ABSTRACT
During infectious processes, antimicrobial proteins are produced by both epithelial cells and innate immune cells. Some of these antimicrobial molecules function by targeting transition metals and sequestering these metals in a process referred to as “nutritional immunity.” This chelation strategy ultimately starves invading pathogens, limiting their growth within the vertebrate host. Recent evidence suggests that these metal-binding antimicrobial molecules have the capacity to affect bacterial virulence, including toxin secretion systems. Our previous work showed that the S100A8/S100A9 heterodimer (calprotectin, or calgranulin A/B) binds zinc and represses the elaboration of the H. pylori cag type IV secretion system (T4SS). However, there are several other S100 proteins that are produced in response to infection. We hypothesized that the zinc-binding protein S100A12 (calgranulin C) is induced in response to H. pylori infection and also plays a role in controlling H. pylori growth and virulence. To test this, we analyzed gastric biopsy specimens from H. pylori-positive and -negative patients for S100A12 expression. These assays showed that S100A12 is induced in response to H. pylori infection and inhibits bacterial growth and viability in vitro by binding nutrient zinc. Furthermore, the data establish that the zinc-binding activity of the S100A12 protein represses the activity of the cag T4SS, as evidenced by the gastric cell “hummingbird” phenotype, interleukin 8 (IL-8) secretion, and CagA translocation assays. In addition, high-resolution field emission gun scanning electron microscopy (FEG-SEM) was used to demonstrate that S100A12 represses biogenesis of the cag T4SS. Together with our previous work, these data reveal that multiple S100 proteins can repress the elaboration of an oncogenic bacterial surface organelle.
INTRODUCTION
Helicobacter pylori is a Gram-negative member of the Epsilonproteobacteria commonly found in the human stomach (1). More than 50% of the world's population is chronically infected with H. pylori, despite a robust immune response to this bacterium (2, 3). The vast majority of infected persons exhibit no symptoms of disease; however, the presence of this pathogen can increase the risk of negative outcomes, including duodenal ulcer, dysplasia, mucosa-associated lymphoid tissue (MALT) lymphoma, and invasive gastric cancer (4). Negative outcomes of H. pylori infection are associated with multiple factors, including host genetics, environmental contributions such as diet and smoking, and bacterial virulence properties.
One major virulence factor that contributes to H. pylori-associated disease outcomes is the cag pathogenicity island-encoded type IV secretion system (cag T4SS) (5). The cag T4SS is a macromolecular nanomachine responsible for translocating substrates, including peptidoglycan and the oncogenic effector molecule CagA, into host epithelial cells (6). The translocation activity of the cag T4SS leads to multiple consequences in host cell biology, including cytoskeletal rearrangements, induction of carcinogenic cell-signaling cascades, alteration of DNA methylation, perturbation of metal homeostasis, and secretion of proinflammatory molecules (6–9).
Transcription and secretion of the cytokine interleukin 8 (IL-8) are dramatically induced by the activity of the cag T4SS (10). IL-8 is a chemokine that recruits innate immune cells such as polymorphonuclear leukocytes (neutrophils) to discrete locations within inflamed tissue (11). Once present, neutrophils perform phagocytosis, elaborate extracellular traps, and produce antimicrobial molecules in an effort to control or eliminate invading pathogens (12). Some antimicrobial molecules produced by innate immune cells include reactive oxygen species, reactive nitrogen species, and antimicrobial proteins (12, 13).
Three S100 EF-hand calcium-binding proteins (S100A8, S100A9, and S100A12) constitute a major class of antimicrobial proteins termed “calgranulins” (14). S100 proteins are homodimeric proteins that also form S100 protein heterodimers. In addition to binding Ca2+ in EF-hand motifs, S100 proteins are distinguished by binding transition metal ions, including Zn2+, Cu2+, and Mn2+, at the dimer interface (15). S100 proteins mediate a variety of intracellular processes, including signaling, proliferation, differentiation, and motility (16), but are also exported from cells. Among their extracellular activities, they exert an antimicrobial activity via chelation of nutrient metals (17). Transition metals are required for all living organisms, and therefore, sequestration of these metals by S100 proteins prevents microbial access to these vital nutrients, thus limiting the growth of the invading pathogen (18). This mechanism of antimicrobial activity constitutes a form of nutritional immunity. Numerous pathogens, including Salmonella enterica serovar Typhimurium, Staphylococcus aureus, Escherichia coli, Borrelia burgdorferi, Listeria monocytogenes, Candida albicans, Acinetobacter baumannii, Staphylococcus epidermidis, Staphylococcus lugdunensis, Enterococcus faecalis, Pseudomonas aeruginosa, and Shigella flexneri, have been shown to be susceptible to the antimicrobial activity of the S100A8/S100A9 heterodimer calprotectin (19–26). Our recent work demonstrated that calprotectin represses the growth and virulence of H. pylori by sequestering nutrient zinc (27). However, the contribution of additional S100 proteins to growth and virulence repression in H. pylori remains unknown.
There are over 25 S100 proteins (17). Previous research has revealed expression of the three calgranulin proteins (S100A8, S100A9, and S100A12) within the gastric mucosas of a small group of H. pylori-infected children (28). Interestingly, S100A12 (calgranulin C, EN-RAGE) has also been identified as a potential biomarker for localized gastrointestinal inflammation and is reported to have calcium-, zinc-, and copper-binding properties (29, 30). S100A12, like S100A8 and S100A9, is highly expressed in innate immune cells, including macrophages and neutrophils (30). Here, we report that the level of S100A12 is elevated in response to H. pylori infection and that the protein localizes to neutrophilic infiltrates within the gastric tissue. S100A12 exhibits the ability to sequester zinc, resulting in repression of H. pylori growth and viability. Concomitantly, H. pylori cells exposed to S100A12-dependent zinc starvation have diminished cag T4SS activity and pilus biogenesis, a result that is reversed by the addition of an exogenous source of nutrient zinc. This report demonstrates the antibacterial activity of S100A12 via nutritional immunity and its role in affecting regulation of an important bacterial virulence factor.
MATERIALS AND METHODS
Bacterial strains, cell lines, and culture conditions.H. pylori strain PMSS1 (a clinical strain with a functional cag T4SS), strain 7.13 (a gerbil-adapted clinical isolate with a functional cag T4SS) used as the parental wild-type (WT), and a 7.13 ΔcagE isogenic mutant, harboring a kanamycin resistance cassette (aphA) inserted within the coding region using previously described methods (10), were used for these studies. Bacterial strains were grown on tryptic soy agar plates supplemented with 5% sheep blood (blood agar plates) or in brucella broth supplemented with cholesterol at 37°C in ambient air containing 5% carbon dioxide. For selection of ΔcagE mutants, blood agar plates were supplemented with 40 μg/ml of kanamycin.
AGS human gastric adenocarcinoma cells (ATCC) were cultured to 60 to 70% confluence in RPMI medium supplemented with 2 mM l-glutamine, 10% fetal bovine serum (FBS), and 10 mM HEPES buffer at 37°C in ambient air containing 5% carbon dioxide.
For growth and viability analyses, bacteria were subcultured in 60% brucella broth plus 40% calprotectin buffer (31) supplemented with cholesterol, referred to as “medium alone,” or medium supplemented with increasing concentrations of purified S100A12 protein alone or with 100 μM zinc chloride or 100 μM copper chloride. Bacterial growth was evaluated at 4 h or 24 h postinoculation by spectrophotometric reading of the optical density at 600 nm (OD600). Bacterial viability was evaluated at 4 h or 24 h by serial dilution and plating onto blood agar plates and quantifying viable CFU per milliliter of culture.
For coculture assays, bacteria were subcultured in 60% brucella broth plus 40% calprotectin buffer supplemented with cholesterol, referred to as “medium alone” (27), or medium supplemented with 500 μg/ml (23.9 μM) purified S100A12 protein alone or with 100 μM zinc chloride or 100 μM copper chloride. For zinc titration experiments, bacteria were grown in the presence of 500 μg/ml (23.9 μM) purified S100A12 protein alone or supplemented with 1.25 μM, 2.5 μM, 5 μM, 10 μM, or 25 μM zinc chloride. Bacteria were grown at 37°C in ambient air containing 5% carbon dioxide overnight before being applied to AGS cells at a multiplicity of infection (MOI) of 100:1 (as determined by OD600 reading and plate counts of viable CFU/milliliter) for 4 to 6 h of coculture.
Purification of S100A12 protein.The s100A12 gene was cloned into the pET1120 vector as previously described (32). Overexpression of the gene product (tagless and without any additional amino acids present in the protein sequence) was achieved in E. coli BL21 grown in Luria broth by inducing expression with isopropyl-thio-β-d-galactoside for 6 h. Cells were lysed by sonication in 50 mM Tris-HCl (pH 8.0). The crude extract was centrifuged at 50,000 × g for 20 min. The supernatant was supplemented with 2 mM CaCl2 and then loaded onto a phenyl Sepharose column (5-ml volume; GE Life Sciences) equilibrated with 50 mM Tris (pH 8.0), 2 mM CaCl2. The column was washed with 20 volumes of the same buffer, and bound S100A12 protein was eluted with 50 mM Tris-HCl (pH 8.0), 5 mM EDTA. The S100A12 obtained was concentrated by ultrafiltration and loaded onto a Superdex 75 column (Amersham, GE Healthcare) equilibrated with 20 mM Tris-HCl, 150 mM NaCl (pH 8.0), and 1 mM EDTA. Fractions containing S100A12 were identified by SDS gel electrophoresis, desalted, and concentrated using Ultrafree-4 5-kDa-cutoff centrifugal filters (Millipore), and aliquots of protein were stored at −80°C until used in culture assays. Protein concentration was determined by monitoring the UV absorbance at 280 nm using an extinction coefficient of 2,980 M−1 cm−1. Typical yields of S100A12 were 25 mg per liter of culture.
Human subjects and ethics statement.Gastric tissues in this study were derived from subjects enrolled according to protocols approved by the Institutional Review Board of Louisiana State University Health Sciences Center and the Institutional Review Committee of Memorial Medical Center at New Orleans as previously described (33). Briefly, human biopsy specimens were collected from adult patients that underwent upper gastrointestinal endoscopy due to dyspeptic symptoms. Three gastric biopsy specimens (antrum, incisura angularis, and corpus) were obtained for histology from each subject. For this study, we included 5 subjects who had normal histology and were negative for H. pylori and 10 subjects who were H. pylori positive and had histological diagnosis of nonatrophic gastritis. None of the subjects had a history or presence of peptic ulcers or gastric cancer. These human gastric biopsy specimens were stratified into H. pylori positive and H. pylori negative based on Steiner silver staining to detect the presence of H. pylori as previously described (33).
Immunohistochemical analyses of S100A12 within gastric tissue.Human gastric biopsy samples were derived from 15 subjects (10 positive for H. pylori, 5 negative for H. pylori) and fixed in 10% neutral buffered formalin prior to embedment in paraffin. One antral biopsy specimen per subject was used for immunohistochemistry (IHC) studies. Tissue was cut into 5-μm sections, and multiple sections were placed on each slide for analysis. After quenching with 0.03% hydrogen peroxide for 20 min at room temperature (RT), tissues were treated with a heat-induced epitope retrieval solution (universal decloaker; Biocare Medical, Concord, CA) using a pressure cooker at 121°C for 20 min and allowed to cool at RT prior to blocking with 10% normal goat serum in 0.1 M PBS (pH 7.4). Primary rabbit polyclonal antibody to S100A12 (ab37657; Abcam) or rabbit monoclonal antibody to human calprotectin subunit S100A8 (MRP8) (ab92331; Abcam) was applied for 1 h. Detection of primary antibody was performed using a rabbit horseradish peroxidase (HRP)-polymer system for 30 min and developed with 3,3′-diaminobenzidine tetrahydrochloride (H-DAB) (Dako, Carpinteria, CA). The sections were counterstained with hematoxylin, rinsed, dehydrated, and mounted with Cytoseal XYL before light microscopy analysis was performed. Tissues were evaluated and scored for S100A12 staining in a blinded fashion. A score of 0 was conferred if no S100A12 was detected, 1 indicated low levels of S100A12 staining, 2 indicated moderate S100A12 staining, and 3 indicated abundant S100A12 staining. Microscope images at a magnification of ×1,000 were obtained from tissues stained for S100A12 or S100A8. Five images per case and antibody were obtained in areas of the mucosa with the highest concentration of positive cells. Micrograph images were analyzed with the ImageJ IHC toolbox plugin to quantify H-DAB staining by color detection and converted to 8-bit format, and densitometry quantification was carried out as previously described (27).
Evaluation of secreted IL-8.AGS cells were cocultured with H. pylori as described above, and coculture supernatants were collected and centrifuged at 8,000 × g to remove any cells or cellular debris. Secreted IL-8 was evaluated by Quantikine human IL-8 sandwich enzyme-linked immunosorbent assay (ELISA) (R&D Systems) per the manufacturer's instructions. IL-8 values were compared to those derived from coculture samples from WT bacteria grown in medium alone as previously described (27).
For evaluation of the host signaling pathway, AGS cells were treated with 25 μg/ml of purified S100A12 (an amount comparable to the maximum that could be carried over from bacterial growth assays) alone or in the presence of 10 ng/ml of purified tumor necrosis factor alpha (TNF-α), or with H. pylori grown as described above in the presence of 500 μg/ml (23.9 μM) of S100A12 alone or in the presence of 10 ng/ml of purified TNF-α for 4 h. As a control, AGS cells alone were treated with 10 ng/ml of purified TNF-α for 4 h concomitantly. Supernatants were collected and processed as described above, and IL-8 values were compared to those derived from untreated AGS cells.
CagA translocation assays.Cocultures of gastric epithelial cells and H. pylori were performed as described above, and cells were scraped into 1 ml radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors and sodium orthovanadate to prevent dephosphorylation of phospho-CagA, centrifuged to collect cells, and resuspended in 100 μl of RIPA buffer containing protease inhibitors and sodium orthovanadate. Total protein concentration was determined by Bradford assay, and 10 μg of protein was separated by SDS-PAGE, transferred to nitrocellulose, and blotted with either a mouse monoclonal antibody to phosphotyrosine 99 (p-Tyr; Santa Cruz), a rabbit polyclonal antibody to CagA (34), or a rabbit polyclonal antibody to total H. pylori lysate (34). Secondary detection was performed using Odyssey infrared detection antibodies (goat anti-mouse or goat anti-rabbit) and the Odyssey imaging system. Densitometry analyses was performed using ImageJ software.
Evaluation of bacterial adherence.Bacterial adherence was evaluated as previously described (27). Briefly, cocultures of 105 AGS gastric epithelial cells and H. pylori (MOI of 100:1) were performed as described above, and cells were scraped into 1 ml of sterile phosphate-buffered saline (pH 7.4), serially diluted, and plated onto blood agar plates. Plates were incubated at 37°C for 2 to 3 days, and resulting colonies were counted to evaluate the number of CFU/105 AGS gastric epithelial cells.
FEG-SEM analyses.Field emission gun scanning electron microscopy (FEG-SEM) sample preparation was performed as previously described (27). Briefly, cocultures of H. pylori and AGS cells grown on poly-l-lysine-treated coverslips were fixed with 2.0% paraformaldehyde, 2.5% glutaraldehyde in 0.05 M sodium cacodylate buffer for 1 h at room temperature. Samples were washed three times with 0.05 M sodium cacodylate buffer before secondary fixation with 1% osmium tetroxide. Samples were sequentially dehydrated by washing with increasing concentrations of ethanol before being dried at the critical point, mounted on SEM stubs and sputter-coated with 20 nm of gold-palladium. A thin line of colloidal silver paint was applied at the sample edge to facilitate grounding of the coverslip and prevent charging during FEG-SEM imaging. Samples were visualized with an FEI Quanta 250 FEG-SEM at high vacuum, and micrographs were analyzed with Image J software. For host cell morphology evaluation, all micrographs were collected at a magnification of ×383. For evaluation of the cag T4SS pilus, images were collected at high magnification (20,000× or above).
Statistical analyses.Statistical analyses of bacterial growth and viability, adherence, IL-8 secretion, and hummingbird quantification were performed using unpaired Student's t test. Pilus quantifications were performed using paired two-tailed Student's t test and one-way analysis of variance (ANOVA). Quantification of immunohistochemical scoring was analyzed by the Mann-Whitney U test, and H-DAB signal was analyzed by one-way ANOVA and unpaired Student's t test with Welch's correction. All data analyzed in this work were derived from at least three separate biological replicates. Statistical analyses were performed using GraphPad Prism software and Microsoft Excel.
RESULTS
S100A12 is elevated in gastric biopsy specimens from H. pylori-infected humans compared to those from uninfected humans.Previous reports have demonstrated that levels of S100A-family proteins, including S100A8 and S100A9, are elevated in response to H. pylori infection (27). In order to determine if S100A12 is induced in response to H. pylori infection, human gastric biopsy samples from H. pylori-infected and H. pylori-negative individuals were subjected to immunohistochemical staining methods to evaluate S100A12 localization within the tissue and its abundance in comparison with the calprotectin subunit S100A8 (MRP8) (Fig. 1). Microscopic analyses of these tissues revealed that S100A12 staining is more prevalent in H. pylori-infected gastric tissue than in uninfected tissue. S100A12 staining was observed primarily in the cytoplasm of polymorphonuclear leukocytes and occasional mononuclear leukocytes (likely macrophages). Tissues were evaluated and scored (0 to 3) for the abundance of S100A12 staining. Human gastric tissue infected with H. pylori exhibited a significant increase in S100A12 score compared to uninfected tissues (P = 0.0027). Quantification of H-DAB by IHC plugin and densitometry analyses revealed that both MRP8 and S100A12 staining were significantly elevated in H. pylori-infected tissues compared to uninfected tissues (P < 0.0001, ANOVA) and that S100A12 staining was comparable to MRP8 staining (P = 0.4424, unpaired Student's t test with Welch's correction). These data indicate that S100A12 is produced by polymorphonuclear cells in response to H. pylori infection and that the abundance of S100A12 is comparable to that of MRP8 (S100A8).
Host S100A12 levels are elevated and comparable to MRP8 levels in H. pylori-infected human stomach tissue as opposed to non-H. pylori-infected stomach tissue. Immunohistochemical (IHC) microscopy evaluation of human gastric biopsy specimens reveals elevated S100A12 levels associated with polymorphonuclear cell infiltrates within H. pylori-infected humans (A to C) compared to uninfected humans (G to I). IHC evaluation of H. pylori-infected human biopsy specimens reveal elevated calprotectin subunit S100A8 (MRP8). Microscopic analyses of tissues subjected to immunohistochemical staining procedures were performed at magnifications of ×200 (A, D, and G), ×400 (B, E, and H), and ×1,000 (C, F, and I). Arrows indicate polymorphonuclear innate immune cells within the gastric tissue, staining positive for S100A12 (C) or MRP8 (F). Representative micrographs are shown (n = 5 to 10). (J) Representative micrographs derived from two groups (H. pylori-infected samples and H. pylori-negative samples) were scored for S100A12 staining (dots indicate scores for each individual; horizontal lines indicate mean score per group). A score of 0 indicates no detectable S100A12, 1 indicates low preponderance of S100A12 staining, 2 indicates moderate presence of S100A12 stain, and 3 indicates abundant S100A12 staining within the gastric tissue. Circles indicate uninfected humans, and squares indicate H. pylori-infected humans. Human gastric tissue infected with H. pylori exhibited a significant increase in S100A12 scoring compared to uninfected tissues (P = 0.0027, Mann-Whitney U analysis; n = 5 biological samples for H. pylori-negative samples, n = 10 for H. pylori-positive samples). H-DAB stain detection and quantification reveal that both MRP8 and S100A12 levels are significantly higher in H. pylori-infected individuals than uninfected individuals (P = 0.0001, one-way ANOVA) and that the levels of S100A12 are comparable to those of MRP8 (P = 0.4424, Student's t test with Welch's correction).
S100A12 decreases growth and viability of H. pylori via zinc chelation.S100A12 has previously been reported to have antimicrobial properties, including the ability to chelate nutrient zinc (35, 36), and sequestration of zinc by S100 proteins has been implicated in repressing H. pylori growth and viability in vitro (27). Therefore, we hypothesized that S100A12 could repress H. pylori growth and viability through zinc chelation. To test this, growth and viability of H. pylori were assessed by culturing bacteria in increasing physiologically relevant concentrations of purified S100A12 protein in the presence or absence of an exogenous source of nutrient zinc. Calgranulins are present within neutrophils at concentrations in the microgram-per-microliter range and can exceed 1 mg/ml (22, 44). Growth was analyzed by spectrophotometric reading at OD600, and viability was analyzed by serial dilution of overnight cultures and plating onto bacteriological media for colony counting to evaluate CFU per ml of culture. In the presence of 750 μg/ml S100A12 protein, H. pylori growth was repressed by 40% and viability was decreased 1.5 log units compared to cells grown in medium alone (P = 0.005 and P = 0.042, respectively) (Fig. 2), results that were statistically significant. Furthermore, in the presence of 1,000 μg/ml of S100A12 protein, H. pylori growth was repressed by 51% and viability was decreased nearly 3 log units compared to cells grown in medium alone (P = 0.005 and P = 0.042, respectively), indicating that S100A12 imposes a dose-dependent growth restriction on H. pylori. The addition of an exogenous source of nutrient zinc ablated the antibacterial activity of S100A12, indicating the repression of growth is dependent upon the sequestration of zinc.
The S100A12 protein inhibits bacterial growth and viability in a dose-dependent manner. WT H. pylori was cultured for 24 h in medium alone (circles) or medium supplemented with 100 μM zinc chloride (squares) plus increasing concentrations of purified S100A12 protein. (A) Bacterial growth was evaluated by spectrophotometric OD600 reading and determining the percent growth compared to controls grown in medium alone. (B) Bacterial viability was evaluated by serial dilution and plating on bacteriological medium to evaluate viable CFU per ml. *, P < 0.05, **, P < 0.01, ***, P < 0.001 (Student's t test; n = 3 biological replicates).
S100A12 inhibits cag T4SS-dependent IL-8 secretion by host cells.The cag T4SS is regulated by the availability of nutrient zinc (27), and since S100A12 represses bacterial growth via zinc sequestration, we hypothesized that it would also have the capacity to repress the function of the cag T4SS. Because one of the consequences of an active cag T4SS is induction of proinflammatory signaling cascades that ultimately lead to host cell IL-8 secretion (10), we chose to test our hypothesis by collecting supernatants from cocultures and analyzing them for IL-8 secretion using a sandwich ELISA (Fig. 3). WT H. pylori strain 7.13 bacteria exposed to 500 μg/ml (23.9 μM) of purified S100A12, a concentration that is subinhibitory for bacterial growth, prior to coculture with AGS cells elicited 39% less IL-8 than bacteria grown in medium alone, a result that was statistically significant (Fig. 3A) (P = 0.047 compared to WT grown in medium alone). Bacteria grown in the presence of S100A12 plus an exogenous source of nutrient zinc restored IL-8 secretion to levels higher than those obtained with bacteria grown in medium alone (Fig. 3A) (P = 0.0435 compared to WT grown in medium alone). An isogenic ΔcagE mutant (which lacks the major ATPase that drives cag T4SS assembly) (10) had significantly reduced IL-8 induction compared to the WT but exhibited no change in IL-8 secretion induction whether grown in medium alone or medium supplemented with S100A12 alone or in conjunction with an exogenous nutrient zinc source (Fig. 3A), indicating that the IL-8 secretion changes seen are dependent upon an intact and functional cag T4SS.
Exposure to S100A12 protein inhibits H. pylori cag T4SS-dependent host IL-8 secretion. Bacteria were exposed to either medium alone or medium supplemented with S100A12 (500 μg/ml = 23.9 μM) alone or in the presence of an exogenous zinc source (1.25 to 50 μM zinc chloride) prior to coculture with gastric cells. IL-8 secretion by human gastric epithelial cells was quantified using an IL-8 ELISA. (A) The wild-type parental strain of H. pylori 7.13 and the isogenic ΔcagE mutant, which lacks a functional cag T4SS, were grown in medium alone or medium supplemented with 500 μg/ml of the S100A12 protein alone or in the presence of 100 μM zinc chloride prior to coculture with AGS cells. Bars represent the means and SEM for each group (n = 3 biological replicates). *, P < 0.05 (Student's t test) compared to WT grown in medium alone. (B) Bacteria exposed to 500 μg/ml (23.9 μM) of S100A12 and increasing concentrations of zinc chloride. Bars represent the means and SEM for each group (n = 3 biological replicates). *, P < 0.05 (Student's t test) compared to WT grown in medium supplemented with 500 μg/ml S100A12.
The addition of 25 μM zinc chloride partially restored induction of IL-8 by bacteria grown in the presence of 500 μg/ml (23.9 μM) S100A12, and the addition of 50 μM zinc chloride completely restored IL-8 induction, indicating that a concentration slightly higher than molar equivalent of zinc chloride was sufficient to restore cag T4SS activity, as determined by comparison to samples grown in the presence of S100A12 alone (Fig. 3B) (P < 0.0001). The IL-8 secretion repression in S100A12-treated bacteria was independent of bacterial adherence to host AGS cells (see Fig. S1 in the supplemental material), but interestingly, the addition of excess exogenous zinc (100 μM) and S100A12 caused a 1.4-fold increase in bacterial adherence, a result that was statistically significant (P = 0.04), indicating that excess zinc availability or zinc stress could modulate bacterial adherence within the gastric niche. Repression of the activity of the cag T4SS by S100A12 was recapitulated with an additional strain of H. pylori (PMSS1), indicating that this phenotype is not strain-specific (see Fig. S2 in the supplemental material). Also, treatment of AGS cells with S100A12 did not alter IL-8 secretion compared to that in untreated cells (P = 0.1296), and treatment of AGS cells with either S100A12 or H. pylori grown in the presence of S100A12 plus purified TNF-α significantly induced IL-8 secretion compared to that in samples without TNF added (P = 0.0033 and P = 0.0132, respectively), indicating that the signaling pathway for IL-8 induction is intact and not perturbed by S100A12 directly (see Fig. S3 in the supplemental material).
Exposure to S100A12 represses cag T4SS-dependent changes in cell morphology.As a consequence of cag T4SS activity, CagA is translocated into host cells and interacts with Src-tyrosine kinase, causing host cells to undergo cytoskeletal rearrangements, including a scatter factor-like “hummingbird” phenotype, characterized by an elongation of the epithelial cell (38, 39). We hypothesized that the cag T4SS-dependent hummingbird phenotype would be attenuated in samples with bacteria exposed to S100A12. To test this, bacteria were grown for 24 h in the presence of purified S100A12 protein at concentrations that were subinhibitory for bacterial growth prior to coculture with AGS human gastric cells (500 μg/ml, 23.9 μM). Cocultures were evaluated by low-magnification (383×) scanning electron microscopy to determine host cell morphology changes (Fig. 4). In uninfected AGS samples, less than 2% of cells demonstrated the hummingbird phenotype, compared to an average of 36% of AGS cells infected with H. pylori grown in medium alone. H. pylori exposed to S100A12 prior to coculture induced the hummingbird phenotype at a lower incidence (11%) than bacteria grown in medium alone (P = 0.009), a result that was reversed by the addition of an exogenous source of nutrient zinc and statistically higher than the control with medium alone (50%; P = 0.0479, unpaired Student's t test). Furthermore, bacteria grown in medium alone exhibited the largest dynamic range of hummingbird cells per field (33 to 126 cells per field), compared to bacteria exposed to S100A12 (4 to 44 cells per field) and bacteria exposed to S100A12 plus zinc chloride (29 to 88 cells per field) (P = 0.05, ANOVA). Together, these results indicate that the zinc sequestration imposed by S100A12 inhibits the activity of the cag T4SS, specifically repressing translocation of substrate into host cells, which leads to cytoskeletal rearrangements.
Pretreatment with S100A12 protein represses H. pylori cag T4SS-dependent host cell morphology changes. H. pylori 7.13 was grown in medium alone or medium supplemented with 500 μg/ml S100A12 protein alone or S100A12 plus 100 μM zinc chloride prior to coculture with AGS cells. To evaluate cag T4SS activity, host cell morphology was analyzed by low-magnification (×383) electron microscopy. (A) Uninfected AGS cells exhibit typical round cellular morphology. (B) AGS cells cocultured with H. pylori grown in medium alone exhibit elongated hummingbird morphology, a consequence of cag T4SS activity. (C) Coculture of AGS cells with H. pylori exposed to S100A12 prior to coculture results in fewer hummingbird cells. (D) Coculture of AGS cells with H. pylori exposed to S100A12 plus zinc chloride restores the hummingbird phenotype. Arrows indicate cells exhibiting hummingbird morphology. Bars, 100 μm. (E) Whisker box plots represent the upper and lower quartile (box), with the center line indicating the mean and the whiskers illustrating the highest and lowest values for number of hummingbird cells per field. (F) Bars indicate the percentage of cells exhibiting the hummingbird phenotype for each field, an analysis that controls for cell density within each replicate (n = 3 to 5 biological replicates, 500+ cells evaluated for uninfected, and at least 1,400 cells evaluated for each of the H. pylori-infected groups). *, P < 0.05 compared to WT grown in medium alone (Student's t test).
S100A12 represses the translocation and phosphorylation of CagA.The observation that S100A12 exposure represses the ability of H. pylori to induce both IL-8 secretion and the hummingbird phenotype in host cells also indicates that CagA translocation could be inhibited in these cocultures. To test this, we employed CagA translocation assays as previously described (38, 39). CagA is phosphorylated by host kinases upon translocation, a consequence that can be detected using a monoclonal p-Tyr antibody (38). Bacteria were grown for 24 h in the presence of concentrations of purified S100A12 protein that were subinhibitory for bacterial growth prior to coculture with AGS human gastric cells. Proteins were collected from the coculture and analyzed by immunoblotting. The p-Tyr antibody was used to evaluate the total amount of phosphorylated CagA compared to total CagA (which was evaluated by a polyclonal rabbit antibody to CagA) associated with host cells. Bacteria exposed to purified S100A12 exhibited decreased CagA translocation (1.8-fold; P = 0.0158) compared to bacteria grown in medium alone prior to coculture with gastric cells (Fig. 5). Interestingly, this defect in CagA phosphorylation was rescued when an exogenous source of zinc was added, a result that was statistically indistinguishable from those derived from bacteria grown in medium alone (P = 0.284). These data indicate that S100A12-associated zinc sequestration represses the translocation of CagA into host cells.
Exposure to S100A12 inhibits CagA translocation into gastric epithelial cells. Analysis of CagA phosphorylation upon translocation into AGS cells was determined by immunoblotting with anti-p-Tyr (α-P-Tyr) and anti-CagA (α-CagA). Total bacterial proteins were evaluated with a rabbit polyclonal antibody to total H. pylori antigen as a loading control (α-Hp). (A) Bacteria were grown in medium alone or medium supplemented with 500 μg/ml S100A12 either alone or in the presence of 100 μM zinc chloride prior to coculture with AGS cells. (B) Densitometry analyses of immunoblots were performed to quantify the percentage of translocated (phosphorylated) CagA in relation to total CagA. Immunoblotting with anti-H. pylori was also performed and used as a loading control. Exposure to S100A12 protein diminished CagA translocation into host cells, a result that was reversed by supplementation with 100 μM zinc chloride. Bars represent the means and SEM for each group (n = 3 biological replicates). *, P < 0.05 compared to medium alone (Student's t test).
Zinc sequestration by S100A12 represses the biogenesis of the cag T4SS.The cag T4SS has an extracellular organelle termed the pilus, which is formed at the host-pathogen interface (37). The H. pylori cag T4SS pilus has previously been shown to be negatively regulated by the S100-family protein calprotectin through a zinc sequestration-dependent pathway (27). To determine if S100A12 has a similar activity on the biogenesis of the cag T4SS, high-resolution field emission gun scanning electron microscopy (FEG-SEM) was utilized to evaluate pilus formation at the host-pathogen interface (Fig. 6). Bacteria grown in medium alone formed an average of 5 pili per cell, and 81% of cells were piliated, while bacteria exposed to S100A12 prior to coculture formed less than one pilus per cell (P < 0.001) and fewer piliated cells total (17%) than those in medium alone (P = 0.0005). Bacteria grown in medium supplemented with both S100A12 and exogenous zinc formed an average of 4 pili per cell and 70% piliated cells, results that were statistically indistinguishable from those obtained with medium-alone samples (P = 0.184 and P = 0.192, respectively). These results were recapitulated with an additional strain of H. pylori (PMSS1), which demonstrated that the S100A12 repression of pilus formation is not strain specific (see Fig. S4 in the supplemental material). These results reveal that S100A12 represses the elaboration of H. pylori cag T4SS pili via zinc chelation.
H. pylori cag T4SS pilus formation is repressed by exposure to the S100A12 protein. High-resolution FEG-SEM analysis of H. pylori cocultured with AGS human gastric epithelial cells. Prior to coculture, bacteria were cultured in medium alone (A) or medium supplemented with 500 μg/ml S100A12 protein (B). S100A12 exposure results in diminished cag T4SS pilus biogenesis at the host-pathogen interface. (C) Bacteria grown in the presence of S100A12 plus 100 μM zinc chloride exhibited restored cag T4SS pilus formation to levels comparable to those obtained in medium alone. Arrows indicate cag T4SS pili formed at the host-pathogen interface. Bars, 500 nm. (D) Dot plot graph indicating enumeration of pili per bacterial cell as quantified from representative micrographs; (E) bar graph indicating the percentage of piliated cells within each field derived from three biological replicates, with at least 20 fields from each replicate (n = 60+ cells). *, P < 0.05 compared to bacteria grown in medium alone (Student's t test).
DISCUSSION
The calgranulin proteins S100A8, S100A9, and S100A12 have been associated with polymorphonuclear leukocytes (neutrophils) and monocytes (40). S100A8/9 (calprotectin) and S100A12 (calgranulin C), are damage-associated molecular pattern molecules (DAMPs) which are characterized by their activity as potent initiators of proinflammatory signaling cascades involved in innate and adaptive immunity (40). Specifically, S100A12 binds to the receptor for advanced glycation end products (RAGE) and induces proinflammatory signaling events (41). In respiratory epithelial cells, S100A12 induces MAPK pathways, NF-κB pathways, and MUC5AC (a major mucin protein) production (42). S100A12 also induces ERK signaling by binding RAGE, and NF-κB signaling via Toll-like receptor 4 (TLR-4), a result that is congruent with other published work demonstrating S100A12 TLR-4 signaling within human monocytes (42, 43). Together, these results indicate that S100A12 binds multiple partners and activates the immune system in more than one fashion.
Because calgranulins are abundant within neutrophils, the concentration of calprotectin within the cytosol of a neutrophil can exceed 1 mg/ml (44), and these proteins are released during inflammation. Calgranulins have been identified as potential biomarkers for a variety of gastrointestinal disorders, including irritable bowel disorder (IBD), colitis, and chronic diarrhea (29, 45, 46). Calprotectin has been used as a biomarker for gastrointestinal inflammation for more than a decade (47). However, recent evidence has emerged indicating that the accuracy of S100A12 as a fecal biomarker for inflammation is greater than that of calprotectin for certain gastrointestinal diseases, such as IBD (48). Chronic gastrointestinal inflammation is strongly linked to increased risk for gastrointestinal cancers; thus, the utility of noninvasive biomarkers (such as inflammation-associated molecules) for cancer screening has grown (49). S100A12 has been identified as a biomarker for colorectal cancer, lymphoblastic leukemia, lymphocytic leukemia, and hepatocellular carcinoma recurrence and metastasis (50–53). S100A12 has also been shown to interact with calcyclin-binding protein (CacyBP), a negative regulator of tumorigenesis that participates in reducing β-catenin signaling (54). As a consequence of S100A12 binding to CacyBP, β-catenin nuclear translocation is upregulated, leading to carcinogenesis. H. pylori infection is associated with an increased risk of gastric cancer and has been implicated in perturbation of β-catenin signaling (55). It is possible that S100A12, deposited at the site of H. pylori infection, binds CacyBP and induces β-catenin translocation to promote oncogenic effects in the stomach. Our work demonstrates that S100A12 is strongly upregulated in gastric tissue in response to H. pylori infection, a result that is congruent with previously published observations indicating that S100A12 proteins were absent in normal gastric mucosa as well as uninfected inflamed gastric mucosa but highly induced in the gastric mucosas of six H. pylori-infected children (28). Together, these data reveal that S100A12 could be used as a potential biomarker for H. pylori-related diseases.
Besides its proinflammatory properties and association with the innate immune response, S100A12 is elevated during various bacterial infections and possesses antimicrobial and antiparasitic activities (36, 56). For example, S100A12 is induced in response to E. coli or S. aureus infection of mammary tissue (57, 58). S100A12 is also induced in response to nontypeable Haemophilus influenzae and Streptococcus pneumoniae infection leading to acute otitis media (59). Patients with severe sepsis and infants with necrotizing enterocolitis (NEC) exhibit increased S100A12 levels (60, 61). Interestingly, elevated S100A12 expression was associated with a dysbiosis of the intestinal microbiome in infants afflicted with NEC (62), indicating that the antimicrobial activity of S100A12 can be exerted on resident microflora as well as invading pathogens.
The antimicrobial activity of S100A12 is related to its biochemical properties. S100A12 is an EF-hand calcium-binding protein that forms an antiparallel noncovalent homodimer (39) and binds calcium, zinc, and copper (14, 39, 63–65). Like other secreted S100 proteins, S100A12 can form higher-order oligomers in the high calcium concentrations outside cells, including tetramers and hexamers (39). In the presence of either calcium or zinc ions, the structural dynamics of S100A12 can change, leading to alternate quaternary structures and potential changes in biochemical activities (39, 65, 66). S100 proteins have recently been recognized as important players in host metal sequestration strategies that contribute to nutritional immunity (18, 20, 22). Calprotectin is the best-characterized of these proteins and has been shown to inhibit the growth of numerous bacteria and fungi (22, 25). However, recent work demonstrates that proteins isolated from human neutrophil cytosol, including calprotectin (but not S100A12), become carbonylated in the presence of reactive oxygen species generated by oxidative burst (67). This oxidative modification occurred at the S100A9 Cys2 and six histidine residues which have been associated with chelation of manganese and zinc (20). Interestingly, carbonylation of these residues inhibited the bacteriostatic activity of calprotectin, indicating that this posttranslational modification could be modulating the metal-binding nutritional immunity capacity of this protein (67). Thus, calprotectin and S100A12 could play similar roles but be active at different times during infection.
Our previously published results indicate that calprotectin represses H. pylori growth and viability via sequestration of nutrient zinc (27). Similarly, S100A12 repressed H. pylori growth and viability in a dose-dependent fashion, a result that was reversed by the addition of an exogenous source of nutrient zinc. S100A12 differs from calprotectin, because in bacterial-growth analyses, the inhibitory dose of S100A12 was higher than that of calprotectin (S100A12 = 750 μg/ml; calprotectin = 300 μg/ml) (27). One factor contributing to this observation is that S100A12 binds zinc with an average dissociation constant of 10 nM (39), while calprotectin binds 3-fold more tightly, with an average dissociation constant of 3.5 nm (site 1 = 3.4 nm; site 2 = 8.2 nm) (20). Thus, calprotectin binds zinc more avidly than S100A12 and may exert greater antimicrobial activity as a result. Calprotectin also binds manganese, a transition metal that is important in mediating the activity of bacterial defense against superoxide stress and subsequent immune evasion (20, 31). S100A12 binds copper in addition to zinc; however, supplementation with exogenous copper did not restore H. pylori growth under increasing concentrations of purified S100A12 protein (data not shown). Experiments where titrations of zinc chloride were provided indicate that 25 to 50 μM zinc reverses the activity of 500 μg/ml (23.9 μM) S100A12, indicating that slightly more than the molar equivalent of zinc is sufficient to restore the cag T4SS activity. Together, these results underscore the importance of nutrient zinc for the regulation of virulence as well as for viability and proliferation of H. pylori, observations that are congruent with previously published work using a chemically defined medium to determine nutrient requirements for growth (68). Although previous studies revealed that S100A12 is induced in response to infection and that this protein has antimicrobial properties (28, 36), our study demonstrates a role in nutritional immunity, specifically via zinc sequestration. It is possible that, as inflammation progresses and excess oxidative species accumulate, the calprotectin subunit S100A9 is carbonylated and S100A12 acts as an auxiliary zinc chelator to facilitate nutritional immunity.
Our previous work determined that the s100A8 and s100A9 genes, which encode the subunits of the heterodimer calprotectin, are induced in both human and murine gastric tissue as a consequence of H. pylori infection (27). That work also implicated the IL-17 receptor (IL-17R) as an important signaling molecule for the induction of calprotectin expression, as evidenced by the lack of elevated s100A8 and s100A9 expression in the IL-17R−/− mouse (27). We are unable to determine the contribution of IL-17R signaling to s100A12 expression due to the fact that the gene encoding S100A12 is not present in the mouse genome (69). However, previous results established that elevated expression of calprotectin is associated with neutrophils responding to H. pylori infection (27). Similarly, our current results confirmed that S100A12 expression is primarily associated with polymorphonuclear cell infiltrates in response to H. pylori infection of human gastric tissues, implying that neutrophils may respond to H. pylori infection by producing a repertoire of S100 proteins that can chelate nutrient zinc and reduce bacterial growth and viability.
In addition to affecting bacterial growth, zinc chelation has multiple consequences for bacterial cell biology. Calprotectin-mediated zinc sequestration alters H. pylori virulence by repressing the elaboration of the cag T4SS pilus and concomitantly abrogating the activity of the cag T4SS (27). S100A12-mediated zinc chelation results in numerous changes in H. pylori activity, including repression of CagA translocation, inhibition of host cell morphological changes (hummingbird phenotype), and repression of host cell IL-8 secretion. All three of these activities are associated with an intact functioning cag T4SS (6). These observations are consistent with high-resolution electron microscopy analyses, which revealed that the cag T4SS pilus biogenesis is repressed in cells exposed to S100A12, a result that is reversed by the addition of zinc chloride. Interestingly, the addition of an exogenous source of nutrient zinc restored both pilus biogenesis and CagA translocation to levels comparable to those observed in the control untreated sample. Furthermore, bacterial exposure to 100 μM zinc chloride with S100A12 prior to AGS coculture significantly induced both host cell IL-8 secretion and hummingbird morphology compared to samples cultured in medium alone, indicating that excess zinc or zinc stress induces cag T4SS activity. This result was mirrored by elevated bacterial adherence to AGS cells after exposure to both S100A12 and excess nutrient zinc (see Fig. S1 in the supplemental material), which could account for the increased cag T4SS activity, which is host cell contact dependent (37). The S100A12-dependent repression of cag T4SS pilus biogenesis does not coincide with any changes in bacterial adherence (see Fig. S1 in the supplemental material) but does correlate with the alteration of cag T4SS activity, a result that agrees with our previous observations of calprotectin-mediated repression of cag T4SS activity and pilus biogenesis (27). Interestingly, the amount of S100A12 protein required to exert a modulatory effect on the cag T4SS was higher (500 μg/ml) than the amount of calprotectin required to achieve the same effect (200 μg/ml) (27), a result that agrees with the growth and viability data. This result also supports a model in which S100A12 has decreased antibacterial activity as a result of its diminished ability to bind nutrient zinc compared to calprotectin. Furthermore, supplementation of an exogenous copper source in H. pylori cultures containing 500 μg/ml of S100A12 did not restore cag T4SS pilus biogenesis (data not shown). These data demonstrate that zinc is an important signal that regulates elaboration and activity of a major toxin secretion system.
Zinc is a required micronutrient for all forms of life. Zinc availability is tightly regulated within the vertebrate host, and thus, zinc homeostasis is gaining greater appreciation for its important role in infectious disease processes (70). Bacteria have evolved strategies to sense and respond to zinc to gain a competitive advantage against resident microbiota (71, 72). Additionally, zinc's regulatory effect on bacterial secretion has been demonstrated. Specifically, zinc stress activates the secretion of the EsxA toxin, a substrate of the type VII ESX-1 secretion system in Mycobacterium tuberculosis which arrests phagosome maturation, lyses cell membranes, and induces cell death (73, 74). Conversely, zinc negatively regulates the type III secretion system in enteropathogenic E. coli, the ZirTS antivirulence secretion pathway in Salmonella, as well as the type VI secretion system in Burkholderia mallei and Burkholderi pseudomallei (75–78). These results, together with our work showing that zinc chelation negatively regulates the H. pylori cag T4SS, implicate zinc as an important global signaling molecule, which is regulated via innate immune molecules, including calprotectin and S100A12 at the host-pathogen interface.
Conclusions.We propose a model (Fig. 7) in which H. pylori infection and cag T4SS activity induce the chemokine IL-8, which ultimately aids in recruiting neutrophils to the site of infection. Neutrophils deposit the S100A12 protein within the gastric niche, thereby chelating available nutrient zinc and starving H. pylori of this vital micronutrient. H. pylori growth, viability, and virulence are repressed as a consequence. Specifically, the zinc sequestration imposed by S100A12 results in decreased cag T4SS pilus biogenesis and CagA translocation, which ultimately inhibits host cell responses, including IL-8 secretion and morphological changes such as the hummingbird phenotype. The molecular mechanism for the repression of the cag T4SS remains unknown, but it could be attributed to changes in transcription of cag T4SS components or to enzymatic activity due to decreased availability of zinc as a cofactor. Ultimately, these changes downregulate the host immune response aiding immune evasion and bacterial persistence to promote chronic infection. Together, the zinc sequestration activity of calprotectin and S100A12 coupled with the bacterial ability to sense the low-zinc environment achieved by the innate immune system could represent an evolutionary negative-feedback loop for the proinflammatory cag T4SS.
Model depicting the role of S100A12 protein in the repression of the H. pylori cag T4SS within the gastric niche. In the absence of S100A12, H. pylori elaborates a functional cag T4SS and induces changes in host epithelial cells, including IL-8 secretion and elongated cell morphology (hummingbird phenotype). The secretion of IL-8 recruits innate immune cells, including neutrophils, which deposit S100A12 as a response to H. pylori infection. S100A12 sequesters the nutrient zinc at the local site of infection, making it unavailable for the bacterial cell, repressing the biogenesis of the cag T4SS pili and the cellular consequences, such as IL-8 secretion and the hummingbird phenotype. This negative-feedback loop aids the bacterium in tempering its virulence to evade immune detection, thereby persisting in the host tissue to establish a chronic infection.
ACKNOWLEDGMENTS
This work was funded primarily by a Career Development Award IK2BX001701 (to J.A.G.) from the Office of Medical Research, Department of Veterans Affairs. Additional support was provided by NIH grant R01 AI101171 (to E.P.S. and W.J.C.), by Merit Award INFB-024-13F from the Office of Medical Research, Department of Veterans Affairs (to E.P.S), and by NIH NRSA F32 AI108192 and Childhood Infections Research Program T32-AI095202 (to B.L.M). Core services, including use of the Cell Imaging Shared Resource, were performed through both Vanderbilt University Medical Center's Digestive Disease Research Center, supported by NIH grant P30DK058404 Core Scholarship, and the Vanderbilt Institute for Clinical and Translational Research program, supported by the National Center for Research Resources, grant UL1 RR024975-01, and the National Center for Advancing Translational Sciences, grant 2 UL1 TR000445-06.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
FOOTNOTES
- Received 24 April 2015.
- Accepted 30 April 2015.
- Accepted manuscript posted online 11 May 2015.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00544-15.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
