cag+ Helicobacter pylori Induce Transactivation of the Epidermal Growth Factor Receptor in AGS Gastric Epithelial Cells*

The gastric pathogen Helicobacter pylori is known to activate epithelial cell signaling pathways that regulate numerous inflammatory response genes. The aim of this study was to elucidate the pathway leading to extracellular signal-regulated kinase (ERK) 1/2 phosphorylation in H. pylori-infected AGS gastric epithelial cells. We find thatH. pylori, via activation of the epidermal growth factor (EGF) receptor activates the small GTP-binding protein Ras, which in turn, mediates ERK1/2 phosphorylation. cag+ strains ofH. pylori are able to induce greater EGF receptor activation than cag− strains, and studies with isogenic mutants indicate that an intact type IV bacterial secretion system is required for this effect. Blockade of EGF receptor activation using tyrphostin AG1478 prevents H. pylori-mediated Ras activation, inhibits ERK1/2 phosphorylation, and substantially decreases interleukin-8 gene expression and protein production. Investigations into the mechanism of EGF receptor activation, using heparin, a metalloproteinase inhibitor and neutralizing antibodies reveal that H. pylori transactivates the EGF receptor via activation of the endogenous ligand heparin-binding EGF-like growth factor. Transactivation of gastric epithelial cell EGF receptors may be instrumental in regulating both proliferative and inflammatory responses induced by cag+ H. pylori infection.

Helicobacter pylori is a pathogenic Gram-negative bacterium that colonizes human gastric mucosa. H. pylori infection has been connected etiologically to peptic ulcer disease (1), mucosaassociated lymphoid tissue lymphoma of the stomach (2), and gastric adenocarcinoma (3). H. pylori colonization invariably causes chronic active gastritis, a disease state characterized by neutrophil infiltration of the gastric mucosa and epithelial layer. Interaction of the bacterium with gastric epithelial cells leads to the production of chemokines, such as interleukin-8 (IL-8), 1 which in turn causes activation and recruitment of neutrophils to the site of infection.
It is now known that attachment of H. pylori to gastric epithelial cells activates multiple signaling pathways that culminate in IL-8 gene transcription. Previous studies have shown that H. pylori is able to induce activation of the transcription factors NF-B and AP-1, key regulators of many inflammatory genes including IL-8 (4 -7). Moreover, we and others, have recently reported that infection of gastric epithelial cell lines with H. pylori results in the rapid activation of p38, JNK, and ERK1/2 mitogen-activated protein (MAP) kinases (6 -8). However, the mechanisms whereby H. pylori are able to activate gastric epithelial cell MAP kinases are still unclear.
The ERK1/2 pathway has been linked to cellular proliferation and differentiation (9), however, our previous findings suggest these proteins also participate in proinflammatory cellular responses in gastric epithelial cells. By blocking upstream ERK1/2 phosphorylation using the MEK inhibitor PD98059 we were able to reduce the amount of IL-8 produced by AGS cells in response to H. pylori infection. Interestingly, this effect was not mediated via blockade of NF-B activation (8). However, it has been demonstrated that H. pylori-mediated AP-1 activation is prevented by PD98059 (6).
The activity of ERK1/2 are generally regulated through the activation of cell surface receptors. One possible mechanism whereby H. pylori may modulate the ERK1/2 pathway is through the EGF receptor. The EGF receptor is a transmembrane receptor with intrinsic tyrosine kinase activity, known to regulate the ERK1/2 pathway, via activation of the small GTPbinding protein Ras. Numerous stimuli are known to cause transactivation of the EGF receptor, including Substance P (10), bradykinin (11), thrombin (12), insulin-like growth factor-1 (13), UV light (14), the dermatonecrotic toxin produced by Pasteurella multocida (15) and the invasive bacterium Salmonella typhimurium (16). Here, we report that H. pylori is capable of inducing EGF receptor phosphorylation, which is, in part, responsible for ERK1/2 activation in H. pylori-infected AGS cells. In this study, we also examine the downstream effects of EGF receptor activation on IL-8 regulation by H. pylori, and explore the mechanism by which this bacterium induces EGF receptor phosphorylation.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-AGS gastric epithelial cells (American Type Culture Collection, Rockville, MD) were grown in F-12 HAM medium (pH 7.4: Sigma-Aldrich) supplemented with 10% fetal bovine serum, 100 units/ml penicillin G sodium, and 100 g/ml streptomycin sulfate. All cultures were maintained at 37°C in a humidified atmosphere of 95% air and 5% CO 2 . Cell culture experiments were carried out using 100-mm dishes and 12-or 24-well polypropylene tissue culture plates (Corning Costar, Cambridge, MA). All experiments were carried out using confluent monolayers unless otherwise stated. In some experiments, cells were treated with AG1478 (Calbiochem, La Jolla, CA), batimastat (British Biotech, Oxford, UK), HB-EGF neutralizing antibody, EGF (both from R&D systems, Minneapolis, MN), heparin or cyclohexamide (both from Sigma-Aldrich).
H. pylori Clinical Isolates and Isogenic Mutants-H. pylori were plated onto Brucella agar supplemented with 5% horse blood (BBL, Becton Dickinson Microbiology, Cockeysville, MD) and incubated at 37°C in a microaerophilic environment. After 3 days the bacteria were harvested into pyrogen-free Dulbecco's phosphate-buffered saline (Cellgro, Mediatech, Herndon, VA). The bacteria were then pelleted by centrifugation at 4,000 ϫ g for 10 min, and resuspended into antibioticfree F-12 Ham's medium. Unless otherwise stated experiments were performed using the cagϩ, vacuolating cytotoxin secreting H. pylori strain 43504 (American Type Culture Collection).
Isogenic H. pylori mutants lacking the picB, cagA, or vacA genes were also studied together with their parental cagϩ, vacuolating cytotoxin secreting wild type strain (number 60190) (17). These strains plus the cagϪ strain J44, were obtained from the culture collection of the Vanderbilt University Campylobacter and Helicobacter Laboratory (Nashville, TN) and have been described previously (17,18). H. pyloriconditioned media were prepared by suspending bacteria in antibioticfree medium for 1 h at 37°C, pelleting at 4,000 ϫ g for 10 min, and then filtering the medium through a 0.2-m filter (Acrodisc; Gelman, Ann Arbor, MI).
Detection of EGF Receptor Phosphorylation-AGS gastric epithelial cells plated onto 100-mm dishes were serum-starved for 24 h prior to experiments. Cells were then incubated with H. pylori (1 ϫ 10 9 ) bacteria for the indicated times. Monolayers were washed 5 times with ice-cold phosphate-buffered saline and lysed (45 min on ice) using 1 ml of lysis buffer. Lysis buffer contained 50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EGTA, 1 mM NaF, 100 M sodium orthovanadate, 100 M phenylmethylsulfonyl fluoride, and a commercially available protease inhibitor mixture tablet (Complete; Roche Molecular Biochemical, Indianapolis, IN). Cells were then scraped, and transferred to Eppendorf tubes. Particulate material was removed by centrifugation, and the lysates collected. EGF receptor immunoprecipitations were performed using 1 mg of lysate protein, which was incubated for 1 h at 4°C with 2 g of a monoclonal anti-EGF receptor antibody (Santa Cruz), followed by an overnight incubation at 4°C with a 50-l aliquot of Protein G Plus-agarose beads (Santa Cruz). The beads were washed three times with lysis buffer, and proteins were eluted by boiling for 5 min in 2 ϫ SDS sample buffer. Immunoprecipitated proteins were fractionated using 6% SDS-PAGE, and transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA). Membranes were then blocked for 1 h with 5% skim milk in Trisbuffered saline with Tween 20 (0.1%) (TBS-T), and incubated overnight with the appropriate antibodies. Phosphorylated tyrosine residues were detected using monoclonal antibody PY99 (Santa Cruz) at a 1:1000 dilution. Blots were stripped and reprobed with a rabbit anti-EGF receptor polyclonal antibody (1:1000) (Santa Cruz) to detect total protein levels.
Western Blot Analysis of Phospho-specific MAP Kinase Activation-AGS cells were grown on 12-well plates and maintained in serum-free medium for 24 h prior to the experiment. H. pylori were added in serum-free medium, and co-cultured for varying lengths of time. At the end of the experiment the monolayers were washed 3 times with phosphate-buffered saline and lysed with SDS sample buffer (containing 62.5 mM Tris-HCl, pH 6.8, 2% w/v SDS, 10% glycerol, 50 mM dithiothreitol, 0.1% (w/v) bromphenol blue). Samples were then sonicated, heated to 100°C for 5 min, and ϳ20 l of lysate loaded onto a 10% SDS-PAGE gel. After electrophoresis, the proteins were transferred onto nitrocellulose membranes and blocked for 1 h at room temperature with a 5% (w/v) solution of dried milk in Tris-buffered saline, pH 7.4, with 0.1% Tween 20 (TBS-T). This was followed by an overnight incubation at 4°C with the phospho-specific MAP kinase antibodies diluted 1:1000 in blocking buffer. The membranes were then washed 3 times with TBS-T, and incubated at room temperature for 1 h with peroxidase-conjugated goat anti-rabbit IgG (1:3000 dilution; Santa Cruz Biotechnology). A SuperSignal chemiluminescent substrate (Pierce, Rockford, IL) was used for detection.
A phospho-specific p44/p42 MAP kinase antibody was used to detect activated ERK1/2. This antibody detects p44 and p42 MAP kinase (ERK1and ERK2) only when they are catalytically activated by phosphorylation at Thr 202 and Tyr 204 . A phospho-specific p38 MAP kinase antibody was used to detect p38 activated by phosphorylation at Thr 180 and Tyr 182 . A phospho-specific p54/p46 MAP kinase antibody was used to detect JNK. This antibody detects p54 and p46 MAP kinase only when they are activated by phosphorylation at Thr 183 and Tyr 185 . All three antibodies were obtained from Cell Signaling Technology, Beverly, MA.
Preparation of an AGS Cell Line Stably Transfected with an IL-8 Reporter Gene-A 1521-bp fragment containing nucleotides Ϫ1481 to ϩ40 of the promoter region of the IL-8 gene was cloned into the pGL2basic luciferase expression vector between KpnI and HindIII restriction sites. The sequence was confirmed by DNA sequencing using primers specific for the pGL2-basic luciferase expression vector (GL primers 1 and 2; Promega Corp., Madison, WI). The IL-8 luciferase reporter gene and pcDNA3.1 vector (Invitrogen, Carlsbad, CA) expressing a geneticin (G418) resistance gene were co-transfected into AGS gastric epithelial cells using LipofectAMINE according to the manufacturer's instructions (Life Technologies, Gaithersburg, MD). AGS clones stably transfected with pcDNA3.1 were selected by adding G418 sulfate (400 g/ml; Life Technologies) to the culture medium, and positive clones co-transfected with the IL-8 luciferase reporter were identified by their ability to respond to IL-1␤ (10 ng/ml; R&D Systems). The AGS clone identified as containing both the geneticin resistance gene and the IL-8 luciferase reporter gene was subsequently maintained in G418 selection media.
Statistical Analyses-Statistical analyses were performed using Sig-maStat for windows version 2.0 (Jandel Scientific Software, San Rafael, CA). Unless stated otherwise, ANOVA followed by protected t tests were used for intergroup comparisons.
Ras Activation Assay-Ras activation was determined using a commercially available kit (Upstate Biotechnologies, Lake Placid, NY). Briefly, 1 mg of lysate protein from treated cells, precleared with 20 l of glutathione-agarose (Santa Cruz), for 30 min at 4°C, was incubated for 30 min at 4°C with 20 l of Raf-1 RBD agarose. The beads were washed 3 times with lysis buffer and resuspended in 40 l of SDS sample buffer, boiled for 5 min, and then loaded onto a 12% SDS-PAGE. Proteins were transferred to nitrocellulose and the membrane was blocked for 20 min with 5% skim milk in phosphate-buffered saline/ Tween (Tween 0.05%). The blot was incubated overnight at 4°C with an anti-Ras monoclonal antibody at a 1:1000 dilution. Activated Ras was then visualized using an horseradish peroxidase-conjugated anti-mouse secondary antibody (Santa Cruz), and a chemiluminescence detection system.
Preparation of Ras-17N-expressing Retroviruses and Infection of AGS Cells-To construct a retroviral vector expressing the dominant negative Ras (Ras-17N), the human Ras-17N fragment was first excised from pZIP-Ras-17N by digestion with BspHI and BamHI. The resulting fragment was then ligated into the retroviral vector pMMP, digested with NcoI and BamHI. To generate Ras-17N-expressing retroviruses, 293T cells were seeded at 5 ϫ 10 6 cells per 100-mm plate and incubated with 10 ml of 10% fetal bovine serum/Dulbecco's modified Eagle's medium for 24 h. The medium was replaced with fresh medium 4 h prior to transfection. The plasmids pMD-gag-pol, pMD-VSVG, and pMMP-Ras-17N were combined in a ratio of 3:1:4 and used to prepare transfection mixtures using Effectene Transfection Reagent (Qiagen Inc., Chatsworth, CA) according to the manufacturer's instructions. Fortyeight h after transfection, the media were collected, filtered through 0.45-m disc filters, and the supernatants were either used immediately or stored at Ϫ80°C. A control retrovirus containing LacZ was generated in an identical manner. pMMP-LacZ, pMD-gag-pol, and pMD-VSVG were all kindly provided by Dr. Richard C. Mulligan (Children's Hospital, Harvard Medical School, Boston, MA).

H. pylori Induces Phosphorylation of the EGF Receptor-To
determine whether the EGF receptor is activated by H. pylori infection, lysates from AGS cells infected with H. pylori for 1 h were immunoprecipitated with a monoclonal anti-EGF receptor antibody. Western blot analysis of the immunoprecipitate using the phosphotyrosine-specific antibody (PY99) demonstrated the presence of a tyrosine-phosphorylated protein of approximately 170-kDa (Fig. 1A, upper panel). Pretreatment of the cells with the EGF receptor inhibitor tyrphostin AG1478 (1 M) for 30 min was able to completely prevent H. pylori-mediated EGF receptor phosphorylation. To confirm that the immunoprecipitated protein was the EGF receptor the blot was rep-robed with an anti-EGF receptor polyclonal antibody (Fig. 1A,  bottom panel). These data show that H. pylori is able to induce phosphorylation of the EGF receptor in AGS cells after 1 h of infection. Time course experiments showed an increase in tyrosine phosphorylation of the EGF receptor as early as 15 min following infection (data not shown).
The EGF Receptor Inhibitor Tyrphostin AG1478 Reduces ERK1/2 Phosphorylation in H. pylori-infected AGS Cells-Phosphorylation of the EGF receptor is known to activate the Ras/Raf/MEK pathway resulting in the phosphorylation of ERK1/2. To determine whether the ERK1/2 activation we had observed previously in H. pylori-infected AGS cells (8) was mediated through activation of the EGF receptor; we used tyrphostin AG1478, an agent that specifically prevents EGF receptor kinase activation.
As shown in Fig. 1B (upper panel) H. pylori was able to induce phosphorylation of ERK1/2 by 15 min and this effect was significantly reduced when the cells were pretreated with AG1478 (1 M). H. pylori-mediated ERK1/2 phosphorylation was prevented at 15 min by AG1478 and reduced by 50% at the 30-min time point as determined by densitometry. Since crosstalk between signal transduction pathways is common, we examined whether this inhibitor could also prevent the phosphorylation of p38 or JNK in H. pylori infected AGS cells. However, in contrast to our findings for ERK1/2, AG1478 had no evident effect on p38 or JNK phosphorylation levels (Fig. 1B, middle and lower panels). These data indicate that H. pylori induced activation of ERK1/2 is, at least in part, due to phosphorylation of the EGF receptor, whereas EGF receptor activation does not mediate H. pylori-induced p38 or JNK phosphorylation.
H. pylori Activated IL-8 Gene Expression and Protein Production Is Down-regulated by Blockade of EGF Receptor Phos-phorylation-Previously, we demonstrated that activation the ERK1/2 pathway by H. pylori is one of the mechanisms whereby the bacterium is able to regulate cytokine production (8). We hypothesized that blockade of the EGF receptor as an upstream activator of the ERK1/2 pathway may have functional effects on IL-8 gene regulation. To investigate this hypothesis, we used AGS cells stably transfected with an IL-8 luciferase reporter gene. As shown in Fig. 2A, treatment of these cells with H. pylori for 4 h caused a 30-fold increase in IL-8 luciferase reporter activity. However, reporter activity was reduced to ϳ22-fold when EGF receptor phosphorylation was prevented by treatment with AG1478 (1 M). Consistent with our transcription data, there was a dramatic up-regulation of IL-8 protein production by AGS cells in response to infection with H. pylori for a 4-h period, which was reduced dose dependently when the cells were pretreated with the EGF receptor inhibitor AG1478 (Fig. 2B).
H. pylori-mediated Activation of the EGF Receptor in AGS Gastric Epithelial Cells Is Dependent on cag Status-We have previously observed that both cagϩ and cagϪ strains of H. pylori were capable of activating the ERK1/2 pathway, but differ in the degree of activation (8). The ERK1/2 phosphorylation seen in AGS cells infected with cagϪ H. pylori was considerably reduced compared with those infected with cagϩ strains. Therefore, we next examined whether activation of the EGF receptor was also dependent on the cag status of the bacteria. Fig. 3 shows the level of EGF receptor tyrosine phosphorylation when cells were infected with either a cagϩ strain (43504) or cagϪ strain (J44) of H. pylori over a 4-h time course. We find that the cagϪ strain induces a relatively weak activation of the EGF receptor, as compared with the more substantial activation seen with cagϩ strain. These data appear to correlate with the ability of the cagϩ and cagϪ H. pylori to FIG. 1. H. pylori promotes phosphorylation of the EGF receptor, which mediates ERK1/2 phosphorylation. A, serum-starved confluent monolayers of AGS cells in 100-mm plates were infected with 1 ϫ 10 9 H. pylori cagϩ strain 43504 for 1 h. Some AGS cells were pretreated with tyrphostin AG1478 (1 M), a specific EGF receptor inhibitor, for 30 min prior to infection with H. pylori. The EGF receptor was immunoprecipitated (IP) with a monoclonal anti-EGF receptor antibody. The immunoprecipitated proteins were resolved on a 6% SDS-PAGE gel, and transferred to nitrocellulose. Phosphorylated EGF receptor (upper panel) was detected by Western blotting (WB) with an anti-phosphotyrosine monoclonal antibody (PY99). EGF receptor levels in the immunoprecipitates (lower panel) were assessed by reprobing with an anti-EGF receptor polyclonal antibody. B, serumstarved confluent monolayers of AGS cells in 12-well plates were pretreated with or without AG1478 at 1 M for 30 min. The cells were then infected with 1 ϫ 10 8 H. pylori cagϩ strain 43504 over a 30-min time course. Whole cell lysates were run on a 10% SDS-PAGE gel, and transferred to nitrocellulose. Levels of phosphorylated ERK1/2, p38, and JNK were assessed by Western blotting with phosphospecific antibodies. Blots were reprobed with control ERK1/2, p38, and JNK antibodies to demonstrate equal loading.
EGF Receptor Activation by H. pylori Is Dependent Upon an Intact cag Secretion System, But Is Independent of the cagA and vacA Gene Products-Having found a distinct difference between cagϩ and cagϪ H. pylori in their ability to induce activation of the EGF receptor, we next examined whether isogenic mutants of H. pylori could induce the same response as a wild type cagϩ strain (60190). As demonstrated in Fig. 4, phosphorylation of the EGF receptor by the picBϪ mutant was considerably less than for the isogenic wild type H. pylori strain. In contrast, the cagAϪ mutant induced a similar activation of the EGF receptor when compared with the wild type strain. These data suggest that full activation of the EGF receptor by H. pylori requires an intact type IV secretory apparatus, but not the translocation of the CagA protein into the host cell.
Since it has been reported that culture supernatant from vacAϩ H. pylori was able to cause dephosphorylation of the EGF receptor (21), we also investigated the effect of a vacAϪ mutant of H. pylori on gastric epithelial cell EGF receptor phosphorylation. As shown in Fig. 4, the absence of the vacA gene had no evident effect on the ability of the bacteria to induce EGF receptor phosphorylation, as compared with the vacAϩ wild type strain.
H. pylori Causes Ras Activation That Is Prevented by the EGF Receptor Inhibitor AG1478 -Activation of the EGF receptor, as well as other cell surface receptors, stimulates the exchange of GDP for GTP on the small G protein Ras. Once in the active state, Ras can interact with several effector proteins such as Raf and phosphatidylinositol 3-kinase. Active Raf stimulates MEK, which in turn leads to the phosphorylation of ERK1/2.
To further delineate the mechanism leading to ERK1/2 phosphorylation, we investigated whether Ras becomes activated following H. pylori infection of AGS cells, and whether we could prevent this activation using AG1478. Our results (Fig. 5A,  upper panel) show a marked activation of Ras, at 30 min, which is inhibited by pretreatment of the cells with AG1478 (1 M). EGF (10 ng/ml), included as positive control, was found to stimulate Ras activation after 10 min. EGF activation of Ras was also prevented by pretreatment with AG1478. Whole cell lysates from these experiments were also probed for phosphospecific and control ERK1/2 to examine the correlation between Ras activation and ERK1/2 phosphorylation and to demonstrate equivalent levels of protein loading for the Ras assay (Fig. 5A, middle and bottom panels, respectively). Interestingly, EGF-mediated ERK1/2 phosphorylation was completely prevented with AG1478, whereas H. pylori activated phospho-ERK1/2 levels, although significantly reduced after AG1478 treatment, were still detectable. This finding suggests that there may be more than one mechanism involved in H. pylori activation of the ERK1/2 pathway. Dominant Negative Ras Inhibits H. pylori-mediated ERK1/2 Phosphorylation-We next examined whether Ras activation by H. pylori was upstream of ERK1/2 phosphorylation. To do this we overexpressed dominant negative Ras in AGS cells using retroviral transfection, and examined ERK1/2 phosphorylation in response to H. pylori infection. As shown in Fig. 5B, ERK1/2 phosphorylation is absent in untreated control cells. However, H. pylori and EGF (10 ng/ml) each caused a strong induction of ERK1/2 phosphorylation in the LacZ (vector control)-transfected cells. EGF activation of ERK1/2 was almost completely abrogated in cells over expressing dominant negative Ras, whereas H. pylori-mediated ERK1/2 activation was only partially inhibited. These findings provide further evidence that ERK1/2 activation by H. pylori can be mediated via both Ras-dependent and Ras-independent pathways.
H. pylori-conditioned Medium Is Unable to Cause EGF Receptor Phosphorylation-To examine whether soluble factors secreted by H. pylori were responsible for phosphorylation of the EGF receptor, we incubated AGS cells with both whole bacteria or conditioned medium for 1 h. H. pylori-conditioned medium was unable to induce EGF receptor phosphorylation (Fig. 6), suggesting that contact between the bacterium and the host cell is necessary for induction of EGF receptor phosphorylation. These data agree with our previous finding, that H. pylori-conditioned medium is unable to induce ERK1/2 phosphorylation (8).
H. pylori-mediated EGF Receptor Phosphorylation Is Prevented by Treatment with Heparin-Previous studies have shown that, in a number of systems where the EGF receptor becomes transactivated, the underlying mechanism involves cleavage of membrane-bound pro-HB-EGF (22). Cleaved HB-EGF then binds the EGF receptor leading to its phosphorylation. HB-EGF requires heparan sulfate proteoglycans as coreceptors of the EGF receptor and the addition of heparin has been shown to compete with heparan sulfate proteoglycans for HB-EGF binding (23). We therefore infected AGS cells with H. pylori in the presence of heparin (100 g/ml). As shown in Fig.  6, heparin markedly inhibited H. pylori-mediated EGF receptor phosphorylation.
HB-EGF Neutralizing Antibodies Prevent H. pylori-mediated EGF Receptor Phosphorylation, and Reduce ERK1/2 Phosphorylation-The data resulting from the heparin co-culture experiments led us to examine whether HB-EGF, an endogenous ligand of the EGF receptor, was mediating EGF receptor activation in H. pylori-infected AGS cells. In support of this hypothesis we found that a HB-EGF neutralizing antibody (25 g/ml) almost completely blocked H. pylori-mediated activation of the EGF receptor (Fig. 7A). The HB-EGF neutralizing antibody was also found to cause a substantial (48.5% by densitometry) reduction in H. pylori-activated ERK1/2 phosphorylation (Fig. 7B).

Batimastat, a Metalloproteinase Inhibitor, Prevents EGF Receptor Activation and Reduces the ERK1/2 Phosphorylation
Induced by H. pylori-Shedding of HB-EGF and other EGF receptor ligands has previously been reported to be dependent on the action of matrix metalloproteinases. We explored the possibility that H. pylori may be activating these proteases, and thus cause the shedding of the soluble form of HB-EGF. We investigated this by treating the cells with batimastat (5 g/ ml), a broad spectrum matrix metalloproteinase inhibitor. As shown in Fig. 8A, we found that batimastat could prevent EGF receptor phosphorylation by H. pylori. In addition, we found that this inhibitor could also markedly reduce ERK1/2 phosphorylation induced by H. pylori (Fig. 8B). DISCUSSION We and others have previously reported that H. pylori activates a number of MAP kinases in gastric epithelial cell lines (6 -8, 24). Activation of these pathways plays a key role in up-regulating the expression of the proinflammatory cytokine IL-8 (8). However, the exact mechanisms whereby H. pylori activates these signaling pathways are still unknown. We now   FIG. 3. cag؉ and cag؊ strains of H. pylori cause a differential activation of the EGF receptor. Serum starved, confluent monolayers of AGS cells in 100-mm plates were infected with 1 ϫ 10 9 of either a cagϩ strain (43504) or a cagϪ strain (J44) of H. pylori, over a 4-h time course. The EGF receptor was immunoprecipitated (IP) with a monoclonal anti-EGF receptor antibody. Immunoprecipitated proteins were run on a 6% SDS-PAGE gel, and transferred to nitrocellulose. Phosphorylated EGF receptor levels, shown in the upper panel, were detected by Western blotting (WB) with an anti-phosphotyrosine monoclonal antibody (PY99). EGF receptor levels, as shown in the second panel, were assessed by reprobing with a rabbit anti-EGF receptor polyclonal antibody. The third panel demonstrates ERK1/2 phosphorylation levels in whole cell lysates from these experiments and control ERK1/2 levels are shown in the bottom panel.

FIG. 4. An intact H. pylori type IV bacterial secretion system, but not CagA or VacA, is required for maximal activation of the EGF receptor.
A, serum-starved confluent monolayers of AGS cells in 100-mm plates were infected with 1 ϫ 10 9 wild type H. pylori cagϩ strain 60190, or its isogenic mutants picBϪ, cagAϪ, or vacAϪ for 1 h. The EGF receptor was immunoprecipitated (IP) with a monoclonal anti-EGF receptor antibody. Immunoprecipitated proteins were run on a 6% SDS-PAGE gel, and transferred to nitrocellulose. Phosphorylated EGF receptor, shown in the upper panel, was detected by Western blotting (WB) with an anti-phosphotyrosine monoclonal antibody (PY99). EGF receptor levels were then assessed by reprobing the blot with an anti-EGF receptor polyclonal antibody, as shown in the lower panel. B, the graph shows the level of EGF receptor phosphorylation in each sample, as quantified by densitometry.
report that H. pylori can induce phosphorylation of the EGF receptor in AGS gastric epithelial cells. H. pylori-induced EGF receptor phosphorylation leads to activation of Ras, which is able to mediate ERK1/2 phosphorylation, which in turn upregulates IL-8 gene expression and protein production. We find that EGF receptor phosphorylation mediated by H. pylori is dependent upon an intact type IV bacterial secretory system. Moreover, we find that the mechanism underlying the induction of EGF receptor phosphorylation by H. pylori involves activation of the EGF receptor ligand HB-EGF.
One important finding of our study is that there is a differential activation of the EGF receptor depending upon whether the H. pylori strain processes a 40-kb region of genes known as the cag pathogenicity island. The cag pathogenicity island encodes for ϳ30 proteins, that based on sequence homology appear to constitute a type IV bacterial secretion system (25). It has previously been reported that the absence of the whole FIG. 6. H. pylori-induced activation of the EGF receptor is not mediated by a soluble bacterial factor, and can be prevented by heparin. Serum-starved confluent monolayers of AGS cells in 100-mm plates were treated with 1 ϫ 10 9 H. pylori cagϩ strain 43504, conditioned media (as described under "Experimental Procedures") or with H. pylori plus heparin (100 g/ml) for 1 h. Analysis of the phosphorylation state of the EGF receptor was performed by immunoprecipitation of the receptor followed by Western blotting with the anti-phosphotyrosine antibody PY99 (upper panel). The blot was then reprobed with a polyclonal anti-EGF receptor antibody (lower panel).

FIG. 7. H. pylori-induced EGF receptor activation is mediated by HB-EGF.
A, serum-starved confluent monolayers of AGS cells in 100-mm plates were treated with 1 ϫ 10 9 H. pylori cagϩ strain 43504 for 1 h. Prior to infection, some cells were pretreated for 30 min and then co-treated with 25 g/ml HB-EGF neutralizing antibody. EGF receptor phosphorylation (upper panel) was determined by immunoprecipitation of the receptor, followed by Western blotting with the antiphosphotyrosine antibody PY99. The blot was reprobed to demonstrate equal levels of EGF receptor in each sample (lower panel). B, serumstarved confluent monolayers of AGS cells in 12-well plates were treated with 1 ϫ 10 8 H. pylori cagϩ strain 43504 for 1 h. Prior to infection, some cells were pretreated for 30 min and then co-treated with 25 g/ml HB-EGF neutralizing antibody. Whole cell lysates from these experiments were analyzed by Western blotting with both phospho-specific ERK1/2 (upper panel) and control ERK1/2 antibodies (lower panel) .   FIG. 5. H. pylori induces activation of GTP-Ras, which mediates ERK1/2 phosphorylation. A, AGS cells grown to ϳ90% confluence on 100-mm plates were serum starved overnight. Cells were then either infected with 1 ϫ 10 9 H. pylori cagϩ strain 43504 over a 2-h time course or treated with EGF (10 ng/ml) for 10 min. Some cells were pretreated with AG1478 (1 M) prior to treatment with H. pylori or EGF. Lysates were subjected to immunoprecipitation with Raf-1 RBD-agarose; GTP-Ras was then detected using a monoclonal anti-Ras antibody (upper panel). Whole cell lysates from the same experiments were probed with both a phospho-specific ERK1/2 antibody (middle panel), to assess levels of phosphorylated ERK1/2 in the lysates and also a control ERK1/2 antibody (bottom panel) to demonstrate equal loading. B, AGS cells were transfected for 48 h with retroviruses carrying either dominant negative Ras or control plasmid LacZ. Transfected cells were then infected with 1 ϫ 10 8 H. pylori cagϩ strain 43504 for 1 h or treated with 10 ng/ml EGF for 10 min. Phosphorylated ERK1/2 was then assessed in whole cell lysates using a phospho-specific ERK1/2 antibody (upper panel). The blot was reprobed with control ERK1/2 to ensure equal loading of protein (middle panel) and with a monoclonal anti-Ras antibody to demonstrate overexpression of dominant negative Ras (bottom panel).
pathogenicity island or deletions of individual cag genes result in an impaired ability to activate NF-B, AP-1, and MAP kinase pathways (6 -8, 26). We have found that the picBϪ (cagEϪ) isogenic mutant of H. pylori is able to mediate a much weaker induction of EGF receptor phosphorylation compared with its parental wild type strain. These data suggest that the H. pylori type IV secretion system participates in host cell EGF receptor activation.
We also examined whether the CagA protein produced by H. pylori could be involved in EGF receptor activation. To date, CagA is the only H. pylori protein known to translocate from the bacterium into the cell via the type IV secretion system; however, its function or target within the host eukaryotic cell is unknown (27)(28)(29). We have demonstrated through the use of an isogenic cagAϪ mutant that this protein does not appear to play a role in inducing EGF receptor phosphorylation. This finding is consistent with our current knowledge of CagA, in that the CagA protein itself does not induce an inflammatory response, and plays no role in MAP kinase activation (8), NF-B activation (26), or IL-8 production (25).
From our study we also conclude that the vacuolating toxin produced by H. pylori plays no role in mediating H. pyloriinduced EGF receptor phosphorylation, since we found no differences between our wild type and vacAϪ mutants in their abilities to induce EGF receptor phosphorylation. It has been reported that H. pylori supernatant containing vacuolating toxin can cause dephosphorylation of the EGF receptor, and down-regulate ERK1/2 activation in EGF-treated Kato III cells (21). In contrast to this, we find that whole intact H. pylori were necessary to induce phosphorylation of the EGF receptor and activate the ERK1/2 signaling pathway. We did not, however, examine the effects of purified vacuolating toxin in our system or investigate the effects of H. pylori supernatant on EGFtreated cells. Instead, we focused on the early signaling events that occur upon contact of H. pylori with gastric epithelial cells.
We have demonstrated for the first time that, through induction of EGF receptor phosphorylation H. pylori is capable of activating the small GTP-binding protein Ras, which activates a signaling cascade resulting in ERK1/2 phosphorylation. However, in our experiments using an EGF receptor inhibitor and dominant negative Ras we were unable to achieve complete blockade of ERK1/2 phosphorylation induced by H. pylori. This suggests that additional pathways exist through which the bacteria are able to induce ERK1/2 phosphorylation. This hypothesis is supported by a recent study showing that H. pylori regulates the histidine decarboxylase gene through ERK1/2 activation, but via a Ras independent pathway (24).
We have found that H. pylori induced phosphorylation of the EGF receptor is mediated through the release of EGF receptor ligand: HB-EGF. Previous studies have also examined the association between EGF receptor ligands and H. pylori. Romano et al. (30) reported that by incubating suspensions of H. pylori or conditioned broth with MKN-28 cells for several hours they were able to up-regulate both amphiregulin and HB-EGF mRNA levels. We find that rapid activation of the EGF receptor by H. pylori (within 15 min) is not mediated through the release of soluble bacterial factors, but instead appears to be contact dependent. Rapid release of HB-EGF from the cell is typically a result of post-translational modifications of membrane-bound HB-EGF, and does not appear to be mediated via production of newly synthesized proteins. Furthermore, replenishment of transmembrane HB-EGF requires 12-24 h (23). As our system examined the acute interaction between H. pylori and gastric epithelial cells we cannot rule out the possibility that soluble factors secreted by the bacteria may play a role in up-regulating EGF receptor ligand production during the later stages of infection.
Transactivation of the EGF receptor by numerous G-proteincoupled receptors has now been identified as a critical element in G-protein-coupled receptor-induced mitogenic signaling (13,22,(31)(32)(33). The mechanism for activation of the EGF receptor under these circumstances appears to be through matrix metalloproteinases. Matrix metalloproteinases such as MMP-3, once activated, are able to cleave the membrane-bound precursor of HB-EGF causing shedding of its mature, soluble form, which in turn engages and activates the EGF receptor, and thus activates the MAP kinase cascade (22). We now report the novel finding that H. pylori-induced EGF receptor phosphorylation is mediated through HB-EGF, and is prevented by the metalloproteinase inhibitor batimastat. Engagement of the host gastric epithelial cell by H. pylori appears to induce HB-EGF release, which in turn activates EGF receptor and ERK1/2 signaling. We speculate that H. pylori may cause transactivation of the EGF receptor by interacting with a receptor already expressed on the surface of the gastric epithelial cell, or through insertion of bacterial proteins into the host cell. HB-EGF-mediated transactivation of the EGF receptor may be instrumental in inducing the augmented proliferative and inflammatory responses seen in infection of the gastric epithelium by cagϩ H. pylori.