a confocal laser scanning microscope (LSM700; Carl Zeiss MicroImaging). For an assay to examine the competition of binding to INT-407 cells by intact rIlpA and rIlpA#3, Alexa Fluor 555-labeled full-length rIlpA (10 μg/ml) was mixed with unlabeled rIlpA#3 at molar ratios of 1:1, 1:5, and 1:10 and then incubated with INT-407 cells as described above.
Results are expressed as means ± standard deviations for three independent experiments. Statistical analysis was performed using analysis of variance (ANOVA). Differences were judged as statistically significant when the P value was <0.05. Statistical analysis for pairwise comparisons was performed using Student's t test (SigmaPlot, version 9; Systat Software Inc.). Differences were considered significant if the P values were <0.05.
V. vulnificus IlpA was isolated by immunoscreening of an expression library with antibodies raised against a fraction containing V. vulnificus OMPs, and it was further identified as an immunogenic protein of this microorganism (8). Some bacterial OMPs have been reported to have dual functions as adhesins and immunogens, such as the surface adhesion protein of Streptococcus pneumoniae (PsaA) (20) and the NadA adhesion protein of Neisseria meningitidis (6). Thus, in the present study, we examined in more detail the role of the IlpA protein in the interactions of V. vulnificus with host cells, particularly in the adherence of this organism to the human intestinal epithelial cell line INT-407 (Fig. (Fig.2A).2A). When the ratio of bacteria to INT-407 cells was 10:1, the percentage of adhered wild-type V. vulnificus carrying pLAFR5, the vector plasmid, was about 10% of the bacteria initially added to the assay mixture. In the same assay with the ΔilpA mutant YS101 carrying pLAFR5, the portion of bacterial cells adhered to INT-407 cells was significantly lowered, i.e., 2.1% of the added bacteria (P = 0.0026; Student's t test). However, when YS101 was carrying pLAFR-ilpA, an lipA+-containing plasmid, cytoadherence of this strain recovered almost (9.7%) to that of the wild type carrying pLAFR5.
The role of the IlpA protein in cytoadherence of wild-type V. vulnificus was further confirmed by examining if masking of IlpA on the bacterial surface with anti-IlpA antibodies inhibited bacterial adherence (Fig. (Fig.2B).2B). Wild-type V. vulnificus was treated for 30 min with various concentrations of anti-IlpA antibodies (ranging from 10 to 30 μg) prior to the adherence assay. As a control, wild-type V. vulnificus that had been incubated with rat preimmune serum at the same concentrations was used for the adherence tests. When wild-type V. vulnificus was preincubated with anti-IlpA antibodies, the levels of cytoadherence decreased gradually, in a dose-dependent manner, to 1.1%. Statistical analysis using ANOVA indicated a significant decrease in bacterial cytoadherence upon incubation with anti-IlpA antibodies (P = 6.2E−08). In contrast, 7.4 to 9.0% of the added V. vulnificus bacteria adhered to the cells treated with preimmune serum, regardless of the concentration of IgG. ANOVA showed the absence of a meaningful difference in this set of data (P = 0.75). The blocking effect of anti-IlpA antibodies on bacterial adherence was not derived from agglutination of bacterial cells in the presence of anti-IlpA antibodies (data not shown).
We examined whether a deficiency of IlpA affects the ability of V. vulnificus to lyse INT-407 cells. After INT-407 cells were incubated with either the wild type carrying pLAFR5 or V. vulnificus ΔilpA carrying pLAFR5 at two different MOIs, 10 and 50, the viability of the INT-407 cells was monitored for 3 h by measuring released LDH (Fig. (Fig.3).3). When INT-407 cells were treated with wild-type V. vulnificus carrying pLAFR5 at an MOI of 10, the percentages of released LDH increased gradually, to 95% lysis, as the incubation time extended to 3 h. When INT-407 cells were exposed to the larger number of wild-type V. vulnificus cells, i.e., an MOI of 50, they were completely lysed during 2 h of incubation. In contrast, cytotoxicity of the ΔilpA mutant carrying pLAFR5 was estimated to be significantly lower than that of the wild type at both MOIs (P < 0.01; Student's t test), except for the 3-h incubation at an MOI of 50 (Fig. (Fig.3B),3B), which was not different from the other incubations (P = 0.59; ANOVA).
In subsequent experiments, we examined whether a loss of cytotoxicity in the ΔilpA mutant could be recovered by introducing a complementation plasmid, pLAFR-ilpA. Inclusion of a complementation plasmid in the ΔilpA mutant increased LDH activity up to that of the wild type carrying the vector plasmid (P < 0.05; Student's t test).
We then investigated whether the IlpA protein binds directly to host cells. INT-407 cells were incubated with rIlpA, washed, and then prepared as lysates. Western blot analysis of these lysates with anti-rIlpA antibodies clearly showed that the rIlpA protein was stably bound to INT-407 cells (Fig. (Fig.4A).4A). As a loading control, the amount of human Cu/Zn SOD in the INT-407 cell lysates used for this assay was also monitored (17). Figure Figure4B4B shows a similar level of human Cu/Zn SOD in each sample, suggesting that the absence of immunoreactive bands in Fig. Fig.4A,4A, lane 2, was not due to a difference in the amounts of human cell lysates loaded on PAGE gels.
An in silico analysis of the amino acid sequence of IlpA revealed major hydrophobic regions, which may be localized within the membrane and connect the extracellular, lipidated N-terminal region and the putatively extracellular C-terminal region (Fig. (Fig.1)1) (http://www.expasy.ch/tools/protscale.html). Therefore, the whole ilpA gene was divided into three parts, as shown in Fig. Fig.1,1, and cloned into an expression vector; the N-terminal hydrophilic, central hydrophobic, and C-terminal hydrophilic regions of IlpA were made as truncated and histidine-tagged forms and were named rIlpA#1, rIlpA#2, and rIlpA#3, respectively. Truncated rIlpA polypeptides were examined for the ability to specifically adhere to INT-407 cells (Fig. (Fig.5A).5A). Only the rIlpA#3 polypeptide, containing the C-terminal domain, was clearly observed in INT-407 cell lysates. In contrast, the other two rIlpA polypeptides, rIlpA#1 and rIlpA#2, were not detected in INT-407 cell lysates. In the same manner, the amount of human Cu/Zn SOD in the INT-407 cell lysates was detected as a loading control for each sample (Fig. (Fig.5B)5B) and indicated that similar amounts of human cell lysates were used in this study.
Binding of truncated rIlpA polypeptides to INT-407 cells was directly visualized using a fluorescence microscope (Fig. (Fig.6A).6A). Three truncated rIlpA polypeptides were labeled by a red fluorescent probe, Alexa Fluor 555, and fluorescent recombinant polypeptides were then added to INT-407 cells. Only the rIlpA#3-treated cells demonstrated fluorescence on the host cells, whereas INT-407 cells incubated with the other two rIlpA polypeptides did not show any fluorescence signal.
Full-length rIlpA protein was labeled with a fluorescent probe and used for the binding experiments. The results demonstrated that the intact IlpA protein binds to INT-407 cells, as shown in the binding assay with rIlpA#3 (Fig. (Fig.6B,6B, panel a). To confirm the specificity of IlpA binding to INT-407 cells, the binding reaction mixture containing fluorescent rIlpA protein and INT-407 cells was examined with various amounts of unlabeled rIlpA#3 added, from 10 to 100 μg (Fig. (Fig.6B,6B, panels c, e, and g). Fluorescent signals on INT-407 cells gradually disappeared as the added amount of rIlpA#3 increased. Addition of unlabeled rIlpA#1 or rIlpA#2 did not affect the degree of rIlpA binding to INT-407 cells (data not shown). These two independent experiments clearly demonstrated that IlpA directly binds to the surfaces of INT-407 cells via its C-terminal region, which functions as an adhesin domain.
In a previous report (8), the ΔilpA mutant was demonstrated to be attenuated in virulence. This mutant stimulated less cytokine production from monocytes due to the absence of the lipidated N terminus of IlpA. In the present study, it was also shown that the C-terminal region of IlpA plays an important role in cytoadherence. To dissect the factor causing virulence attenuation of the ΔilpA mutant, the phenotypes of the ΔilpA mutant carrying pilpA[1-169], in which a C-terminally truncated IlpA polypeptide was produced, were analyzed (data not shown). This strain showed a similar level of adherence of INT-407 cells (1.7%) to that shown by the ΔilpA mutant carrying pLAFR5 (2.1%) (Fig. (Fig.2A).2A). This result confirmed the important role of the C-terminal region of IlpA in cytoadherence, as shown in Fig. Fig.55 and and66.
Next, we examined if the C-terminal region is also involved in stimulation of proinflammatory cytokine production in human PBMCs. When human PBMCs were treated with a lysate of the wild type carrying pLAFR5, the secretion of IFN-γ by PBMCs was 93 pg/ml (Fig. (Fig.7).7). In contrast, treatment of PBMCs with a lysate of the ΔilpA mutant carrying pLAFR5 resulted in a significantly lowered level of IFN-γ production (4 pg/ml) (P = 0.044; Student's t test), which was comparable to those of the control cells, such as PBMCs treated with medium only or with BSA. When a lysate of the ΔilpA mutant carrying pLAFR-ilpA was used to challenge PBMCs, IFN-γ production was restored to the level induced by wild-type V. vulnificus lysate (69 pg/ml) (P = 0.47; Student's t test). Lysates of the ΔilpA mutant carrying pilpA[1-169] were effective at inducing cytokine production by human PBMCs (56 pg/ml), showing similar levels to those shown by the wild type carrying the vector and the ΔilpA mutant carrying pLAFR-ilpA (P = 0.32 and 0.38, respectively). Human PBMCs treated with ConA, which served as a positive control for this experiment, showed IFN-γ production at 79 pg/ml. The capacity of the amino-terminal domain to reproduce most of the immunostimulating activity of IlpA suggests that the C-terminal domain of IlpA may not have a major role in immunostimulating activity of IlpA.
Since the plasmid encoding the C-terminally deleted IlpA, pilpA[1-169], was found to be defective in recovery of cytoadherence of the ΔilpA mutant (Fig. (Fig.2A)2A) but efficient in recovery of immunostimulation of the ΔilpA mutant (Fig. (Fig.7),7), we further examined whether this truncated IlpA was involved in bacterial cytotoxicity to INT-407 cells. The ΔilpA mutant carrying pilpA[1-169] exhibited lowered levels of cytotoxicity, 27 to 57% of those of the wild type carrying the vector plasmid, except for the 3-h incubation at an MOI of 50 (Fig. (Fig.3).3). Therefore, all of the results for the ΔilpA mutant carrying pilpA[1-169] suggest that decreased cytotoxicity of the ΔilpA mutant was caused mainly from reduced cytoadherence.
The severe and rapid cytopathological characteristics of V. vulnificus infection have led to this organism being considered a model system with which to investigate host-pathogen interactions. Among various virulence-related phenotypes, the ability of bacterial cells to adhere to host cells via bacterial adhesins is critical for the initial pathogenic interaction. The significant decrease in cytoadherence of a V. vulnificus ΔilpA mutant to INT-407 human cells suggests a potential role of the IlpA protein as an adhesin for this pathogenic microorganism.
The role of IlpA in adherence of V. vulnificus to human intestinal epithelial cells, such as INT-407 cells, was clearly demonstrated by comparative cytoadherence assays of wild-type and ΔilpA mutant V. vulnificus (Fig. (Fig.2A).2A). In addition, a ligand-binding immunoblot assay using rIlpA (Fig. (Fig.4A)4A) and microscopic observation using a fluorescent rIlpA (Fig. (Fig.6B)6B) demonstrated an association of rIlpA with INT-407 cells. As a result of reduced adherence, the mutant exhibited significantly less cytotoxicity than the wild type (Fig. (Fig.33).
IlpA has previously been reported to be an immunogenic protein of V. vulnificus which stimulates cytokine production in human blood cells (8). The V. vulnificus IlpA protein is presumed to be modified by acylation of its first cysteine residue with fatty acid chains after its N-terminal leader peptide of 22 amino acid residues is processed. Its lipid moiety is essential to activate the TLR2-mediated signaling pathway and then to trigger cytokine production by human monocytes. Therefore, the IlpA protein may play important roles in V. vulnificus pathogenicity via at least dual functions during the initial interactions of bacteria with host cells, i.e., cytoadherence to intestinal epithelial cells and immunostimulation in monocytes.
Some membrane-bound lipoproteins have been reported to have multiple functions. One of the best-characterized adhesin/immunogen proteins is the 33-kDa PsaA protein of S. pneumoniae, which directly binds to the target host cells, such as nasopharyngeal epithelial cells (9, 23, 29). Mutation in the psaA gene caused pleiotropic effects on the growth, oxidative stress response, adherence, and virulence of S. pneumoniae (14). PsaA was also found to be one of the main pneumococcal antigens and causes a distinct immune response in a mouse model which is different from those mediated by other pneumococcal antigens (25). Another group of well-known bacterial adhesins are pili or fimbriae (28). The thin-curled fimbriae of Salmonella enterica and the fimbrillin of Porphyromonas gingivalis are able to trigger cytokine production from human cells via the action of TLR2 (10, 33). Therefore, many membrane-bound lipoproteins as well as surface proteins of diverse bacteria play multiple roles, and the IlpA protein of V. vulnificus may be added to the list of bacterial proteins performing at least two distinct roles, facilitating cytoadherence and eliciting cytokine production.
Both an in silico analysis of the amino acid sequence of IlpA and a comparison with the structure of the IlpA homolog lipoprotein 9 of Staphylococcus aureus (35) allowed us to localize the major hydrophobic regions that might reside in the outer leaflet of the membrane and connect two extracellular domains, a lipidated N-terminal region and a putatively extracellular C-terminal region (data not shown). Ligand-binding immunoblot assays using three different truncated rIlpAs (Fig. (Fig.5A)5A) showed that the polypeptide rIlpA#3, including the C-terminal one-third of the IlpA protein, was as efficient in binding to INT-407 cells as the intact IlpA protein (Fig. (Fig.4A).4A). The direct binding of rIlpA#3 was confirmed by microscopic observation using fluorescent probe-labeled truncated rIlpAs (Fig. (Fig.6).6). On the other hand, two other truncated rIlpA polypeptides, rIlpA#1 and rIlpA#2, did not show any binding to host cells. Therefore, the extracellular domains of IlpA may be accessible to interaction with the host cell surface, and some of the secondary structures in the extracellularly exposed rIlpA#3 possibly participate in binding to the specific host receptor, which has not yet been identified.
Using recombinant PsaA protein covalently bound to fluorescent spheres, E-cadherin was identified as a host cellular receptor for PsaA of S. pneumoniae (1). A Campylobacter jejuni adhesin, JlpA lipoprotein, was found to trigger inflammatory/innate immune responses in host cells during C. jejuni infection via direct interaction with surface-exposed heat shock protein 90α and subsequent activation of NF-κB and p38 mitogen-activated protein (MAP) kinase in host cells (13). A family of opacity outer membrane proteins (Opa proteins) plays a significant role during cytoadherence of Neisseria gonorrhoeae to human cells (20). Their interaction with host cells is mediated by several cellular components, including heparin sulfate proteoglycan, extracellular matrix components, and CD66 family receptors (4). Interaction between Opa and these host components activates a signaling pathway resulting in transcellular transversal of this pathogen across human cells (34).
In addition to IlpA, two adhesins have been documented for V. vulnificus, including OmpU (7) and PilA (26). Further investigation using V. vulnificus strains deficient in multiple loci of these adhesin genes will be needed to evaluate the contribution of each adhesin to overall cytoadherence of V. vulnificus to diverse cell types. Due to a lack of this information, it is premature to suggest that each adhesin functions with a specific cell type or broad range of cell types. Human epidermoid carcinoma (HEp-2) cells were used for studies of OmpU (7) and PilA (26), whereas the INT-407 cell line, derived from human intestinal epithelial cells, has been used for IlpA. However, the effects of ilpA mutation on adherence and cytotoxicity are not confined to the specific host cell line used in this study, since the same effects were also observed in similar experiments using HEp-2 cells (Y.-S. Han, S.-J. Park, and K.-H. Lee, unpublished data). Fibronectin, especially its RGD repeat, was found to be a receptor for the OmpU protein in V. vulnificus (7) as well as in Vibrio cholerae (32). However, no information is available on a receptor for V. vulnificus PilA.
Among the three adhesins of V. vulnificus, only IlpA has been shown to stimulate the production of proinflammatory cytokines in human monocytes. However, the possibility that the other adhesins of V. vulnificus have immunostimulating activity cannot be excluded. A V. cholerae ompU mutant was defective in activation of a proinflammatory response (2), and the V. vulnificus outer membrane protein fraction, including OmpU, was effective at eliciting immunization of mice against V. vulnificus (15). In addition, a Pseudomonas aeruginosa pilA mutant induced less secretion of cytokines and showed a lack of neutrophil chemotaxis in a murine model (22).
As mentioned above, IlpA stimulates cytokine production from human blood cells, and TLR2 of the host cells is a critical component for this process (8). A direct association between IlpA and TLR2 has not been proven and needs to be verified by further investigation. Similarly, further study is required to identify the host receptor molecule(s) for the adhesion domain of IlpA and to elucidate subsequent signaling processes in host cells. Therefore, we are currently screening for a host receptor(s) for V. vulnificus IlpA. The study of adhesin-receptor interactions will elucidate the complex interactive networks of diverse adhesins of V. vulnificus during the initial pathogenicity process, in which each adhesin might interact with unique or common receptor molecules to efficiently connect the required steps for successful cytoadherence to and entry into specific cell types.
This study was supported by the Basic Science Research Program through the National Research Foundation (NRF), funded by the Ministry of Education, Science and Technology (grant 2009-0070681 to K.-H.L.), and partly by the Gyeonggi Regional Research Center (GRRC) program (Protein Research Center for Bio-Industry) of Gyeonggi Province, Republic of Korea.
Editor: S. R. Blanke
Published ahead of print on 22 March 2010.