Activation of Peroxisome Proliferator-activated Receptor (cid:1) Suppresses Nuclear Factor (cid:2) B-mediated Apoptosis Induced by Helicobacter pylori in Gastric Epithelial Cells*

Helicobacter pylori colonization leads to epithelial cell hyperproliferation within inflamed mucosa, but levels of apoptosis vary, suggesting that imbalances between rates of cell production and loss may contribute to dif-ferences in gastric cancer risk among infected populations. Peroxisome proliferator-activated receptor (cid:1) (PPAR (cid:1) ) regulates inflammatory and growth responses of intestinal epithelial cells. We determined whether activation of PPAR (cid:1) modified H. pylori -induced apoptosis in gastric epithelial cells. PPAR (cid:1) was expressed and functionally active in gastric epithelial cell lines sensitive to H. pylori -induced apoptosis. PPAR (cid:1) ligands 15d-PGJ 2 and BRL-49653 significantly attenuated H. pyl (cid:3) ri - induced apoptosis, effects that could be reversed by co-treatment with a specific PPAR (cid:1) antagonist. Cyclopentanone prostaglandins that do not bind and activate PPAR (cid:1) had no effects on H. pylori -induced apoptosis. The ability of H. pylori to activate

Chronic gastritis induced by Helicobacter pylori significantly increases the risk for non-cardia gastric cancer (1,2), and host responses that may affect the threshold for carcinogenesis include alteration of epithelial cell proliferation and apoptosis. Mucosal hyperproliferation has been reproducibly demonstrated in H. pylori-infected human (3)(4)(5)(6)(7)(8) and rodent gastric tissue (9 -11), and proliferation scores normalize following successful eradication in humans (3)(4)(5)(6). However, maintenance of tissue integrity requires that enhanced cell production be ac-companied by increased rates of cell loss; consequently, studies have also examined the effect of H. pylori on apoptosis. In contrast to hyperproliferation, H. pylori has been associated with increased (12)(13)(14)(15), unchanged (16), or even decreased (17) levels of apoptosis in vivo, and within a particular population, substantial heterogeneity exists among apoptosis scores (12)(13)(14)(15). These observations suggest that increases in proliferation that are not balanced by concordant increases in apoptosis over years of colonization may heighten the retention of mutated cells, ultimately enhancing the risk for gastric malignancy in certain populations. Differing levels of apoptosis may depend upon bacterial strain-specific factors, because carriage of isolates that possess the disease-associated gene cagA has been associated with enhanced proliferation but attenuated apoptosis in some (8,18) but not all (19), studies. In vitro, however, H. pylori consistently stimulate apoptosis (15, 20 -23), suggesting that mediators within inflamed mucosa modify the direct effect of this organism on epithelial cells and contribute to variability in levels of apoptosis observed in vivo.
One specific host pathway through which inflammatory mediators may influence H. pylori-induced apoptosis is the transcription factor peroxisome proliferator-activated receptor ␥ 1 (PPAR␥). PPAR␥ and the related isoforms PPAR␣ and PPAR␦ constitute a family of nuclear hormone receptors with important roles in the regulation of fatty acid oxidation and glucose utilization (24,25). PPARs form functional heterodimers with the retinoid X receptor family of nuclear receptors (24). It is now appreciated that PPARs are important in regulating pathways beyond energy homeostasis (26). For example, although PPAR␥ was originally identified as a transcription factor essential for adipocyte differentiation (27), there is now increasing evidence to indicate a role for this receptor in regulating other cell types including macrophages, lymphocytes, and epithelial cells. In colonic epithelial cells, activation of PPAR␥ inhibits intracellular signaling cascades, such as NF-B, that regulate inflammation and apoptosis (28,29). Thiazolidinediones (e.g. pioglitazone and rosiglitazone (BRL-49653)) are a family of synthetic compounds with anti-diabetic activity that represent an important class of high affinity, PPAR␥-selective agonists (30). Putative endogenous PPAR␥ ligands include the cyclopentanone prostaglandin 15-deoxy 12,14 -J 2 (15d-PGJ 2 ), which is derived from PGD 2 , a metabolite of cyclooxygenases (COX) (31,32).
Over-expression of COX-2 is a promoting event for colorectal cancer (33), and aberrant expression of COX-2 within H. pyloricolonized mucosa has also been implicated in gastric carcinogenesis. Levels of COX-2 are increased within gastric mucosa of infected but not uninfected persons, suggesting that prostaglandins are important mediators of the host response to H. pylori (34 -37). COX-2 expression is further increased in H. pylori-induced pre-malignant (atrophic gastritis and intestinal metaplasia) and malignant (adenocarcinoma) lesions (38 -40); the chronic use of aspirin or nonsteroidal medications that inhibit COX-2 decreases the risk for distal gastric cancer (41)(42)(43). The molecular mechanism by which COX-2 enhances gastric cancer risk may involve attenuation of apoptosis, since H. pylori-induced apoptosis is augmented in the presence of COX-2 inhibitors in vitro (44) and within gastric mucosa of COX-2-deficient mice (45). Collectively, these findings suggest that prostaglandin products generated by COX-2 may contribute to the heterogeneous levels of apoptosis found within H. pylori-colonized mucosa. Because H. pylori strains invariably induce gastritis and COX-2-generated ligands of PPAR␥ may contribute to apoptosis in vivo, the aims of this study were to determine whether PPAR␥ activation affects the ability of H. pylori to induce apoptosis and to identify the molecular pathways required for these events.
H. pylori Culture-Experiments were performed with the cagA ϩ vacA s1a iceA1 H. pylori reference strain 60190 (ATCC no. 49503) as well as with 8 (6 cagA ϩ , 2 cagA Ϫ ) well characterized clinical strains. Clinical strains were selected from a larger population of isolates that have been described previously as part of an ongoing prospective study designed to study mechanisms of H. pylori pathogenesis (8,46). Because we sought to analyze the importance of H. pylori genes related to carcinogenesis, we selected strains that varied in cagA status. H. pylori were grown in Brucella broth with 5% fetal bovine serum for 48 h, harvested by centrifugation, resuspended to a concentration of 1 ϫ 10 8 colony forming units/ml, and added to gastric cells at a bacteria:cell concentration of 100:1 based on previous reports that H. pylori reproducibly induce apoptosis at this ratio (20,21,23).
Western Analysis-Cell protein extract concentrations were normalized by the Bradford assay (Pierce), and 25 g of protein was separated by SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (PVDF, Pall Corporation, Ann Arbor, MI). Membranes were incubated with affinity-purified anti-PPAR␥ antibody (sc-1984; Santa Cruz Biotechnology, Santa Cruz, CA) or antiphospho-IB antibody (Cell Signaling Technology, Beverly, MA). Primary antibodies were detected using goat anti-mouse (Santa Cruz) or goat anti-rabbit (Cell Signaling) horseradish peroxidase-conjugated secondary antibodies, respectively, and visualized by the ECL detection system (Cell Signaling) according to the manufacturer's instructions.
Assessment of Apoptosis and IL-8 Production-Gastric cells were cultured on chamber slides with or without H. pylori for 24 -48 h, fixed with ice-cold methanol for 10 min, incubated with 1 g/ml propidium iodide for 5 min, and evaluated by fluorescent microscopy. DNA fragmentation was quantified using a commercially available ELISA (Roche Molecular Biochemicals) that detects nucleosomal fragments in cytoplasmic fractions of cells undergoing apoptosis but not necrosis, as described previously (23). For quantitation of IL-8, AGS cells in 6-well plates were co-cultured with or without H. pylori for 6 h in triplicate, and IL-8 protein was measured in supernatants by ELISA (R & D Systems, Minneapolis, MN) (46).
Transfections and Luciferase Assays-AGS cells (5 ϫ 10 5 ) were transfected with a mixture containing 20 g/ml LipofectAMINE (Life Technologies, Inc.), 0.66 g/ml PPRE3-tk-luciferase or 0.66 g/ml pRL-TK (Promega, Madison, WI), and 0.66 g/ml pCDNA3 (Invitrogen, La Jolla, CA) in Opti-MEM (Life Technologies, Inc.) for 5 h. The transfection mixture was then replaced with complete media containing either vehicle (0.1% Me 2 SO or 0.1% acetate) or the indicated ligand. After 24 h, cells were harvested in 1ϫ luciferase lysis buffer. Relative light units from firefly luciferase activity were determined using a luminometer and normalized to the relative light units from Renilla luciferase using the Dual Luciferase kit (Promega).
Immunofluorescence-AGS cells were cultured on collagen-coated glass cover slides, and cells treated with H. pylori in the presence or absence of BRL-49653 were washed twice with ice-cold PBS, fixed in 1% paraformaldehyde in PBS for 10 min at 4 o C, and then permeabilized with methanol for 5 min at Ϫ20°C as described previously (48). Airdried slides were washed with PBS, incubated in 10% non-immune goat serum in PBS (Zymed Laboratories Inc., San Francisco, CA) for 1 h, and then incubated with rabbit anti-NF-B p65 antibody (1:300) (Santa Cruz) in PBS with 1% goat serum overnight at 4°C. Slides were washed three times for 5 min in PBS and incubated with goat anti-rabbit IgG-fluorescein isothiocyanate (1:2000) (Zymed Laboratories Inc.) in 10% goat serum in PBS at room temperature. After washing again, slides were dehydrated and mounted using Vectashield mounting medium (Vector Laboratories, Burlingame, CA). Immunofluorescence was observed using a fluorescence microscope, and for each sample, at least 100 cells were evaluated by an independent observer unaware of experimental conditions. Results are expressed as the mean number of cells with nuclear localization of the NF-B p65 subunit/total number of cells counted.
Statistics-Two-tailed statistical tests were used to evaluate the data. Results are expressed as the mean Ϯ S.D. The Mann-Whitney U test was used for statistical analyses of inter-group comparisons; p values Յ 0.05 were considered significant.

PPAR␥ Is Expressed and Functionally Active in Gastric Epithelial Cell Lines Sensitive to H. pylori-induced Apoptosis-
Reverse transcription-PCR and Western blot analyses were performed to determine whether PPAR␥ is expressed in gastric epithelial cells. PPAR␥-specific amplification products (Fig. 1A) and immunoreactive protein bands (Fig. 1B) were present in both AGS and MKN28 cells. To determine the functional activ-ity of the endogenous PPAR␥ receptor, AGS cells were transfected with a reporter vector (PPRE3-tk-luc) containing three tandem repeats of the PPAR response element (PPRE) from the acyl-CoA oxidase gene upstream of a luciferase cDNA. Treatment of transfected cells with the PPAR␥ selective agonists 15d-PGJ 2 or BRL-49653, but not the related cyclopentanone prostaglandin PGA 1 , resulted in a dose-dependent increase in luciferase activity; the synthetic compound BRL-49653 was more potent in inducing PPRE3-tk-luc expression than 15d-PGJ 2 (Fig. 1C). Thus, PPAR␥ is expressed and transcriptionally responsive to both endogenous and synthetic ligands in gastric epithelial cells.
We tested the hypothesis that H. pylori could induce the phenotypic response of interest (apoptosis) in AGS and MKN28 cells; therefore, we co-cultured cells in the absence or presence of reference strain 60190 and quantified cytoplasmic nucleosomal release. H. pylori significantly increased oligonucleosomal fragments in both cell types at 24 and 48 h (p Ͻ 0.001 for each time point versus controls, Fig. 1D). In addition, the magnitude of H. pylori-induced apoptosis was greatest for AGS cells, and therefore this cell line was used in all subsequent experiments. To independently assess apoptosis, we examined propidium iodide-stained cells using fluorescence microscopy. Similar to DNA fragmentation results, strain 60190 reduced the population of adherent AGS cells and induced chromatin condensation and nuclear segmentation, features consistent with apoptosis (Fig. 1E). These results indicated that we had identified an in vitro system that would allow us to examine the effects of PPAR␥ activation on H. pylori-induced apoptosis.
H. pylori-induced Apoptosis Is Attenuated by Co-treatment with PPAR␥ Ligands-To determine whether activation of PPAR␥ altered H. pylori-induced apoptosis, we pretreated cells with increasing concentrations of either 15d-PGJ 2 , BRL-49653, PGA 1 , or vehicle alone prior to the addition of H. pylori and quantified apoptosis by ELISA. H. pylori alone significantly increased apoptosis (p Ͻ 0.001 versus vehicle-treated cells, Fig.  2A). Treatment with 15d-PGJ 2 or BRL-49653 alone did not significantly increase AGS cell apoptosis (data not shown). However, pre-incubation with either 15d-PGJ 2 or BRL-49653 significantly (p Յ 0.03 for each) inhibited H. pylori-induced apoptosis in a dose-dependent manner, and this reduction was not found following exposure to PGA 1 (Fig. 2A). To more completely confirm the role of PPAR␥ activation, we attempted to reverse these events by using the PPAR␥ antagonist GW9662 (50). Pre-incubation with GW9662 partially reversed the suppressive effects of 15d-PGJ 2 and BRL-49653 (Fig. 2B), providing further evidence that specific activation of PPAR␥ inhibits the ability of H. pylori to stimulate apoptosis in AGS cells.
Attenuation of H. pylori-induced Apoptosis by PPAR␥ Activation Is Not Strain-specific-Apoptosis levels within inflamed tissue have been shown to vary depending upon H. pylori strain characteristics (i.e. cagA) (8,18). To determine whether attenuation of apoptosis by PPAR␥ activation was affected by strain variation, we measured AGS cell apoptosis during co-culture with a panel of cagA ϩ or cagA Ϫ isolates in the absence or presence of BRL-49653. Apoptosis was induced by all strains, although absolute levels varied between different isolates (Fig.  3). Specifically, cagA ϩ strains induced higher levels of nucleo- somal release than cagA Ϫ strains, which is consistent with previous reports (23). However, BRL-49653 decreased H. pylori-induced apoptosis in all samples, albeit to varying degrees, indicating that the ability of PPAR␥ to attenuate apoptosis is likely not related to the presence of cagA in the infecting isolate.
PPAR␥ Agonists Inhibit NF-B Activation by H. pylori-NF-B is a transcription factor for which activation is tightly controlled by inhibitory IB proteins. NF-B can induce (51,52) or inhibit (53-55) apoptosis depending upon the specific cell type, and H. pylori activates NF-B in gastric epithelial cells (56 -58). Because PPAR␥ ligands suppress NF-B activation in intestinal cells (28,29), we determined whether inhibition of H. pylori-induced apoptosis by PPAR␥ activation in gastric cells occurs via a NF-B-dependent mechanism. We first examined the kinetics of NF-B activation by co-culturing strain 60190 with AGS cells and quantitating NF-B nuclear translocation by immunostaining and fluorescence microscopy. H. pylori induces nuclear translocation of p50/p65 NF-B heterodimers (56). Therefore, we used an antibody that recognizes the IB binding region of the p65 subunit. Nuclear translocation of p65 occurred quickly (30 min) following treatment with phorbol 12-myristate 13-acetate (data not shown), indicating that all of the components required for NF-B activation were functional in AGS cells. H. pylori also significantly (p Յ 0.001 for each time point) increased nuclear translocation of NF-B, and the percentage of H. pylori-infected cells with nuclear positivity for p65 peaked at 4 h and then decreased by 6 h (Fig. 4A). These results confirm the findings from previous investigations (56,57) and demonstrate that H. pylori activates NF-B in AGS cells.
We next examined whether activation of PPAR␥ attenuates H. pylori induction of NF-B. Incubation with BRL-49653 alone had no discernible effect compared with control AGS cells, and as predicted, strain 60190 significantly (p Ͻ 0.001) increased p65 nuclear translocation at 4 h (Fig. 4, B and C). In contrast, pre-incubation with BRL-49653 prior to the addition of H. pylori dramatically reduced the number of cells containing nuclear p65 (80 -90% reduction versus H. pylori alone, p Ͻ 0.001, Fig. 4, B and C). These results were confirmed by immunoblot analysis using a phosphorylation state-specific antibody for IB. Immunoreactive bands for phosphorylated IB were observed shortly (15 min) following H. pylori co-culture (Fig. 4D); however, H. pylori-stimulated IB phosphorylation was abrogated by BRL-49653 (Fig. 4D).
To determine whether inhibition of NF-B activation by PPAR␥ agonists has functional consequences, we quantitated the release of a known downstream target of NF-B, IL-8. H. pylori alone potently stimulated IL-8; however, BRL-49653 attenuated this effect (Fig. 4F), results that mirrored its ability to inhibit H. pylori-stimulated NF-B nuclear migration and IB phosphorylation (Fig. 4, B-E). These findings indicate that in AGS cells, activation of PPAR␥ inhibits NF-B signaling through prevention of IB phosphorylation and degradation, which leads to a corresponding reduction of NF-B-dependent inflammatory mediators.
Inhibition of NF-B Suppresses H. pylori-induced Apoptosis-Having established that PPAR␥ activation attenuates induction of NF-B by H. pylori, we next asked whether H. pylori-induced apoptosis was mediated by NF-B. PDTC is a compound that inhibits NF-B activation in AGS cells, thereby preventing NF-B-mediated transcriptional activation (56). BAY 11-7082 is an agent that selectively inhibits tumor necrosis factor-␣-induced phosphorylation of IB␣ (59). We therefore used these inhibitors to demonstrate a link between NF-B activation and induction of apoptosis in H. pylori-infected cells. PDTC or BAY 11-7082 alone had no significant effect on apoptosis (Fig. 5). In contrast, pre-treatment with these compounds markedly inhibited H. pylori-induced apoptosis (Ն95% inhibition by 100 mol/liter PDTC or 10 mol/liter BAY 11-7082; p Ͻ 0.001 versus H. pylori alone, Fig. 5). These findings indicate that H. pylori stimulates apoptosis in AGS cells via NF-B-dependent cascades. Thus, inhibition of NF-B is likely the predominant mechanism through which PPAR␥ activation attenuates H. pylori-induced programmed cell death. DISCUSSION Alteration of epithelial cell proliferation and apoptosis is a manifestation of H. pylori-induced gastritis. However, the precise mechanisms underlying these effects remain incompletely clarified. In isolated cell culture systems, H. pylori directly stimulates apoptosis (15, 20 -23); for example, urease can induce apoptosis by binding to major histocompatibility complex class II molecules expressed on the surfaces of gastric epithelial cells (60). In vivo, however, levels of apoptosis are considerably more variable (12)(13)(14)(15), suggesting that apoptosis within gastric tissue is regulated by host inflammatory mediators.

FIG. 2. H. pylori-induced apoptosis is inhibited by 15d-PGJ 2 and BRL-49653 but not PGA 1 , and PPAR␥ antagonism partially reverses attenuation of H. pylori-induced
The immune response elicited by H. pylori in humans and Helicobacter felis in mice is Th1-predominant (61-63). H. pylori infection of mice deficient in interferon-␥ (IFN-␥), a Th1 lymphocyte-derived cytokine, leads to decreased levels of gas- tric inflammation compared with wild-type mice (64,65), and in vivo neutralization of IFN-␥ in mice infected with H. felis similarly reduces the severity of gastritis (61). H. felis infection of mice deficient in Fas antigen is also associated with reduced levels of inflammation, and mucosal apoptosis scores are decreased in parallel compared with infected wild-type mice (66). A paradigm invoked by these findings is that cytokines (i.e. IFN-␥) within inflamed mucosa may increase apoptosis by regulating Fas-Fas ligand interactions between inflammatory cells and gastric epithelial cells. However, additional host factors also may influence the ability of H. pylori to alter apoptosis in vivo. Mice that possess secretory phospholipase A 2 that are challenged with H. felis demonstrate decreased levels of apoptosis compared with secretory phospholipase A 2 Ϫ/Ϫ strains (67). Apoptosis in response to H. pylori may reflect heterogeneity of class II major histocompatability complex host genotypes, as binding of H. pylori to these IFN-␥-inducible molecules can directly stimulate apoptosis (22,60). Particular polymorphisms of the human interleukin-1␤ gene promoter are genetic risk factors for atrophic gastritis and gastric adenocarcinoma among H. pylori-infected persons (68). Because IL-1␤ can stimulate multiple signaling pathways involved in apoptosis, including those regulated by Fas, NF-B, mitogen-activated protein kinases, and COX-2 (20, 56, 69 -71), differing mucosal levels of IL-1␤ may affect the ability of H. pylori to stimulate apoptosis.
In the current study, we have identified PPAR␥ as an important modifier of H. pylori-induced apoptosis. Activation of PPAR␥ by structurally distinct ligands consistently suppressed the apoptotic cellular response to H. pylori, and one such agonist, 15d-PGJ 2 , is a terminal metabolite of COX-2. COX-2 expression is enhanced within H. pylori-colonized tissue (34 -40), and within gastric mucosa of COX-2-deficient mice infected with H. pylori, apoptosis is increased (45), emphasizing the reciprocal relationship between COX-2 expression and programmed cell death. Our results have provided potential mechanistic insights into these in vivo observations by demonstrating that a COX-2 generated ligand of PPAR␥ attenuates the ability of H. pylori to induce apoptosis in gastric epithelial cells. These events occur via the inhibition of IB phosphorylation, with concomitant suppression of NF-B activation, a precedent that has been established in other models of gastrointestinal inflammation and injury. PPAR␥ ligands potently inhibit NF-B-dependent cytokine secretion in intestinal epithelial cells in vitro (28). Pharmacologic activation of PPAR␥ similarly reduces the levels of inflammation in murine models of colitis (28,72) and intestinal reperfusion injury (29); in the latter report, decreased injury was associated with suppression of NF-B activation. Our current results are concordant with these observations and, for the first time, demonstrate that H. pylori influences apoptosis through activation of NF-B.
PPAR␥ ligands are likely to have effects on cell signaling pathways independent of their ability to bind to and activate PPAR␥. For example, cyclopentanone prostaglandins such as 15d-PGJ 2 and PGA 1 can directly bind to and inhibit IB kinase activity (73). However, this is unlikely to be the mechanism for the effects reported in the current study, as only 15d-PGJ 2 , but not PGA 1 , was capable of inhibiting H. pylori-induced apoptosis. Because the concentration of BRL-49653 required to inhibit H. pylori-induced apoptosis is ϳ10 -20-fold higher than the dose required to induce maximal transcriptional activity in cell-based transactivation assays, an alternative hypothesis is that the effects seen with BRL-49653 are independent of PPAR␥. However, by co-treating BRL-49653 with the PPAR␥ antagonist GW9662, we were able to substantially reverse the decrease in H. pylori-induced apoptosis seen with BRL-49653 treatment alone. Collectively, these results suggest that a majority of the anti-apoptotic response induced by PPAR␥ ligands in H. pylori-infected cells was due to specific modulation of PPAR␥, a finding consistent with a recent report that provided direct genetic evidence of an anti-inflammatory role for PPAR␥ in colonic mucosa (29).
Contact between H. pylori and gastric epithelial cells results in brisk activation of NF-B (56 -58), and this is dependent upon activation of NF-B-inducing kinase via activation of tumor necrosis factor receptor-associated receptors 2 and 6 (58). Activated NF-B-inducing kinase then phosphorylates and activates IKK␣ and IKK␤, which in turn phosphorylate IB␣, leading to its proteosome-mediated degradation, with subsequent release and nuclear translocation of NF-B. Stimulation and activation of NF-B does not require protein synthesis, and therefore this system is particularly utilized in immune, inflammatory, and acute phase responses where rapid activation of defense genes following exposure to pathogens can be critical for the survival of an organism. Activation of NF-B can also regulate cellular growth responses including apoptosis. Although NF-B has been shown to inhibit tumor necrosis factor-␣ stimulated apoptosis (53)(54)(55), expression of Fas ligand and subsequent apoptosis in T lymphocytes is dependent upon activation of NF-B and AP-1 (52). NF-B activation has also been demonstrated to enhance apoptosis in human embryonic kidney cells (51). Consequently, cell lineage characteristics in conjunction with additional intracellular mediators may ultimately dictate the apoptotic cellular response to stimuli that activate NF-B.
The ability of PPAR␥ agonists to suppress H. pylori-induced apoptosis does not correlate with cagA genotype. cagA is the terminal open reading frame of an ϳ40-kilobase locus containing 31 genes (the cag pathogenicity island) (74,75). Several cag island genes possess homology to components of a type IV bacterial secretion system (74,75), which, in other prokaryotic species, functions as a conduit for the export of proteins across the bacterial membrane. CagA is translocated into and phosphorylated within host cells following H. pylori:epithelial cell contact (76 -78). Carriage of cagA ϩ strains significantly enhances the risk for severe gastritis, atrophic gastritis, and distal gastric cancer compared with that incurred by cagA Ϫ strains (46, 79 -83). Although cagA per se is not required for H. pylori induction of COX-2 and prostaglandin release by gastric epithelial cells, these events are dependent upon adjacent genes within the cag island (84). These findings raise the hypothesis that strains containing a functional cag island augment PPAR␥ activation by inducing higher levels of COX-2 generated prostaglandins in vivo, which may be responsible for reduced levels of apoptosis found in association with these strains in certain human populations (8,18). However, other levels of control are likely to be important in regulating these pathways. For example, PPAR␥ ligands can reduce phorbol 12-myristate 13-acetate-dependent transcriptional activation of COX-2 in vitro by suppressing AP-1 activity (85), suggesting that PPAR␥ may regulate its own expression by decreasing eicosanoid production. Activation of PPAR␥ in MKN45 gastric epithelial cells also represses expression of the tyrosine kinase c-Met, which may alter additional signaling pathways that regulate cell growth and death (86).
In conclusion, we have demonstrated that PPAR␥ ligands suppress H. pylori-induced apoptosis in vitro, an effect likely dependent on the ability of PPAR␥ to inhibit H. pylori-mediated increases in NF-B activity. Because PPAR␥ regulates a multitude of host responses such as inflammation, cell growth, and cell death, selective activation of this receptor may not only contribute to varying levels of cellular turnover within inflamed tissue but also to the diverse pathologic outcomes (gastritis alone, peptic ulcer disease, distal gastric cancer) associated with chronic H. pylori colonization.