Differential Activation of Mitogen-Activated Protein Kinases in AGS Gastric Epithelial Cells by cag⁺ and cag⁻ Helicobacter pylori¹

Helicobacter pylori chronically infects over half the world’s population, but in the majority of cases, infection results only in asymptomatic chronic active gastritis (1). Symptomatic gastroduodenal disease, specifically peptic ulceration or gastric neoplasia, develops in only approximately 10% of infected individuals (2, 3, 4). The host’s immune and inflammatory response to bacterial virulence factors is likely to play a critical role in determining the clinical outcome of H. pylori infection (5, 6). In this study we examine the ability of H. pylori to activate mitogen-activated protein (MAP)³ kinases, key elements in the regulation of cellular responses to external inflammatory and proliferative signals. We find that different strains of H. pylori vary in their ability to activate MAP kinase pathways in AGS gastric epithelial cells. The differential activation of MAP kinase and other host cellular signaling pathways is a possible mechanism for strain-specific variations in the outcome of gastric H. pylori infection.

MAP kinases are a family of ubiquitous, highly conserved, cell signaling molecules (7, 8, 9). Upon activation by upstream kinases, MAP kinases phosphorylate downstream kinases and/or mediators, including transcription factors. MAP kinases can be activated by a wide variety of extracellular stimuli and transmit signals from the cell surface to the nucleus to regulate gene expression. Key cellular functions that are regulated at least in part by MAP kinase signaling include cell proliferation, cell survival, and cytokine production. Three main groups of MAP kinases have been characterized to date: the extracellular signal-regulated kinases (ERK), the c-Jun N-terminal kinases (JNK), and the p38 MAP kinases (7, 8, 9). These MAP kinase subfamilies form three parallel cascades that can be activated simultaneously or independently. ERK MAP kinases are strongly activated by growth factors and phorbol ester, but weakly by inflammatory stimuli. In contrast, JNK and p38 MAP kinases are stimulated by inflammatory cytokines and stress stimuli, but minimally by growth factors.

We and others have shown previously that H. pylori infection activates IL-8 gene expression in gastric epithelial cells in vitro and in vivo (10, 11, 12, 13, 14, 15, 16). IL-8 mRNA and protein levels are increased in the gastric mucosa of patients with H. pylori gastritis, and immunohistochemical studies demonstrate increased IL-8 protein in gastric epithelial cells from infected individuals (11, 13, 16, 17). IL-8 is a potent neutrophil-activating chemotactic cytokine or chemokine. Thus, IL-8 release by infected gastric epithelial cells may be instrumental in regulating neutrophil infiltration of the gastric mucosa in H. pylori gastritis. H. pylori also increases IL-8 mRNA levels and protein production in cultured monolayers of AGS and other gastric epithelial cell lines (10, 12, 13, 15, 18). Contact with the bacterium initiates epithelial cell signaling events that result in activation of the transcription factor NF-κB (10, 14, 15, 19, 20, 21, 22). Activated NF-κB then translocates to the nucleus where it up-regulates IL-8 gene transcription. NF-κB activation occurs within minutes of H. pylori infection of AGS gastric epithelial cells in vitro, and activated NF-κB is also evident in gastric epithelial cells from patients with H. pylori gastritis (10, 14, 15, 19, 22).

The cag pathogenicity island of H. pylori is a 40-kb region immediately upstream to the cagA gene that encodes over 40 putative bacterial proteins (12). Based on sequence homology, cag region gene products appear to constitute a bacterial secretion system that may be involved in the export or surface expression of bacterial virulence factors. In the developed world, approximately 70% of clinical H. pylori isolates carry the cag pathogenicity island. Carriage of cag⁺ strains has been associated in most published studies with more severe gastritis and a greater risk for peptic ulceration and gastric cancer than H. pylori cag⁻ infection (6, 12, 23, 24, 25, 26). Gene products of the cag pathogenicity island are also known to participate in epithelial cell activation by H. pylori. H. pylori cag⁺ strains are more potent in activating epithelial cell IL-8 production than cag⁻ bacteria (16, 27). Furthermore, disruption of specific cag region genes markedly reduces H. pylori-mediated tyrosine phosphorylation of gastric epithelial cell proteins, NF-κB activation, and IL-8 gene transcription (12, 14, 19, 21, 22, 28).

A number of recent studies implicate MAP kinases as upstream mediators of NF-κB activation and cytokine gene expression (9, 29, 30, 31, 32, 33, 34). This led us to examine whether MAP kinases participate in regulating H. pylori-induced IL-8 production by gastric epithelial cells. The aims of this study were to determine whether H. pylori activates MAP kinases in AGS gastric epithelial cells, whether MAP kinase activation is required for H. pylori-mediated NF-κB activation and IL-8 production, and whether cag⁺ and cag⁻ strains of H. pylori differ in their ability to activate epithelial cell MAP kinases.

AGS gastric epithelial cells (American Type Culture Collection, Manassas, VA) were grown in Ham’s F-12 medium (pH 7.4; Sigma, St. Louis, MO) supplemented with 10% FBS, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate (15). MKN-28 cells (provided by R. Peek, Jr., Nashville TN) were grown in DMEM with 10% FBS, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate. All cultures were maintained at 37°C in a humidified atmosphere of 95% air and 5% CO₂. Cell culture experiments were conducted in 6-, 12-, or 96-well polypropylene tissue culture plates (Corning Costar, Cambridge, MA).

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 (15, 35). After 3–7 days the bacteria were harvested into pyrogen-free Dulbecco’s PBS (Cellgro, Mediatech, Herndon,VA). The bacteria were pelleted by centrifugation at 4000 × g for 10 min, and bacterial numbers were determined by resuspension in PBS to an OD_{600 nm} of 1.5, corresponding to 3.6 × 10⁸ CFU/ml as described previously (15). Defined numbers of bacteria were then resuspended in antibiotic-free Ham’s F-12 medium. Unless otherwise stated experiments were performed using the cagA⁺ and vacA⁺ H. pylori strain 43504 (American Type Culture Collection) (15).

H. pylori clinical strains were isolated from gastric mucosal biopsies obtained during upper gastrointestinal endoscopy as previously described (16). The presence of cagA was determined by PCR of bacterial genomic DNA and vacuolating cytotoxin activity was determined by neutral red assay (16, 36). All of the cag⁺ clinical isolates used in this study were toxigenic, while the cag⁻ strains did not produce a functional cytotoxin. Isogenic H. pylori mutants lacking the picB or cagA genes were also studied together with their parental cag⁺, toxigenic, wild-type strain (no. 60190) (28). The H. pylori clinical isolates, strain 60190, and the picB⁻ and cagA⁻ mutants were obtained from the culture collection of the Vanderbilt University Campylobacter and Helicobacter Laboratory (Nashville, TN) and have been described previously (16, 28).

H. pylori filtrate was prepared by suspending the bacteria in antibiotic-free medium for 30 min at room temperature, pelleting the bacteria at 4000 × g for 10 min and then filtering the medium through a 0.2-μm pore size filter (Acrodisc, Gelman, Ann Arbor, MI). In some experiments H. pylori were heat treated for 30 min by boiling or were treated with 2 or 20 μg/ml chloroamphenicol (Amersham, Arlington Heights, IL) for 60 min at 37°C.

AGS cells were grown to confluence on twelve-well plates and maintained in serum-free medium for 24 h before the experiment. Cells were fed with fresh serum-free medium 2 h before stimulation. At the end of the experiment the monolayers were washed three times with PBS and lysed with SDS buffer (containing 62.5 mM Tris-HCl (pH 6.8), 2% (w/v) SDS, 10% glycerol, 50 mM DTT, 0.1% (w/v) bromophenol blue). Samples were then sonicated, heated to 100°C for 5 min, and loaded onto a 10% SDS-PAGE gel. After running the gel, the proteins were transferred onto nitrocellulose membranes (Bio-Rad, Hercules, CA). The membranes were blocked for 3 h at room temperature with a 5% (w/v) solution of dried milk in Tris-buffered saline (pH 7.4). This was followed by an overnight incubation at 4°C with the phospho-specific MAP kinase Abs diluted 1/1000 in blocking buffer. The membranes were then washed three times with Tris-buffered saline and incubated at room temperature for 1 h with peroxidase-conjugated goat anti-rabbit IgG (1/2000 dilution; Jackson ImmunoResearch Laboratories, West Grove, PA). A SuperSignal chemiluminescent substrate (Pierce, Rockford, IL) was used for detection.

Phospho-specific p44/p42 MAP kinase Ab was used to detect ERK1/2. This Ab detects p44 and p42 MAP kinase (ERK1 and ERK2) only when they are catalytically activated by phosphorylation at Thr²⁰² and Tyr²⁰⁴. Phospho-specific p38 MAP kinase Ab was used to detect p38 activated by phosphorylation at Thr¹⁸⁰ and Tyr¹⁸². Phospho-specific p54/p46 MAP kinase Ab was used to detect JNK. This Ab detects p54 and p46 MAP kinase only when they are activated by phosphorylation at Thr¹⁸³ and Tyr¹⁸⁵. All three Abs were obtained from New England Biolabs (Beverly, MA). Nonphosphospecific Abs to ERK, p38, and JNK were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Controls for these Western blot experiments consisted of AGS cells harvested at the zero time point for kinetic experiments or cultured in the absence of H. pylori and harvested at the same time as the test cells (usually after 1 h of incubation) for all other experiments.

Confluent monolayers of AGS cells were lysed in 1 ml of lysis buffer (150 mmol NaCl, 1% Nonidet P-40, 0.8 mmol MgCl₂, 5 mmol EGTA, 1 mmol Na₃VO₄, 15 μg/ml leupeptin, 1 mmol PMSF, and 50 mmol HEPES, pH 7.5). ERK2 (p42), p38, or JNK1 (p46) MAP kinases were precipitated with 2 μg of rabbit specific IgG (Santa Cruz Biotechnology). After incubation for 2 h at 4°C, 20 μl of protein G-Sepharose (Santa Cruz Biotechnology) were added, and samples were further incubated for 1 h. Immunopellets were washed twice in lysis buffer, twice in kinase buffer (30 mM NaCl, 0.1 mM Na₃VO₄, 2 mM DTT, 20 mM MgCl₂, and 30 mmol HEPES, pH 7.5.) and then resuspended in 40 μl of kinase buffer. The kinase reaction was started by addition of 20 μM ATP, 100 μCi/ml [γ-³²P]ATP (DuPont-NEN, Boston, MA), and 10 μg myelin basic protein (Sigma) as substrate for ERK and p38, and 2 μg of GST c-Jun_1–79 (Stratagene, La Jolla, CA) as substrate for JNK1. Samples were subjected to SDS-PAGE (12%) and analyzed by autoradiography.

In some experiments AGS cells were treated with the specific p38 MAP kinase inhibitor SB203580 (10 μM; Calbiochem, La Jolla, CA) for 30 min before exposure to H. pylori and during the incubation period of the experiment (37). A similar protocol was used for experiments using the MAP kinase inhibitor PD98059 (25 μM; Calbiochem). PD98059 blocks ERK1/2 activation by specifically inhibiting MEK1, the kinase that catalyzes ERK1/2 phosphorylation (38).

IL-8 protein levels in AGS cell-conditioned medium were measured by ELISA as previously described (15, 35, 39). Statistical analyses were performed using SigmaStat 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.

AGS cells were pretreated with the specific MAP kinase inhibitors PD98059 (25 μM) and/or SB203580 (10 μM) for 30 min preceding infection with H. pylori (10⁸ bacteria/ml). Incubation of AGS cells with H. pylori was maintained for 1 h in the presence of the inhibitors, after which time the cells were washed three times with PBS and lysed. Equal amounts of protein were loaded onto a 10% SDS-PAGE, the gel was run, and the proteins were then transferred to nitrocellulose membranes. The blots were blocked for 1 h in a 5% (w/v) solution of dried milk in Tris-buffered saline solution containing 0.1% Tween-20, followed by immunoblotting with a rabbit polyclonal Ab against IκBα (Santa Cruz Biotechnology) at a 1/1000 dilution. After washing the blot three times with Tris-buffered saline, a peroxidase-conjugated goat anti-rabbit Ab (Jackson ImmunoResearch Laboratories) at a 1/2000 dilution was applied. Immunoreactive bands of the 36-kDa IκBα protein were detected using SuperSignal chemiluminescent substrate.

AGS cells were pretreated for 30 min with the MAP kinase inhibitors PD98059 (25 μM) and/or SB203580 (10 μM). The AGS monolayers were then infected with H. pylori for 1 h. To prepare nuclear extracts AGS cells were washed three times with ice-cold PBS and scraped in ice-cold TNE buffer (40 mmol/L Tris-HCl (pH 7.5), 0.15 mol/L NaCl, and 1 mmol/L EDTA). Cells were pelleted by centrifugation and then resuspended in 400 μL of buffer A (10 mmol/L HEPES (pH 7.9), 10 mmol/L KCl, 0.1 mmol/L EDTA, 0.1 mmol/L ethylene glycol-bis(B-aminoethyl ether)-N,N,N′,N′-tetra-acetic acid, and 0.5 mmol/L PMSF), and incubated on ice for 10 min. Nonidet P-40 (23 μl of a 10% solution) was added, and the cells were incubated for an additional 2 min on ice. After centrifugation at 12,000 × g for 5 s, the nuclear pellet was resuspended in 85 μl of buffer B (20 mmol/L HEPES (pH 7.9), 0.42 mol/L KCl, 1 mmol/L EDTA, 1 mmol/L ethylene glycol-bis(B-aminoethyl ether)-N,N,N′,N′-tetra-acetic acid, and 0.1 mmol/L PMSF) and incubated on ice for 30 min. After a further centrifugation at 12,000 × g, for 2 min, the supernatants were recovered as nuclear extracts that were immediately frozen on dry ice and stored at −80°C.

Single-stranded complementary oligonucleotides bearing the human IL-8 gene NF-κB site were prepared by custom oligonucleotide synthesis (Genosys Biotechnologies, The Woodlands, TX). After annealing, 100 ng of the double-stranded oligonucleotide was labeled in a Klenow fill-in reaction in the presence of [α-³²P]dCTP. The probe was then purified on a Sephadex G-25 spin column (Boehringer Mannheim, Indianapolis, IN) and diluted to yield approximately 15,000 cpm and 0.1 ng of DNA/μL. Binding reactions (20 μl) contained 0.1 ng (∼15,000 cpm) of double-stranded probe, 5–12 μg of extracted protein (the protein concentration of the samples were adjusted to ensure equal loading), 2 μg of poly(dI-dC) (Pharmacia, Piscataway, NJ), 10 mmol/L 2-ME, and 1% Ficoll. After first incubating the protein extracts for 10 min at room temperature, the radiolabeled probe was added. After an additional 30 min at room temperature, the reaction mixtures were then loaded on a nondenaturing 6% polyacrylamide gel in 0.2 mol/L glycine, 25 mmol/L Tris-HCl, and 1 mmol/L EDTA. The gel was run, dried, and exposed to autoradiography film for 6–18 h at −80°C with an intensifying screen.

Supershift assays with anti-NF-κB p50 (sc-114), and anti-NF-κB p65 (sc-109; both from Santa Cruz Biotechnology) were performed to confirm the identity of the complexes binding the IL-8 promoter κB probe. Each Ab (2 μg) was added to the DNA probe at the start of the 30-min incubation.

Confluent monolayers of AGS gastric epithelial cells were infected with H. pylori strain 43504 at bacterial densities ranging from 10⁶-10⁸/ml. After 1 h cell lysates were prepared and examined by Western blot analysis using phospho-specific Abs to ERK1/2, p38, and JNK isoforms p54 and p46. Control cells showed low or undetectable levels of activated MAP kinases (Fig. 1 A). Contact with H. pylori resulted in a marked and dose-dependent increase in the phosphorylation of all three MAP kinases. In the case of ERK, two immunoreactive bands are evident representing phosphorylated p44 (ERK1, upper band) and p42 (ERK2, lower band) MAP kinases. For JNK two immunoreactive bands are again evident representing phosphorylated p54 (upper band) and p46 (lower band) MAP kinases.

Activation of AGS cell MAP kinases by H. pylori was confirmed using in vitro kinase assays. As shown in Fig. 1 B the kinase activities of ERK2, p38, and JNK1 in cell extracts were each markedly increased within 30 min of H. pylori inoculation of AGS monolayers.

We and others have reported that contact with intact bacteria is required for H. pylori-mediated activation of the transcription factor NF-κB in gastric epithelial cells (10, 15, 18, 20, 21). We next examined whether contact with H. pylori was also required for epithelial cell MAP kinase activation. As shown in Fig. 2 H. pylori that were killed by heat treatment for 30 min at 100°C were no longer capable of MAP kinase induction. Soluble factors contained in a cell-free H. pylori filtrate also failed to activate epithelial cell MAP kinases. Pretreatment of H. pylori with chloramphenicol, a bacteriostatic agent that inhibits bacterial protein synthesis, had no apparent inhibitory effect on MAP kinase phosphorylation. These data demonstrate that MAP kinase activation requires live, intact bacteria and does not result from soluble factors present in bacterial culture filtrate. However, de novo bacterial protein synthesis is not required.

Fig. 3 A illustrates the time course of MAP kinase activation following gastric epithelial cell contact with H. pylori. Phosphorylation of ERK1/2 and p38 are clearly evident within 5 min of H. pylori inoculation. Both JNK isoforms are activated within 30 min. Phosphorylation of ERK1/2, p38, and JNK appears to be maximal at 30–60 min.

H. pylori strains that carry the cag pathogenicity island (cag⁺) show greater potency in activating gastric epithelial cell inflammatory responses (12, 14, 16, 18, 21, 22, 27, 28, 40). Infection with cag⁺ strains may also be associated with a greater risk for symptomatic gastroduodenal disease (6, 12, 23, 24, 25, 26). This led us to determine whether cag⁺ and cag⁻ strains of H. pylori differ in their ability to activate epithelial cell MAP kinases. As shown in Fig. 3,A incubation of AGS cells with the cag⁺ H. pylori strain 43504 resulted in marked activation of ERK1/2, p38, and JNK. Incubation with the same concentration of a cag⁻ H. pylori (strain J68) again resulted in the rapid activation of ERK1/2 (Fig. 3 B). However, the intensity of ERK1/2 phosphorylation was less than that observed with the cag⁺ strain. The cag⁻ isolate also induced far less phosphorylation of p38 and very little phosphorylation of JNK compared with the cag⁺ strain.

The differences in MAP kinase activation induced by strains 43504 and J68 led us to study a panel of cag⁺ and cag⁻ H. pylori clinical isolates. As illustrated in Fig. 4 all three of the cag⁺ H. pylori isolates induced a marked activation of the MAP kinases ERK1/2, p38 and JNK. Each of the three cag⁻ H. pylori isolates was less potent in inducing MAP kinase activation. Furthermore, a consistent pattern of differential MAP kinase phosphorylation was observed with the three cag⁻ strains; ERK1/2 phosphorylation was somewhat less intense than with the cag⁺ strains, p38 phosphorylation was substantially less, and there was minimal phosphorylation of JNK.

Having found distinct patterns of MAP kinase activation by cag⁺ and cag⁻ H. pylori clinical isolates, we next examined the participation of specific cag pathogenicity island gene products in this process. The picB region of the cag pathogenicity island is known to be involved in gastric epithelial cell NF-κB activation and IL-8 up-regulation (12, 14, 22, 28). As shown in Fig. 5 the cag⁺ reference strain 60190 (wild type) induced the phosphorylation of ERK1/2, p38, and JNK in AGS gastric epithelial cells as observed with other cag⁺ H. pylori. The cagA⁻ isogenic mutant was similar to the wild type in terms of its ability to activate epithelial cell MAP kinases apart from a possible reduction in ERK1/2 phosphorylation. However, the picB⁻ isogenic mutant was less potent in inducing epithelial cell MAP kinase activation. Compared with the parental wild-type strain, ERK1/2 phosphorylation was only slightly reduced, p38 phosphorylation was moderately reduced, and JNK phosphorylation was markedly reduced. Thus, disruption of picB in this cag⁺ strain resulted in a pattern of MAP kinase activation similar to that seen with cag⁻ strains (see Fig. 4).

H. pylori infection is known to stimulate gastric epithelial cell IL-8 production, and MAP kinases have been reported to regulate the upstream signaling events that control cytokine transcription. We asked, therefore, whether MAP kinase activation might be involved in H. pylori-induced IL-8 production by gastric epithelial cells. In these experiments we used the specific MAP kinase inhibitors SB203580, which blocks p38 kinase activity, and PD98059, which blocks MEK1 kinase activity, thereby preventing ERK1/2 phosphorylation. In control experiments, SB203580 (10 μM) inhibited p38 MAP kinase activity in H. pylori-infected (strain 43504; 10⁸ bacteria/ml) AGS cells as demonstrated using the in vitro p38 kinase assay (data not shown). Similarly, the ability of PD98059 (25 μM) to block MEK1 kinase activity in H. pylori-infected AGS cells was confirmed by Western blotting using the phosphospecific ERK1/2 Ab (data not shown).

AGS cells were pretreated for 30 min with the MAP kinase inhibitors used singly or in combination. The gastric epithelial cell monolayers were then inoculated with H. pylori (strain 43504; 10⁸ bacteria/ml). IL-8 levels were measured in the conditioned medium harvested 7 h later.

H. pylori induced a 36-fold increase in AGS cell IL-8 production (Fig. 6). The p38 inhibitor SB303580 reduced epithelial cell IL-8 production by 82% (p < 0.001), the MEK1 inhibitor PD98059 reduced IL-8 production by 78% (p < 0.001), and a combination of the two inhibitors completely abrogated H. pylori stimulation of IL-8 production (42% of control IL-8 level; p < 0.0001).

Using the MKN-28 gastric epithelial cell line, the p38 inhibitor reduced H. pylori-stimulated IL-8 production by 39% (from 251 ± 7 to 152 ± 5 pg/ml, mean ± SD; n = 4; p < 0.001), the MEK1 inhibitor by 60% (from 251 ± 7 to 100 ± 14 pg/ml; n = 4; p < 0.001), and a combination of the two inhibitors by 100% (from 251 ± 7 to 0 ± 8 pg/ml; n = 4; p < 0.001).

Previous studies have demonstrated that cag⁺ H. pylori activate gastric epithelial cell NF-κB, leading to up-regulation of IL-8 mRNA transcription and protein production (12, 14, 15, 22, 28). We now find that cag⁺ bacteria also activate MAP kinases, leading to an increase in IL-8 production by gastric epithelial cell lines. In view of these parallel findings we next asked whether MAP kinase activation was part of the upstream signaling pathway leading to IκB kinase activation, IκB degradation, and the subsequent activation and nuclear translocation of NF-κB.

As shown in Fig. 7 A detectable levels of IκBα were evident in control AGS gastric epithelial cells by Western blot analysis. Exposure of the gastric epithelial cell to H. pylori resulted in a marked reduction in IκBα levels consistent with IκBα kinase activation leading to IκBα phosphorylation and degradation. This marked reduction in IκBα levels was not prevented by pretreatment of epithelial cells with the MAP kinase inhibitors PD98059 and/or SB203580.

Fig. 7 B shows an EMSA of nuclear extracts from AGS gastric epithelial cells examining NF-κB binding to an oligonucleotide probe containing the IL-8 promoter κB binding sequence. Nuclear extracts from control cells contained little activated NF-κB (lane 1). NF-κB binding to the IL-8 promoter site was markedly increased following H. pylori infection (lane 2). Activation and nuclear translocation of NF-κB did not appear to be influenced by the MAP kinase inhibitors PD98059 and/or SB203580 (lanes 3–5). EMSA supershift studies using Abs to p50 and p65 κB proteins confirmed that p65-containing NF-κB is the predominant form binding to the IL-8 promoter site.

We have shown that contact with H. pylori rapidly activates ERK1, ERK2, p38, JNK p46, and JNK p54 MAP kinases in AGS gastric epithelial cells. We also find that H. pylori cag⁺ strains are more potent than cag⁻ strains in inducing MAP kinase activation. Differential MAP kinase activation by cag⁺ and cag⁻H. pylori strains is particularly evident for JNK phosphorylation and appears to be dependent upon genes within the cag pathogenicity island. Inhibitors of p38 and MEK1 MAP kinases prevent H. pylori-induced IL-8 production. However, p38 and MEK1 MAP kinase activity do not appear to be essential for H. pylori-induced NF-κB activation.

H. pylori has adapted to interact specifically with gastric-type epithelial cells. H. pylori infection is limited to areas of the gastrointestinal tract that are lined by gastric epithelium, and the bacterium is known to activate several gastric epithelial cell signaling events. Previous studies have shown that adherence of H. pylori to AGS gastric epithelial cells induces tyrosine phosphorylation of host proteins, cytoskeletal reorganization, NF-κB activation, and up-regulation of expression of a variety of inflammatory response genes including IL-8 (10, 13, 15, 21, 41). We now report that H. pylori also induces the phosphorylation of ERK, JNK, and p38 MAP kinase family members. Threonine and tyrosine phosphorylation of MAP kinases was evident within 5 min of H. pylori inoculation. Contact with intact bacteria appears to be required for MAP kinase activation. A similar requirement for bacterial contact or adherence was described previously for gastric epithelial cell NF-κB activation by H. pylori (15). Our experiments using the bacterial protein synthesis inhibitor chloramphenicol indicate that MAP kinase activation does not require de novo bacterial protein production. Instead, epithelial cell activation appears to result from contact with preformed bacterial factors, which is consistent with the observed rapid onset of MAP kinase activation.

We did not examine specifically whether H. pylori LPS activates MAP kinases in AGS cells. However, in a previous study we were unable to demonstrate AGS cell NF-κB activation (or IL-8 production) in response to LPS (15). Furthermore, soluble factors present in bacterial culture filtrate did not activate AGS cell MAP kinases. These findings suggest that H. pylori LPS is not responsible for MAP kinase activation in gastric epithelial cells. This is in contrast to human monocytic cells, which are activated by H. pylori culture filtrate and by purified H. pylori LPS as shown by other investigators and by us (35, 42, 43, 44).

We find that H. pylori cag⁺ strains are more potent than cag⁻ strains in inducing gastric epithelial cell MAP kinase phosphorylation. A consistent pattern of differential MAP kinase activation was observed with the panel of cag⁺ and cag⁻ isolates examined in this study. The most striking differences between the cag⁺ and cag⁻ strains were seen for JNK phosphorylation, which was minimal after exposure to the cag⁻ bacteria. p38 MAP kinase activation was also far less evident with the cag⁻ strains. In previous studies, disruption of the picB region of the cag pathogenicity island resulted in a marked reduction in the ability of H. pylori to activate epithelial cell NF-κB and up-regulate IL-8 protein production (22, 28). We now find that a picB⁻ mutant also shows reduced potency in MAP kinase activation compared with its cag⁺ parental strain. Again, an almost complete loss of JNK activation and a markedly reduced potency of p38 activation were evident. Thus, the pattern of MAP kinase activation induced by the picB⁻ mutant recapitulates the pattern observed using the cag⁻ clinical isolates. These findings indicate that gene products of the cag pathogenicity island are necessary for the observed differences in MAP kinase induction by H. pylori cag⁺ and cag⁻ strains.

H. pylori vacuolating cytotoxin has been reported to inhibit EGF-mediated signal transduction and ERK2 activation in Kato III gastric epithelial cells (45, 46). Broth culture supernatants from a vacA⁺ H. pylori strain, but not its isogenic vacA⁻ mutant, inhibited epithelial cell EGF receptor activation, and ERK2 phosphorylation in response to EGF. We now find that H. pylori bacteria, but not culture filtrate, activate ERK1 and ERK2 in gastric epithelial cells. Furthermore, ERK activation is more pronounced following exposure to cag⁺ strains compared with cag⁻ strains that secrete less vacuolating cytotoxin. Thus, vacuolating cytotoxin produced by cag⁺ bacterial strains in our experimental system is unable to prevent H. pylori-induced ERK activation. The interplay between bacterial activation of gastric epithelial cell MAP kinases and the blockade of related signaling pathways by vacuolating cytotoxin requires further investigation.

MAP kinase signaling regulates the expression of many proinflammatory cytokines, including IL-8. This led to our experiments using the specific MAP kinase inhibitors SB203580, which blocks p38 kinase activity, and PD98059, which blocks MEK1 activity, thereby preventing ERK1/2 phosphorylation. Unfortunately, a suitable JNK inhibitor was not available for use in our studies. Both SB203580 and PD98059 substantially reduced IL-8 production by H. pylori-stimulated AGS cells. Thus, activation of both the p38 and MEK1-ERK1/2 pathways are required for a maximal gastric epithelial cell IL-8 response to H. pylori. When both pathways were blocked by combining the two inhibitors, H. pylori-induced IL-8 production was abolished.

Cross-talk between the MAP kinase and NF-κB pathways has been demonstrated in a number of recent studies. For example, the MAP kinase family members MAP kinase kinase kinase (MEKK1) and NF-κB-inducing kinase (NIK) can each directly activate the IκB kinase signalsome, resulting in IκB phosphorylation and release of activated NF-κB (29, 30, 31, 47, 48, 49). Thus, we examined whether MAP kinase activation by H. pylori was an upstream signaling event in the pathway leading to NF-κB activation and, hence, IL-8 gene transcription. Although the MAP kinase inhibitors SB203580 and PD98059 were both effective in blocking IL-8 production, neither had any apparent effect on H. pylori-induced IκBα degradation or on NF-κB activation and binding to the IL-8 promoter κB site. These data suggest that p38 and MEK1 MAP kinases are not required for H. pylori-mediated IκBα degradation or NF-κB activation. Thus, the MAP kinase and NF-κB pathways may exert independent regulatory effects on gastric epithelial cell IL-8 production following H. pylori infection. In other systems, p38 MAP kinases and NF-κB have been shown to regulate cytokine gene expression by independent pathways (32, 50, 51). However, the mechanism by which p38 regulates cytokine production without altering NF-κB activation and DNA binding is not known.

Bacterial adherence to or at least contact with the gastric epithelial cell appears to be necessary for H. pylori-induced MAP kinase activation. However, the adhesins or other bacterial factors, host receptors, and intermediary host signaling molecules that are engaged to activate MAP kinase family members are unknown. Further studies are needed to elucidation these upstream events and define the specific bacterial and host factors that interact to activate the epithelial cell signaling pathways. The potential downstream effects of epithelial cell MAP kinase activation are multiple and varied. In this study we examined MAP kinase-mediated activation of IL-8 production. MAP kinases regulate cell proliferation, differentiation, and programmed death, in addition to stress and inflammatory responses. Hence, activation of gastric epithelial cell MAP kinases by H. pylori may be instrumental in inducing gastroduodenal inflammation, ulceration, and neoplasia. The differential activation of MAP kinases by cag⁺ and cag⁻ strains may be an important determinant of strain-specific differences in the host response to H. pylori infection of the gastric mucosa.