Previous Article | Next Article ![]()
Infection and Immunity, January 2006, p. 566-577, Vol. 74, No. 1
0019-9567/06/$08.00+0 doi:10.1128/IAI.74.1.566-577.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Discipline of Immunology & Microbiology, School of Biomedical Sciences, Faculty of Health, The University of Newcastle, Newcastle, Australia,1 Vaccines, Immunology/Infection, Viruses and Asthma Group, The Hunter Medical Research Institute, Newcastle, Australia,2 Environmental & Life Sciences, Faculty of Science & Mathematics, The University of Newcastle, Newcastle, Australia3
Received 15 April 2005/ Returned for modification 16 June 2005/ Accepted 20 October 2005
|
|
|---|
|
|
|---|
Random peptide phage display has been successfully employed to identify new receptors and receptor ligands (44, 52, 53, 82) and in epitope mapping and mimicking of protein antigens and antibodies (16, 42, 61) and drug discovery (40, 52, 78). In this study we employed random peptide phage display in an attempt to identify novel surface antigens that may be used as therapeutic targets and diagnostic markers for Listeria spp. and L. monocytogenes. Subtractive panning of the phage display library was performed on immobilized L. monocytogenes cells, and a single peptide sequence was identified. The peptide sequence corresponds to a sequence found in repeat 3 of the insulin-like growth factor II receptor (IGFIIR; also known as the cation-independent mannose 6-phosphate [M6P] receptor CIM6PR or CD222), which is one of the binding sites for M6P in this receptor. Phage display results were confirmed by demonstrating that Listeria spp. specifically bound a synthetic peptide (M6Pbs peptide), designed to match the sequence of the M6P binding site, as well as the native receptor. The specificity of the interaction was further investigated by employing competitive binding and invasion assays using the M6Pbs peptide and M6P. We demonstrate a role for this receptor in adherence and invasion of mammalian cells by L. monocytogenes. This is the first time that the IGFIIR has been characterized as a receptor for Listeria spp.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Bacterial strains employed in the study
|
Preparation of bacterial strains for coating and panning. Bacterial strains were grown overnight (o/n) (10 ml tryptic soy broth, 37°C, static conditions). An L. monocytogenes cocktail was obtained by combining 100-µl amounts of each serotype (1.1 ml total). An L. innocua cocktail (500 µl of each serotype) and an L. ivanovii cocktail (250 µl of each isolate) were similarly produced (1 ml total). Bacteria were washed four times (10 mM phosphate buffer [pH 8.1], 6,000 x g, 10 min), and the pellets were resuspended (1 ml of phosphate buffer with 0.1% sodium azide).
Preparation of bacterial cells for binding and invasion assays. Bacterial strains were grown o/n. Log-phase bacterial cultures were obtained by inoculating 0.5 ml of the culture into 9.5 ml of tryptic soy broth. Cultures were grown (37°C with shaking at 200 rpm) to an optical density at 600 nm of 0.5 to 0.7, washed four times (phosphate-buffered saline [PBS], 6,000 x g, 10 min), and resuspended (1 ml PBS).
Labeling of bacterial cells. O/n bacterial cultures were gently washed three times (10 mM phosphate buffer [pH 8.1], 1,400 x g, 10 min, 4°C) and resuspended. Bacteria (1 ml) were labeled by incubating with 5- (and 6-) carboxyfluorescein diacetate-succinimidyl ester (CSFE; Molecular Probes, Eugene, Oregon) (2 µl, 5 mM, 30 min, 37°C). Bacteria were washed, resuspended (1 ml PBS), and stored at 4°C for up to a week.
Binding of labeled peptides by bacterial cells. Peptides were synthesized by Auspep Pty. Ltd. (Parkville, Australia) and labeled with fluorescein isothiocyanate (FITC) during synthesis. Peptides were the identified peptide PLAQSGGSSYI (M6Pbs peptide), a control peptide with a randomized sequence of the M6Pbs peptide YPSGSILGSQA (M6Pbs control peptide), and a second control peptide corresponding to the IGFII binding site of the IGFIIR, FGQTRISVGKA (IGFIIbs peptide) (27). Lyophilized peptides were resuspended (5 mg/ml) with 50% dimethyl sulfoxide (DMSO) (M6Pbs peptide and M6Pbs control peptide) or PBS (IGFIIbs peptide). Bacterial cultures were washed (1 ml PBS) and resuspended with each of the peptides (40 µl, 2 h, 4°C). Bacteria were washed, resuspended (1 ml PBS-0.1% sodium azide), and stored at 4°C for up to a week.
Preparation of mammalian cells for fluorescence-activated cell sorting (FACS) and invasion assays.
MS9, MS9II, and McCoy cells were grown in serum-free supplemented Dulbecco's modified Eagle medium (DMEM) containing fetal calf serum (FCS; CSL, Parkville, Australia) (10%), HEPES (25 mM), L-glutamate (2 mM), sodium pyruvate (0.1g/liter), and penicillin-streptomycin (100 µg/ml). To maintain IGFIIR expression, MS9II cell medium contained methotrexate (SIGMA, Sydney, Australia) (3.2 µM). For invasion assays, cells were seeded into 24-well tissue culture plates (
105 cells/well) and grown until >90% confluent.
Phage display. Phage displays were produced using a Ph.D. 12mer random phage display peptide library (NEB, Beverly, Massachusetts). Panning was performed using Dynex Immunolon microtiter wells coated with L. monocytogenes cocktail (150 µl/well, diluted 1:100, in phosphate buffer, o/n, room temperature [RT]). After coating, wells were blocked (5 mg/ml bovine serum albumin [BSA], 300 µl/well, 2 h, RT), washed once (distilled water), and stored at 4°C until further use. Nonspecific phages were removed by absorbing the phage library sequentially on uncoated wells and wells coated with BSA (1 mg/ml, 30 min, RT, rocking). Selection of antigen-positive phages was performed by the addition of the absorbed phage library to immobilized L. monocytogenes cocktail (60 min, RT, rocking). Unbound phages were discarded, and wells were washed three times (Tris-buffered saline-0.5% Tween 20). Subtractive selection was carried out by the sequential application of L. innocua cocktail and L. ivanovii cocktail (100 µl, 1:1,000 in Tris-buffered saline) to wells (30 min, RT), with washing between steps. Low-affinity L. monocytogenes-specific phage peptides were removed by rapid exposure and removal of L. monocytogenes cocktail (100 µl). The wells were washed before incubation with L. monocytogenes cocktail (30 min, RT) to elute specific high-affinity bound phages. The phage eluate was titered, amplified, and panned twice more according to the manufacturer's instructions. DNA was prepared from single phage clones after two rounds of panning by phenol extraction (Tris-buffered phenol, 1:1) and ethanol precipitation before DNA sequencing (Newcastle Sequencing, University of Newcastle, Newcastle, Australia).
Binding assays using fluorescently labeled peptides. Microtiter wells were coated with bacterial preparations (1:100) and BSA (1 mg/ml, 150 µl/well, in phosphate buffer). M6Pbs peptide and M6Pbs control peptide were serially diluted (100 to 6.25 µg/ml) in PBS-0.05% Tween 20 (PBS/T) containing 25% DMSO,while the IGFIIbs peptide was serially diluted in PBS/T. Binding was performed by incubating peptides with immobilized bacteria (100 µl/well, o/n, 4°C). Wells were washed five times (PBS/T), and distilled water (100 µl) was added before the fluorescence was read (excitation 485 nm, emission 538 nm) using a fluorescence enzyme-linked immunosorbent assay plate reader (Fluoromark; Bio-Rad).
Affinity purification of IGFIIR from FCS by use of magnetic beads coated with bacteria. Magnetic beads (Dynal, Victoria, Australia) (0.5 ml) were prepared as described in the manufacturer's instructions, coated with L. monocytogenes cocktail (1 ml, 1:100, 0.1 M phosphate buffer [pH 8.0], o/n, 4°C, rotating wheel) and washed five times (50 ml PBS) to remove unbound material. For coating of beads with cell surface protein extract, an o/n culture of L. monocytogenes (250 ml) was washed twice (50 mM phosphate buffer [pH 7.6], 1,400 x g, 4°C), resuspended (3 ml, 50 mM phosphate buffer [pH 7.6], 1% sodium dodecyl sulfate [SDS]), and incubated in a water bath (2 h, 70°C). Cell debris was removed (3,000 x g, 10 min), the cell surface protein extract was dialyzed (50 mM phosphate buffer [pH 7.6], o/n, 4°C), and magnetic beads were coated with the extract (1:100). Affinity purification of IGFIIR was performed by incubating coated beads with FCS (250 ml, o/n, RT, rotating wheel). Unbound material was removed by washing three times (50 ml PBS, 4°C), and bound material was eluted (1 ml, 0.1 M glycine-HCl, pH 2.5). Eluates were concentrated using microcon concentrators (molecular weight, 30,000; Millipore), and concentrates were made up to 0.25 ml (10 mM Tris-0.8% NaCl, pH 8.0) and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting.
Preparation of mammalian cells for protein analysis. Mammalian cells were harvested (10 ml, PBS-0.05% Triton X-100) and sonicated (10,000 Hz, 2 min), and the protein content of cell lysates was determined using a bicinchoninic acid assay (Pierce, Rockford, IL) according to the manufacturer's instructions. The concentrations of cell lysates were adjusted (0.5 mg/ml) prior to analysis by SDS-PAGE and Western blotting.
SDS-PAGE and Western blotting. SDS-PAGE (47) was performed using 10%- or 4- to 20%-gradient polyacrylamide gels (Gradipore, French's Forest, Australia). Briefly, samples were heated in sample buffer and separated on gels, and proteins were visualized using Coomassie blue or a silver staining kit (Bio-Rad, Sydney, Australia). For Western blotting, gels were transferred onto nitrocellulose (Bio-Rad) in transfer buffer (12 mM Tris, 0.96 M glycine [pH 8.0], 20% methanol), and transfer was performed (o/n, 4°C, 60 V). The membrane was blocked (5% skim milk powder-PBS/T, 4 h, RT), and bovine IGFIIR was detected with monoclonal mouse anti-bovine IGFIIR antibody (Affinity BioReagents, Golden, CO) (1 µg/ml, 1% skim milk-PBS/T) as primary antibody (o/n, RT) and goat anti-mouse immunoglobulin G horseradish peroxidase conjugate (SIGMA) (1:5,000, 1% skim milk-PBS/T) as secondary antibody (1.5 h, RT). Detection of human or murine IGFIIR was performed with goat polyclonal anti-human or anti-mouse IGFIIR, respectively (Affinity BioReagents) (1 µg/ml, 1% skim milk-PBS/T) and rabbit anti-goat immunoglobulin G horseradish peroxidase conjugate (SIGMA, 1:100,000, 1% skim milk-PBS/T). Blots were washed three times between steps (5 min, PBS/T, rocking). Bands were detected using Supersignal chemiluminescent substrate (Pierce).
FACS scanning. Fluorescent cells were detected with a FACScan flow cytometer and analyzed using CellQuest software (BD Biosciences). Bacterial binding was calculated by determining the relative fluorescent intensity (relative fluorescent intensity = geographical mean of the sample/geographical mean of the control).
Binding of FITC-labeled peptides by bacteria. FITC peptide-labeled bacteria (100 µl) were taken up in paraformaldehyde (900 µl, 1%), and control procedures were performed by incubating bacterial cells in 25% DMSO-PBS (as the control for the FITC-M6Pbs peptide and FITC-M6Pbs control peptide) or PBS (FITC-IGFIIbs peptide).
Effect of M6Pbs peptide on bacterial adherence. CSFE-labeled bacteria (50 µl, 109 to 1010 cells/ml) were incubated with M6Pbs peptide (40 µl, 5 mg/ml, 30 min, 37°C) or the control (25% DMSO in PBS). Unbound peptide was removed by washing three times (5 min, PBS, 6,000 x g), and bacteria were incubated with mammalian cells (400 µl, 106 to 107cells/ml, DMEM, 20 min, 37°C). Samples were washed three times (5 min, PBS, 700 x g) and resuspended (200 µl, 2% paraformaldehyde).
Effect of M6P on bacterial adherence. CSFE-labeled bacteria (50 µl, 109 to 1010cells/ml) and mammalian cells (50 µl of 106 to 107cells/ml) were added to 400 µl of PBS, PBS containing 50 mM glucose-6 phosphate (G6P) (controls), or PBS containing 0, 10, 20, or 50 mM M6P (20 min, 37°C). Samples were washed three times with PBS, PBS containing 10 mM G6P (control samples), or PBS containing 10 mM M6P (M6P samples) (5 min, 700 x g) and resuspended (200 µl, 2% paraformaldehyde).
Invasion assay-gentamicin protection assay.
Mammalian cell monolayers were washed three times (DMEM, 37°C) before the addition of log-phase bacterial cells at a multiplicity of infection of
10:1 (1 h, 37°C) followed by the addition of gentamicin (20 µg/ml final concentration, 2 h, 37°C). Mammalian cells were washed three times (serum-free DMEM, 37°C), and cells were lysed by the addition of ice-cold 0.2% Triton X-100 (15 min, 4°C). Intracellular bacterial cells were recovered by scraping the wells vigorously and plating recovered bacteria (tryptic soy agar plates, o/n, 37°C). Bacterial colonies were counted, and the percentage of relative invasion was determined using the following equation: number of recovered bacteria x 100%/number of bacteria in inoculum.
Statistical analysis. Analysis of statistical significance for studies showing dose-dependent variations (Fig. 1; see also Fig. 6 and 8) was performed using the test Nptrend (23). The two-sample Wilcoxon rank-sum test was performed to determine statistical significance of differences observed between two groups (Fig. 1 and 2; see also Fig. 5 and 7).
![]() View larger version (23K): [in a new window] |
FIG. 1. Binding of FITC-labeled M6Pbs peptide to immobilized bacteria. Bacterial preparations were immobilized on microtiter wells, and binding of synthetic, FITC-labeled peptides was determined by measurement of fluorescence spectrophotometry. (A) Binding of the M6Pbs peptide by various Listeria spp. was dose dependent (P < 0.01) and was significantly increased compared to control results at 100 µg/ml (***, P < 0.001), 50 µg/ml (**, P < 0.01), and 25 µg/ml (*, P < 0.05). (B and C) Only background binding of the M6Pbs-control peptide (B) and control IGFIIRbs peptide (corresponding to a sequence within the IGFII binding site of the IGFIIR) (C) by immobilized bacterial preparations and BSA was observed in comparisons to binding of the M6Pbs peptide at 100 µg/ml (***, P < 0.001), 50 µg/ml (**, P < 0.01), and 25 µg/ml (*, P < 0.05). Data presented are means ± standard deviations (n = 3 in duplicate).
|
![]() View larger version (28K): [in a new window] |
FIG. 6. The effect of M6P on binding of Listeria spp. to McCoy cells. CSFE-labeled Listeria spp. were incubated with McCoy cells in the presence of M6P (0 to 50 mM), and binding was measured by FACS. Bacterial binding was determined by calculating the relative fluorescent intensity. Significant dose-dependent inhibition of binding was observed for L. monocytogenes cocktail (P < 0.05) and serotype 1/2a (P < 0.01). There was a smaller but significant dose-dependent increase in binding for L. innocua cocktail and serotype 6b (P < 0.05). Data presented are means ± standard deviations (n = 3 in duplicate).
|
![]() View larger version (27K): [in a new window] |
FIG. 8. The effect of M6P on invasion of Listeria spp. into McCoy cells. Invasion assays of Listeria spp. into McCoy cells were performed in the presence of M6P (0 to 50 mM). Significant inhibition of invasion of McCoy cells by L. monocytogenes cocktail (P < 0.01) and serotype 1/2a (P < 0.01) and increases in invasion by L. innocua cocktail (P < 0.01) and serotype 6b (P < 0.01) were observed. Data presented are means ± standard deviations (n = 3 in duplicate).
|
![]() View larger version (28K): [in a new window] |
FIG. 2. Binding of FITC-labeled M6Pbs peptide by various bacterial strains. Binding of FITC-labeled peptides to Listeria spp. and controls was measured by FACS, and relative fluorescent intensity was calculated. (A) Levels of binding of the M6Pbs peptide by various Listeria spp. were significantly (**, P < 0.01) elevated compared to control (E. coli and B. cereus) results. (B) Binding of the M6Pbs-control peptide and IGFIIbs peptide (controls) by cocktails of L. monocytogenes (**, P < 0.01), L. innocua (*, P < 0.05), and L. ivanovii (**, P < 0.01) was significantly lower than binding of the M6Pbs peptide. Data presented are means ± standard deviations (n = 2 in duplicate).
|
![]() View larger version (28K): [in a new window] |
FIG. 5. Binding of Listeria spp. to MS9 and MS9II cells in the presence of M6Pbs peptide by FACS. Bacteria were labeled with CFSE and incubated with MS9 and MS9II cells. Bacterial binding was determined using FACS by calculating the relative fluorescent intensity. No differences were observed between binding to MS9II cells and binding to MS9 cells for any of the bacterial strains used. Binding was further investigated using inhibition of binding by the M6Pbs peptide. M6Pbs peptide was bound by CSFE-labeled bacteria, which were then incubated with MS9 and MS9II cells. Significant inhibition of binding to MS9II cells in the presence, compared to the absence, of the M6Pbs peptide was observed for L. monocytogenes cocktail (*, P < 0.01) and serotype 1/2a (**, P < 0.05) but not L. innocua cocktail or B. cereus. Data presented are means ± standard deviations (n = 2 in duplicate).
|
![]() View larger version (23K): [in a new window] |
FIG. 7. Invasion of Listeria spp. into MS9 and MS9II cells. Log-phase cultures of L. monocytogenes at a multiplicity of infection of 10:1 were incubated with MS9 and MS9II cells, and gentamicin was added to kill extracellular bacteria. Intracellular listerial cells were released by lysis of mammalian cells and recovered by plating onto media. Invasion was further investigated using listerial strains deficient in known virulence factors InlA and/or InlB. Percent relative invasion was calculated using the number of recovered listerial cells x 100%/the number of cells in the inoculum. Invasion of L. monocytogenes cocktail was significantly (**, P < 0.01) increased in MS9II cells compared to MS9 cell results. Invasion of L. monocytogenes strains lacking InlB was significantly reduced compared to the wt EGD strain results (*, P < 0.05). Noninvasive L. innocua was used as a control and was not recovered from lysed mammalian cells (not shown). Data presented are means ± standard deviations (n = 3 in duplicate).
|
|
|
|---|
|
View this table: [in a new window] |
TABLE 2. Panning of phage-displayed peptidesa
|
Binding of labeled peptides by immobilized bacterial cells. The identified peptide was synthesized, labeled with FITC (FITC-M6Pbs peptide), and used in binding assays to show that listerial cells specifically bound the peptide. The identified peptide was bound by all Listeria spp. but not by other bacterial species (such as the gram-positive species Bacillus cereus and the gram-negative strain Escherichia coli ER2738) or by BSA (for 100 µg/ml, P < 0.001; for 50 µg/ml, P < 0.01; and for 25 µg/ml, P < 0.05) (Fig. 1A). To further demonstrate the specific association between the M6Pbs peptide, Listeria cells, and the IGFIIR, two FITC-labeled control peptides were used in this study. One of the control peptides (M6Pbs-control peptide) had a randomized sequence of the M6Pbs peptide (YPSGSILGSQA), and a second peptide (FITC-IGFIIbs peptide) corresponded to a sequence within the IGFII binding site of the IGFIIR (FGQTRISVGKA) (10). By contrast to the M6Pbs peptide results, neither of the control peptides were bound by Listeria spp. or other bacteria (for 100 µg/ml, P < 0.001; for 50 µg/ml, P < 0.01; and for 25 µg/ml, P < 0.05) (Fig. 1B and C).
Analysis of binding of labeled peptides by bacterial cells. Specific binding of the FITC-M6Pbs peptide by Listeria spp. was confirmed by FACS. All Listeria spp. bound this peptide at significantly higher levels than controls (P < 0.01). L. monocytogenes deficient in virulence factor InlA and/or InlB bound the M6Pbs peptide at levels similar to those seen with wild-type (wt) strains, demonstrating that loss of these virulence factors had no impact on the binding of this peptide. M6Pbs peptide did not bind to controls (B. cereus and E. coli ER2783) (Fig. 2A). To further determine that the binding of FITC-M6Pbs peptide by Listeria spp. was specific, binding of the FITC-labeled control peptides (M6Pbs-control peptide and IGFIIbs peptide) by cocktails of L. monocytogenes, L. innocua, and L. ivanovii serotypes was measured and compared to FITC-M6Pbs peptide binding by use of the same preparations. Binding of FITC-labeled control peptides by cocktails of L. monocytogenes, L. innocua, and L. ivanovii serotypes was significantly reduced compared to binding of FITC-M6Pbs peptide (P < 0.001, P < 0.05, and P < 0.01, respectively) (Fig. 2B).
Listeria spp. can bind the native receptor. To confirm that Listeria spp. can bind the native IGFIIR, whole bacterial cells and surface protein extracts were coated onto magnetic beads and used to isolate the native receptor from FCS by affinity purification. Purified IGFIIR protein is not commercially available; however, a truncated form missing the cytoplasmic domain of the receptor is circulating in fetal serum and can be affinity purified from FCS by use of M6P affinity matrices (11, 49, 50). Although the M6Pbs peptide sequence used is homologous to that of the human IGFIIR, demonstration of binding of the bovine IGFIIR by Listeria spp. is relevant because this protein is highly conserved in mammalian species (38, 84). The IGFIIR was eluted from listerial preparations and detected by SDS-PAGE and Western blotting using a monoclonal anti-bovine IGFIIR antibody. A strong band of approximately 200 kDa, corresponding to the molecular mass reported for the circulating form of the receptor (49), was detected in affinity-purified material from L. monocytogenes and L. innocua preparations but not in materials purified from L. ivanovii or controls (Fig. 3). There were also some additional faint (119 kDa and below) bands detected in the Western blot (Fig. 3A) that may have resulted from the detection of breakdown products of the IGFIIR.
![]() View larger version (41K): [in a new window] |
FIG. 3. Affinity purification of bovine IGFIIR by use of magnetic beads coated with various bacterial preparations. Magnetic beads were coated with bacterial preparations and incubated with FCS. Affinity-purified material was eluted, and binding of IGFRII was detected by SDS-PAGE and Western blotting. The Western blot was developed using a specific monoclonal anti-bovine IGFIIR antibody. Bovine IGFIIR ( 200 kDa) was bound by L. monocytogenes cocktail cells (panel A, lane 1) and cell surface proteins (panel A, lane 2) and cells of L. monocytogenes serotype 1/2a (panel B, lane 5) and L. innocua cocktail (panel B, lane 3) but not by cells of L. ivanovii cocktail (panel B, lane 4) or controls (E. coli ER2738 [panel B, lane 1] or B. cereus [panel B, lane 2]). Standard molecular masses are represented in kilodaltons. The gels depicted are representative of duplicate experiments.
|
Detection of the IGFIIR in mammalian cells. The presence of the IGFIIR in MS9II and McCoy cells but not in MS9 cells was confirmed by SDS-PAGE and Western blotting using antibodies against human or murine IGFIIR. A distinct band of approximately 300 kDa was detected in Western blots of cell lysates prepared from MS9II and McCoy cells but not in MS9 cells (Fig. 4). The detected band (300 kDa) corresponds to the reported molecular mass for this receptor in mammalian cells (84). Additional smaller-sized bands (75 kDa and below), which may have been due to breakdown products of the IGFIIR specifically interacting with the antibodies, were also detected.
![]() View larger version (26K): [in a new window] |
FIG. 4. Detection of IGFIIR in MS9, MS9II, and McCoy cells. Cell lysates of MS9, MS9II, and McCoy cells were prepared by sonication and separated on a 4- to 20%-gradient SDS-PAGE gel prior to Western blotting. The Western blots were developed using a monoclonal anti-human IGFIIR antibody for MS9 and MS9II cells and a polyclonal anti-mouse IGFIIR antibody for McCoy cells. The IGFIIR ( 300 kDa) was detected in cell lysates from MS9II (panel A, lane 2) and McCoy (panel B, lane 1) cells but not MS9 (panel A, lane 1) cells. Standard molecular masses are represented in kilodaltons. The gels depicted are representative of duplicate experiments.
|
Binding of bacterial preparations to mammalian cells in the presence of soluble M6P. The peptide sequence identified by phage display is located in repeat 3 of the IGFIIR, which has been identified as a carbohydrate binding site with specificity for M6P (11, 37). Binding of L. monocytogenes to McCoy cells was inhibited using soluble M6P as a competitive inhibitor. CSFE-labeled listerial preparations were incubated with mammalian cells, and adherence was measured by FACS (Fig. 6). Binding of L. monocytogenes (cocktail and serotype 1/2a) to McCoy fibroblast cells was significantly inhibited by M6P in a dose-dependent manner (P < 0.05 and P < 0.01, respectively). Surprisingly, adherence of L. innocua (cocktail and serotype 6b) to McCoy cells increased significantly in the presence of M6P (P < 0.05). To determine that inhibition was a result of M6P specifically inhibiting binding of Listeria spp. to the M6P binding sites of the IGFIIR, G6P was used as an inhibitor in control experiments. G6P is a closely related analogue of M6P, but it has no affinity for the M6P binding sites of the IGFIIR. Binding of Listeria spp. to McCoy cells was not inhibited in the presence of G6P (results not shown).
The IGFIIR is involved in invasion of mammalian cells by Listeria spp. A role for the IGFIIR in invasion of mammalian cells by Listeria spp. was demonstrated. Invasion of L. monocytogenes into MS9II cells was significantly increased compared with invasion of MS9 cells. Furthermore, invasion of McCoy cells by L. monocytogenes was significantly inhibited by soluble M6P.
Invasion of Listeria spp. into MS9 and MS9II cells. Invasion of wt L. monocytogenes (cocktail, serotype 1/2a, or strain EGD) into MS9/MS9II cells represented approximately 1% of the inoculum, but invasion was significantly (0.2 to 0.3%) reduced in strains lacking InlB (P < 0.05) compared to wt strain results (cocktail, serotype 1/2a, and strain EGD) (Fig. 7). Invasion of L. monocytogenes cocktail into MS9II cells was significantly greater than MS9 cell invasion, as determined using raw (P < 0.01) (Fig. 7) or normalized (P < 0.01) (not shown) data. Normalizing data is an accepted method of analyzing data (76). Results of invasion of L. monocytogenes serotype 1/2a and strain EGD and strains lacking inlA, inlB, and inlAB into MS9II cells compared to MS9 cell invasion were not significantly different, as determined using raw data (Fig. 7). However, significantly greater invasion into MS9II cells compared to MS9 cell invasion for L. monocytogenes serotype 1/2a (P < 0.05) and strains lacking inlA (P < 0.01), inlB (P < 0.05), and inlAB (P < 0.05) was determined by normalizing raw data (not shown). Normalized relative invasion values were obtained by setting relative values for invasion into MS9 cells at 1%, and relative invasion into MS9II cells was calculated relative to 1%. The noninvasive species L. innocua was used as a control and did not invade MS9 or MS9II cells (not shown).
Invasion of Listeria spp. into McCoy cells in the presence of soluble M6P. Invasion studies were performed by incubating log-phase listerial cultures with McCoy cells in the presence of various concentrations of M6P as a competitive inhibitor. Invasion of McCoy cells by L. monocytogenes (cocktail and serotype 1/2a) was inhibited by M6P in a dose-dependent manner (P < 0.01). Inhibition of entry occurred at a concentration of 10 mM M6P (relative invasion reduced from 9.5% to 5%) but was most prominent at 50 mM (relative invasion reduced to 1.7%) (Fig. 8). Invasion of McCoy cells by L. innocua cocktail and serotype 6b significantly increased in the presence of M6P in a dose-dependent manner (Fig. 8).
|
|
|---|
Thus, all Listeria spp. bind the M6Pbs peptide; however, some differences were observed in the binding of the IGFIIR protein from different mammalian species by some Listeria spp. L. ivanovii bound the M6Pbs peptide (corresponding to the human IGFIIR sequence) (Fig. 2) but not the native bovine IGFIIR (Fig. 1 and 3).
The functions of the IGFIIR in mammalian cells are to bind and internalize insulin-like growth factor II (IGFII) (26, 36, 43, 75), to internalize and to sort lysosomal enzymes and other M6P-tagged proteins (11, 12, 24, 45, 46), and to regulate transforming growth factor ß activity (33, 51). The majority of the receptor is found inside the cell, with only 5 to 10% located on the cell surface (45). The surface-associated IGFIIR is an important factor in the regulation of cell growth and in tumor suppression, because it removes IGFII from circulation and regulates transforming growth factor ß activity (27).
The IGFIIR is made up of 15 cysteine-based repeats, a transmembrane domain, and a short, hydrophobic cytoplasmic domain (59). The peptide sequence identified in our studies is located in repeat 3 of the extracellular domain that is one of the three M6P binding sites of the IGFIIR (25, 68). Recently, the amino acid sequence in repeat 3 was determined (25, 38, 65); the identified (M6Pbs) sequence is located between two conserved cysteine residues (C334 and C375) and upstream of a conserved residue (Y368) which is essential for carbohydrate recognition (38, 65, 68, 84).
Adherence of Listeria to mammalian cells is known to occur through a wide range of different listerial virulence factors and cell surface components such as fibronectin binding proteins (31, 41), carbohydrates (15, 20, 22, 67), protein 60 (9, 72), the autolysin ami (58), lipoproteins (69), and teichoic and lipoteichoic acids (1, 34, 66, 67). These listerial cell surface components interact with a diverse range of host cell receptors, many of which have not yet been identified. Our results demonstrate that the IGFIIR is another such host cell receptor for Listeria spp. that is involved in both binding and invasion.
Binding of L. monocytogenes, but not L. innocua, to the receptor on MS9II cells could be competitively inhibited by the M6Pbs peptide but not the control peptides (M6Pbs-control peptide and IGFIIbs peptide). Furthermore, binding of L. monocytogenes to McCoy cells was significantly inhibited by M6P in a dose-dependent manner. Competitive inhibition studies using soluble M6P as an inhibitor of binding of other ligands are established techniques for the investigation of IGFIIR function (37, 62, 73). The M6P effects were specific, because G6P, a closely related analogue that does not bind to the IGFIIR (38), had no effect. These results demonstrate that the IGFIIR is involved in adherence of L. monocytogenes. Surprisingly, adherence of L. innocua strains to McCoy cells increased significantly in the presence of M6P. When considered together, these results suggest that the usage of the IGFIIR differs between different Listeria spp.
A fundamental role for the IGFIIR in invasion of L. monocytogenes is also evident by the demonstration that there was a trend to increased invasion into MS9II cells compared to MS9 cell invasion for all L. monocytogenes serotypes and mutants tested. Furthermore, invasion of McCoy cells by L. monocytogenes strains was significantly inhibited by M6P in a dose-dependent manner. Internalization of L. monocytogenes by mammalian cells is known to be mediated by the major virulence factors InlA and InlB through their interaction with specific host cells receptors E-cadherin and the hepatocyte growth factor receptor, respectively (17, 18, 56, 57). However, invasion into mouse fibroblast cells is not mediated by InlA, because the murine E-cadherin contains an amino acid substitution that abolishes interaction with InlA (48). L. monocytogenes strains deficient in InlA invaded MS9 and MS9II at approximately the same level as wt strains, confirming that invasion into murine cells is not mediated by InlA (48). Invasion of MS9II and MS9 cells by L. monocytogenes bacteria deficient in InlB and both InlA and InlB was significantly reduced.
Both of these observations indicate that internalization into murine fibroblast cells is primarily mediated by InlB (8, 19). The observation that invasion into MS9 cells by mutants lacking inlB and inlAB was not completely abolished also indicates that yet other receptors may be involved. The residual invasiveness of L. monocytogenes mutants lacking the major virulence factors InlA and InlB was also evident in other studies (4, 7, 29, 35) and suggests that there are a number of other host cell receptors involved in L. monocytogenes invasion (2, 34).
A significant increase in both binding and invasion of L. innocua of murine fibroblast cells was observed in the presence of soluble M6P in a dose-dependent manner. A possible explanation for this observation is that, in contrast to pathogenic L. monocytogenes, the nonpathogenic species L. innocua cannot utilize hexose phosphate sugars (14, 70, 77). We speculate, therefore, that this organism may have nonspecifically incorporated the soluble M6P on its surface, which led to uptake of L. innocua through the IGFIIR. The M6P binding sites of the IGFIIR have been targeted previously to introduce M6P-coated foreign substances into mammalian cells by use of M6P-coated vesicles (63, 71) or albumin modified with M6P (5, 6). This provides evidence that the IGFIIR can function effectively to transport large particles from the extracellular space into the host cytoplasm.
Interactions of L. monocytogenes with mammalian cells, which involve binding of lectins or lectin-like molecules, including carbohydrate binding components and D-galactose- and glucosamine- or fucosamine-containing proteins, have previously been demonstrated (20-22, 80). Such components have been shown to play an important role in listerial pathogenicity and innate immune responses (32). However, there is as yet no evidence that Listeria spp. contain M6P, as a component of the cell wall, teichoic acids, lipoteichoic acids, or other virulence factors, that is available to interact with IGFIIR during pathogenesis. The carbohydrate binding site of the IGFIIR in domain 3 has been shown to recognize related structures such as phosphodiesters or mannose 6-sulfate (54, 55, 81), and the interaction of Listeria spp. with the IGFIIR may be based on such a related structure that is present on the cell surface of Listeria bacteria.
In conclusion, we have shown that binding and invasion of mammalian cells by L. monocytogenes may involve the IGFIIR and that this may be a novel mechanism of pathogenesis.
We thank TECRA International for kindly supplying the strains used in this study. We are most grateful to Keith Ireton (University of Toronto, Canada) for assistance in experimental design and manuscript preparation. We also thank R. MacDonald (Nebraska Medical Centre) for advice on IGFIIR purification techniques and preparation of the manuscript, W. Sly for kind permission to use MS9 and MS9II cell lines, and J. Trapani for the provision of these cell lines.
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2010 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»