HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Su, B. Articles by Sherman, P. M. Search for Related Content PubMed PubMed Citation Articles by Su, B. Articles by Sherman, P. M. Agricola Articles by Su, B. Articles by Sherman, P. M.

Cytoskeletal rearrangements in gastric epithelial cells in response to Helicobacter pylori infection

Research Institute, Hospital for Sick Children, Departments of Paediatrics and Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada M5G 1X8

Correspondence Philip M. Sherman sherman{at}sickkids.ca

Received February 20, 2003
Accepted June 13, 2003

Helicobacter pylori causes host epithelial cell cytoskeletal rearrangements mediated by the translocation and tyrosine phosphorylation of an outer-membrane protein, CagA, and by the vacuolating cytotoxin, VacA. However, the mechanisms by which H. pylori mediates cytoskeletal rearrangements in infected host cells need to be more clearly defined. The aim of this study was to determine the effects of H. pylori isolates from children on the architecture of host gastric epithelial cells. Gastric epithelial (AGS) cells were infected with type I (cagE⁺, cagA⁺, VacA⁺) H. pylori, a type II H. pylori strain (cagE^-, cagA^-, VacA^-) or a cagE isogenic mutant. Double-labelled immune fluorescence was used to detect adherent H. pylori and the distribution of F-actin, -actinin and Arp3. Both type I and type II H. pylori strains induced stress fibres in gastric epithelial cells that were not observed in uninfected cells. Type I H. pylori also induced cell elongation (hummingbird phenotype) after 4 h of infection, whereas the type II H. pylori strain did not. Less elongation occurred when AGS cells were exposed to a cagE isogenic mutant, compared with the parental strain. Confocal microscopy showed Arp3 accumulation in AGS cells infected with wild-type H. pylori, but not in response to infection with the cagE mutant. These findings indicate that type I H. pylori induce a stress fibre-like phenotype in infected gastric epithelia by a mechanism that is different from the induction of host-cell elongation. In addition to CagA and VacA, cagE also impacts on the morphology of infected gastric epithelial cells.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Helicobacter pylori is a Gram-negative bacterium that infects the human stomach and causes gastritis, peptic ulceration and gastric cancers (Blaser & Berg, 2001). Type I H. pylori strains contain a cag pathogenicity island (PAI) that encodes proteins that constitute a type-IV secretion system and the putative virulence factors cagA and cagE (Censini et al., 2001). Studies using isogenic mutants have demonstrated that certain genes encoded on the cag PAI, including cagE but not cagA, are responsible for nuclear factor-B (NF-B) activation, resulting in the transcription of pro-inflammatory genes such as interleukin-8, interleukin-1ß, interferon- and tumour necrosis factor- (Maeda et al., 2001). H. pylori vacuolating cytotoxin (VacA) is also involved in inducing cytoskeletal rearrangements in infected epithelia (Pai et al., 2000; Ashorn et al., 2000).

H. pylori delivers CagA into the cytoplasm of host cells, through the type-IV secretion system, where it is phosphorylated on tyrosine residues (Backert et al., 2001; Odenbreit et al., 2000). CagA induces a growth factor-like phenotype, referred to as ‘hummingbird’ (Segal et al., 1999), in gastric epithelial cells. This morphological change is similar to that induced by hepatocyte growth factor (HGF), which occurs in an Src kinase-dependent fashion (Selbach et al., 2002). Once CagA is translocated into the host-cell cytosol and tyrosine-phosphorylated, the SHP-2–Rho pathway is activated to cause changes in host-cell morphology (Higashi et al., 2002; Lacalle et al., 2002). Backert et al. (2001) showed that, although the translocation and phosphorylation of CagA is important for the induction of the hummingbird phenotype in gastric (AGS) cells, it is not sufficient. This finding suggests, therefore, that there are additional bacterial factors involved in eukaryotic cellular responses to H. pylori infection.

In this study, we report that infection with paediatric H. pylori clinical isolates induces the accumulation of Arp3 in epithelial cells in a cagE-dependent manner. In addition to CagA and VacA, cagE is necessary for the induction of the hummingbird cell-elongation phenotype in response to H. pylori infection.

	METHODS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Reagents and antibodies.

Polyclonal H. pylori rabbit antiserum was purchased from DAKO. Monoclonal anti--actinin (IgM) antibody, fluorescent (FITC)-conjugated phalloidin and FITC-conjugated anti-mouse IgM were purchased from Sigma. Polyclonal anti-human Arp3 and goat anti-rabbit antiserum conjugated to rhodamine (TRTIC) were purchased from Santa Cruz Biotechnology. Secondary antibodies against rabbit IgG or against goat IgG conjugated with Cy5 or Cy3, respectively, were purchased from Jackson Laboratory.

Bacterial growth and infection conditions.

H. pylori strains employed in this study included the type I strain LC11 (cagA⁺, cagE⁺, VacA⁺) and the type II strain LC20 (cagA^-, cagE^-, VacA^-), both originally isolated from children in Toronto, respectively with duodenal ulceration and gastritis alone (Dytoc et al., 1993). H. pylori strain 8823 (cagA⁺, cagE⁺, VacA⁺) and its isogenic cagE mutant were a kind gift from Richard Peek, Jr (Vanderbilt University, Nashville, TN, USA).

Bacteria were cultured for 72 h on Columbia agar plates containing 5 % sheep blood (PML Microbiologicals) under microaerophilic conditions (5 % O₂, 85 % N₂ and 10 % CO₂) and then inoculated into Brucella broth supplemented with 10 % fetal bovine serum (FBS) at 37 °C under microaerophilic conditions with shaking overnight. For the cagE mutant, the culture medium contained 20 µg kanamycin ml^-1 (Gibco). For harvesting, bacteria were washed once with sterile PBS (pH 7.4) and then resuspended in antibiotic-free F-12 medium (Gibco) with 0.1 % FBS.

Tissue culture.

Human gastric carcinoma-derived AGS epithelial cells (CRL-1739; ATCC, Manassas, VA, USA) were grown in Ham's F-12 medium (Gibco) with 10 % FBS at 37 °C in 5 % CO₂. T84 cells (ATCC CCL-248), a polarized intestinal epithelial cell line derived from a colon cancer, were grown in a 1 : 1 mixture of Dulbecco's minimum essential medium (Gibco) and Ham's F12 medium. HEp-2 cells (epithelial cells originally derived from a carcinoma of the larynx; ATCC CCL-23) were cultured in MEM medium. Each of the cell lines was cultured in medium containing 2 mM L-glutamine, 10 % FBS, penicillin (100 U ml^-1) and streptomycin (100 µg ml^-1) (all from Gibco).

Immune fluorescence.

Tissue-culture cells were seeded onto coverslips and placed into 24-well plates (Nunc) in medium containing 0.1 % FBS without antibiotics for 20 h at 37 °C prior to bacterial infection. Cells then were washed once with sterile PBS and H. pylori was added at an m.o.i. of 100 : 1, for varying times at 37 °C. At the end of the infection period, tissue-culture cells were washed six times with PBS to remove non-adherent bacteria. Cells were then fixed in 2 % paraformaldehyde for 15 min and permeabilized with 0.2 % Triton X-100 for 10 min.

F-actin was stained by FITC-labelled phalloidin at 20 ng ml^-1 and -actinin was detected with anti--actinin mAb IgM (1 : 100) for 30 min followed by incubation with FITC-labelled anti-mouse IgM (1 : 100) for 30 min. Arp2/3 was detected with polyclonal anti-human Arp3 followed by anti-goat antibody conjugated with Cy3. HEp-2 cells were stained for the signal transducer and activator of transcription (Stat)-6, which was employed as a marker for cytoplasmic protein in epithelial cells (Schindler, 2002), using a specific polyclonal antibody (Santa Cruz) at a dilution of 1 : 100.

Adherent bacteria were visualized by staining bacteria with polyclonal anti-H. pylori antibody (1 : 100) followed by goat anti-rabbit antibody conjugated with TRITC (1 : 100). Alternatively, H. pylori was stained green by using anti-rabbit antibody conjugated with Cy5 (Santa Cruz Biotechnology).

Images were detected either under fluorescent microscopy (Leitz Dialux 22; Leica) or under laser scanning confocal microscopy performed using a Zeiss LSM 510. Data presented represent findings arising from three separate experiments.

	RESULTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Induction of a stress fibre phenotype in AGS cells by both type I and type II H. pylori strains

Stress fibres were observed when AGS cells were infected with a type I H. pylori, strain LC11 (30 min, m.o.i. 100 : 1) (Fig. 1). By double immunostaining, H. pylori-infected AGS cells showed stress fibre formation, compared with uninfected control cells. This change in morphology was observed as early as 30 min following infection and lasted for up to 4 h post-infection. The altered morphology was observed for more than 80 % of AGS cells following incubation with H. pylori. These changes occurred before the time when bacterial infection causes reduced viability of the gastric epithelia by inducing programmed cell death (Jones, 1999).

View larger version (90K): [in this window] [in a new window]   Fig. 1. Type I H. pylori induces stress fibres in AGS cells. Gastric epithelial cells were incubated with H. pylori strain LC11 for 30 min and then stained for F-actin with FITC-labelled phalloidin (green) and bacteria were stained with anti-H. pylori immune serum followed by goat anti-rabbit serum conjugated with rhodamine (red). Images were recorded by fluorescent microscopy (Leitz Dialux 22). (a) AGS cells alone. (b) Attached bacteria and stress fibres. Results are representative of three separate experiments. Approximate original magnifications, x1000.

Confocal microscopy of horizontal sections cut sequentially from the apical to basal aspect of AGS cells (Z-sections) showed that H. pylori adhered to and entered host cells and induced the accumulation of actin filaments in host epithelia (Fig. 2).

View larger version (93K): [in this window] [in a new window]   Fig. 2. Type II H. pylori induces stress fibres in AGS cells. Confocal microscopy of a horizontal section from top to bottom of the cell (Z-sections) showing double staining of actin filaments (green) and H. pylori (red). Sections are presented in sequence from the apical side (a) to the basolateral aspect (l) of an AGS cell. Images were recorded by using a Zeiss LSM 510 confocal microscope. Approximate original magnifications, x640.

The stress fibre phenotype was observed in more than 90 % of AGS cells infected with either a type I H. pylori, strain LC11 (Fig. 1), or a type II H. pylori, strain LC20 (Fig. 2). These data indicate that H. pylori-induced epithelial cell stress fibre formation occurred independently of the VacA, cagA and cagE status of the infecting organism.

Type I H. pylori-induced elongation of AGS cells is dependent on cagE

Analysis of cell morphology was also performed by the detection of -actinin using immunofluorescence and corresponding phase-contrast images. Epithelial cells infected with a type I H. pylori strain displayed extended, slender and elongated cell processes (Fig. 3c, d), also referred to as the hummingbird phenotype (Segal et al., 1999), which were not observed in uninfected cells (Fig. 3a, b). Epithelial cell elongation occurred after infection with H. pylori strain LC11 (m.o.i. 100 : 1) for 4 h, but not after 30 min. In addition, cell elongation was not observed when AGS cells were infected under the same experimental conditions with the type II H. pylori strain LC20 (cagA^-, cagE^-, VacA^-) (data not shown).

View larger version (133K): [in this window] [in a new window]   Fig. 3. Type I H. pylori-induced elongation in infected AGS cells is dependent on cagE. (a)–(d) AGS cells were incubated with culture medium only (a, b) or with H. pylori strain LC11 (cagA+, cagE+, VacA+) (c, d) for 4 h. Strain LC11 induced elongation of cells and the formation of processes extending from the elongated cells (c, d), whereas strain LC20 did not (data not shown). (e)–(h) AGS cells were incubated with wild-type strain 8823 (e, f) or with the cagE mutant (8823cagE : : kan) (g, h) for 4 h, demonstrating the induction of cell elongation with the wild-type and the absence of changes in morphology with the mutant. Panels (a), (c), (e) and (g) represent phase-contrast microscopy images. Panels (b), (d), (f) and (h) show cells stained for -actinin. Images were recorded by fluorescent microscopy (Leitz Dialux 22). Representative fields are shown from the results of four separate experiments. Approximate original magnifications, x400.

In order to delineate further the virulence factors involved in changing cell morphology, an isogenic cagE mutant was incubated with AGS cells for 4 h. As shown in Fig. 3, the type I wild-type H. pylori strain 8823 caused elongation of AGS cells (Fig. 3e, f), compared with uninfected controls. However, less elongation occurred when AGS cells were infected with the isogenic cagE mutant (Fig. 3g, h).

In order to determine whether H. pylori induced cytoskeletal rearrangements in other epithelial cell lines, immunofluorescent analysis was undertaken of T84 cells and HEp-2 cells infected with H. pylori strain LC11 (4 h, m.o.i. 100 : 1). As shown in Fig. 4, -actinin staining of H. pylori-infected T84 cells (Fig. 4a, b) and staining of the cytoplasmic protein Stat6 in infected HEp-2 cells (Fig. 4c, d) showed elongation of about 80 % of cells in all three cell lines. These data indicate that multiple epithelial cell lines elongate in response to infection with a type I H. pylori isolate.

View larger version (48K): [in this window] [in a new window]   Fig. 4. T84 cells and HEp-2 cells elongate in response to H. pylori infection. T84 (a, b) and HEp-2 (c, d) cell monolayers were incubated with medium alone (a, c) or with H. pylori strain LC11 (b, d) for 4 h at 37 °C. Both cell lines showed elongated morphology in response to H. pylori infection, comparable to that observed using AGS cells. Cells in (a) and (b) were stained for -actinin and those in (c) and (d) were stained with antibody to detect the cytoplasmic protein Stat6. Findings presented represent the results of three separate experiments. Approximate original magnifications, x600.

H. pylori induces accumulation of Arp3 in AGS cells in a cagE-dependent manner

AGS cells were infected with H. pylori (4 h, m.o.i. 100 : 1) and then stained for Arp3 to determine whether the F-actin-related protein is also affected. As shown by confocal microscopy (Fig. 5), H. pylori-infected AGS cells showed Arp3 protein accumulation that co-localized with adherent bacteria. Cells infected with a cagE mutant strain did not co-localize with Arp3, indicating that H. pylori-induced accumulation of Arp3 protein occurred in a cagE-dependent manner.

View larger version (49K): [in this window] [in a new window]   Fig. 5. Accumulation of Arp3 beneath adherent H. pylori in infected AGS cells. AGS cells were infected with H. pylori strain LC11 (4 h, m.o.i. 100 : 1) and stained for H. pylori (a, d, g; fluorescein-conjugated secondary antibody) and Arp3 (b, e, h; red). Merged images are shown in (c), (f) and (i). AGS cells were uninfected (a–c) or infected with wild-type strain 8823 (d–f) or the mutant 8823cagE : : kan (g–i). Arp3 accumulated beneath adherent bacteria (e, f). cagE is involved in cell elongation because the cagE mutant did not induce cell changes (g–i), compared with the morphological changes induced by the wild-type parental strain (d–f). Images were recorded on a Zeiss LSM 510 confocal microscope. Approximate original magnifications, x640.

	DISCUSSION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This study shows that H. pylori strains induce morphological changes in multiple epithelial cell lines. There was evidence of both stress fibre formation (after 30 min of infection) and elongation of infected cells (after 4 h post-challenge). H. pylori-induced stress fibre formation was independent of the cagA, cagE or VacA status of the infecting strain. In contrast, elongation was observed only when cells were infected with a type I H. pylori strain. Furthermore, this elongation was characterized by Arp3 accumulation beneath adherent bacteria in a cagE-dependent manner.

Stress fibre formation has been reported previously in epithelial cells infected with H. pylori (Segal et al., 1999). Herein, we observed similar effects with both type I and type II H. pylori strains, thereby confirming the work of Palovuori et al. (2000). Cell morphology changes were observed not only using AGS cells, but also in other epithelial cell types, including HEp-2 cells and T84 cells that have previously been used as models of H. pylori infection (Papini et al., 1998; Fahey et al., 2002). These findings indicate that the morphological changes induced by H. pylori are not restricted to gastric-derived tissues.

Isogenic pairs of H. pylori strains were compared to demonstrate that elongation in infected epithelial cells was dependent on cagE. Previous studies have shown the role of CagA in mediating morphological changes (Segal et al., 1999), where the protein is translocated into host cells through a type-IV secretion system to become phosphorylated on tyrosine residues (Censini et al., 2001). Our findings, using type I and type II clinical isolates from infected children, support the observation that CagA is involved in inducing host-cell cytoskeletal rearrangements. CagE, its gene encoded in the cag PAI, mediates the activation of NF-B and interleukin-8 secretion (Maeda et al., 2001). The present study shows that cagE can also be involved in manipulation of the host-cell cytoskeleton. It remains to be determined whether cagE acts as a transporter for CagA and thereby contributes to the observed changes in host-cell morphology (Guillemin et al., 2002). Regardless, these findings indicate that multiple genes on the cag PAI of H. pylori are involved in mediating cytoskeletal rearrangements in infected epithelial cells.

The molecular control of H. pylori-induced cytoskeletal rearrangements involves activation of Rho-GTPase Rac1 and Cdc42 (Churin et al., 2001; Hotchin et al., 2000). The Arp2/3 complex is a seven-subunit protein complex containing two actin-related proteins, Arp2 and Arp3, that initiates the formation of actin-filament networks in response to intracellular signals (Dayel et al., 2001). Bacterial pathogens have developed a variety of mechanisms to utilize the cytoskeleton of host cells to their benefit (Steele-Mortimer et al., 2000). For example, induction of ruffles at the plasma membrane allows the invasion of non-phagocytic cells by Salmonella typhimurium (Jones et al., 1993). Shigella flexneri induces epithelial cell signalling through activation of small GTPases of the Rho family and c-Src to cause major cytoskeletal rearrangements leading to bacterial entry. The surface protein IcsA of Shigella flexneri binds and activates neuronal Wiskoff–Aldrich syndrome protein (N-WASP) to induce F-actin rearrangements in an Arp2/3-dependent mechanism (Sansonetti, 2001). Also, the surface protein ActA of Listeria monocytogenes directly activates the Arp2/3 complex (Cossart, 2000; Boujemaa-Paterski et al., 2001). Our findings indicate that H. pylori induces the accumulation of Arp3 underneath adherent bacteria and suggest, therefore, that Arp3 is important in altering host-cell morphology.

The Yersinia enterocolitica protein invasin binds ₅ß₁ integrin to cause actin polymerization through Cdc42 activation and recruitment of a WASP and Arp2/3 complex (Wiedemann et al., 2001) to achieve internalization into cells (Alrutz et al., 2001). The ₅ß₁ integrin mediates H. pylori adherence and invasion into cultured cells, which also requires tyrosine kinase activity and actin polymerization (Su et al., 1999). Thus, H. pylori binding to ₅ß₁ integrin could lead to cytoskeletal rearrangements through Cdc42 activation and recruitment of the Arp2/3 complex.

In summary, this study shows that, in addition to cagA and VacA, cagE is involved in the cytoskeletal rearrangements of host cells that occur in response to infection with H. pylori strains originally isolated from children. In addition, there is Arp2/3 accumulation beneath adherent bacteria that occurs in a cagE-dependent fashion.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This study was supported by a Canadian Institutes of Health Research (CIHR)-University Industry (Canadian Association of Gastroenterology/AstraZeneca Canada) Research Initiative Award and by an operating grant from the CIHR. P. J. M. C. is the recipient of a CIHR doctoral studentship award and is a CIHR Strategic Training Fellow in Cell Signaling in Mucosal Inflammation and Pain (STP-53877). P. M. S. is the recipient of a Canadian Research Chair in Gastrointestinal Disease.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 

Maeda, S., Akanuma, M., Mitsuno, Y., Hirata, Y., Ogura, K., Yoshida, H., Shiratori, Y. & Omata, M. (2001). Distinct mechanism of Helicobacter pylori-mediated NF-B activation between gastric cancer cells and monocytic cells. J Biol Chem 276, 44856–44864.[Abstract/Free Full Text]