Helicobacter pylori Colonization Ameliorates Glucose Homeostasis in Mice through a PPAR γ-Dependent Mechanism

Background There is an inverse secular trend between the incidence of obesity and gastric colonization with Helicobacter pylori, a bacterium that can affect the secretion of gastric hormones that relate to energy homeostasis. H. pylori strains that carry the cag pathogenicity island (PAI) interact more intimately with gastric epithelial cells and trigger more extensive host responses than cag− strains. We hypothesized that gastric colonization with H. pylori strains differing in cag PAI status exert distinct effects on metabolic and inflammatory phenotypes. Methodology/Principal Findings To test this hypothesis, we examined metabolic and inflammatory markers in db/db mice and mice with diet-induced obesity experimentally infected with isogenic forms of H. pylori strain 26695: the cag PAI wild-type and its cag PAI mutant strain 99–305. H. pylori colonization decreased fasting blood glucose levels, increased levels of leptin, improved glucose tolerance, and suppressed weight gain. A response found in both wild-type and mutant H. pylori strain-infected mice included decreased white adipose tissue macrophages (ATM) and increased adipose tissue regulatory T cells (Treg) cells. Gene expression analyses demonstrated upregulation of gastric PPAR γ-responsive genes (i.e., CD36 and FABP4) in H. pylori-infected mice. The loss of PPAR γ in immune and epithelial cells in mice impaired the ability of H. pylori to favorably modulate glucose homeostasis and ATM infiltration during high fat feeding. Conclusions/Significance Gastric infection with some commensal strains of H. pylori ameliorates glucose homeostasis in mice through a PPAR γ-dependent mechanism and modulates macrophage and Treg cell infiltration into the abdominal white adipose tissue.


Introduction
Helicobacter pylori is the dominant member of the gastric microbiota and has persistently colonized the stomach in humans since our early evolution [1]. However, currently in developed countries there has been a sharp decrease in the prevalence of H. pylori gastric colonization [2,3,4]. Colonization with strains bearing the cag (cytotoxin-associated gene) pathogenicity island (cag PAI) is associated with increased risk of distal gastric pathologies such as non-cardia gastric adenocarcinoma, gastric lymphoma and peptic ulceration [5,6]. Conversely, there also is increasing evidence of H. pylori protection against esophageal and cardial pathologies [2,7,8,9], childhood asthma [10,11,12] and childhood allergies [11,13]. The mechanisms underlying this protective effect of H. pylori acting as a commensal bacterium are largely unknown, although for asthma the suppression of T helper 2 responses by a neutrophil-activating protein of H. pylori favorably modulates allergic asthma in mice [14].
The cag PAI encodes a type IV secretion system that mediates interactions between the bacterium and the gastric epithelium [15], including the secretion of the cag effector protein (CagA) and peptidoglycan. H. pylori virulence factors, such as CagA and the vacuolating protein (VacA), mediate the interactions of H. pylori with host cells [16,17], although immune modulation mediated by H. pylori also may utilize other pathways [18,19]. Transgenic expression of CagA in mice leads to gastric hyperplasia by causing aberrant epithelial cell signaling [20,21]. Inside the host cell, phosphorylation of tyrosines within CagA EPIYA repeats [22] induces the hummingbird phenotype [23], whereas CRPIA motifs (conserved repeat responsible for phosphorylation-independent activity) contribute to the epithelial proliferative and pro-inflammatory responses [24]. CagA proteins lacking EPIYA motifs induce the Jak/STAT3 pathway, with effects on c-myc [25]. CagA alters tight junction barrier function in polarized epithelial cells, affecting adhesion and basement membrane integrity [26,27], and exposing the gastric lamina propria to luminal antigens. A newly characterized strain of H. pylori (V225d) contains an atypical but active cag PAI; passage through mice led to loss of host cell interactive phenotypes via a 15-kb deletion within the cag PAI [1]. Thus, the deletion of a significant portion of the cag PAI in H. pylori resulted in suppressed host cell interactive phenotypes. In addition to illustrating the value of comparing effects of cag PAI-positive and -negative strains on cell phenotype, these findings suggest the potential importance of H. pylori cag PAI in regulating immunity and metabolism.
In addition to its interaction with the gastric epithelium, H. pylori also interacts with gastric neuroendocrine cells secreting gastrin, somatostatin, leptin, and ghrelin, and may influence metabolic processes. Specifically, gastric H. pylori colonization decreases plasma levels of ghrelin, a hormone involved in energy homeostasis [28,29] as well as the density of gastric ghrelin-producing cells in obese patients [30]. Plasma ghrelin concentrations increase following H. pylori eradication, suggesting that eradication may contribute to increased appetite and weight gain, and potentially affect body mass index [29]. Ghrelin expression is negatively regulated by leptin, a multifunctional adipokine with cytokine-like features [31]. Leptin is chiefly synthesized by adipocytes, but 5 to 10% is produced in the stomach [32,33,34,35]. Plasma leptin concentrations are not fully dependent on adiposity [35], suggesting a contribution of gastric-derived leptin.
Obesity is increasing in both developed and developing countries [36], and the incidence of type II diabetes has grown concomitantly [37]. This global epidemic coincides with the decreasing prevalence of H. pylori [2,3,38,39,40], suggesting that gastric colonization with this bacterium might contribute to antiobesity and anti-diabetic actions. Obesity is characterized by insulin resistance and low-grade chronic inflammation in white adipose tissue (WAT) with accumulation of macrophages and elevated levels of circulating pro-inflammatory cytokines [41]. Macrophages mediate chronic inflammation in WAT, and are implicated in obesity-induced inflammation and insulin resistance [42,43]. In the gastric mucosa, H. pylori is pro-inflammatory [44,45], but down-modulates the immune response by impairing phagocytosis [46] and enhancing apoptosis of macrophages [47]. By targeting cells involved in the immune response, H. pylori enhances its own persistence in the host [14,48], suggesting possible global roles for H. pylori in the induction of antiinflammatory or regulatory responses. Since cag + H. pylori strains trigger stronger inflammatory responses than do cag 2 strains [1], we hypothesized that cag status could affect energy homeostasis through its neuroendocrine and immunological effects. To test this hypothesis, we examined the effects of gastric H. pylori infection on appetite-controlling hormones and peroxisome proliferator-activated receptor c (PPAR c), a nuclear receptor and transcription factor that acts as an important thermostat for inflammation and metabolism. We used two mouse models of obesity-related inflammation (i.e., leptin receptor deficient db/db mice and mice with diet-induced obesity, DIO). Leptin regulates immune responses by direct effects on immune cells and regulates feeding and the neuroendocrine system by acting on its receptor in the hypothalamus [49]. The db/db mouse model lacks the long isoform of the leptin receptor (ObRb), which associates with the Janus kinase 2 to mediate intracellular signaling. This mutation causes hyperphagia and decreased metabolic rate, predisposition to diabetes, and endocrine dysregulation. The DIO mouse model consists on feeding high-fat diets (40 calories from fat) to induce obesity and white adipose tissue inflammation [50]. By using these mouse models of obesity and diabetes in combination with H. pylori infection, we examined the role of the predominant gastric bacterium in regulating the initiation, progression and outcomes of obesity and metabolic disorders.

Materials and Methods
Diets and mouse model systems BKS.Cg2+Leprdb/+Leprdb/OlaHsd(db/db) mice, which lack the long isoform of the leptin receptor (ObRb), were fed purified diets that represent a modification of the AIN-93G rodent diet in which the nutritional requirements were met or exceeded (Table S1). We utilized 30 genetically obese db/db pre-diabetic mice in the first experiment. We also performed a follow-up experiment using twenty C57BL/6 wild-type mice in a model of diet-induced obesity (DIO), as described [50]. In the second study we only used the cag PAI mutant strain because the results of the first study had revealed more differences with the mutant than the wild-type strain. A third experiment used PPAR c-expressing (wild-type, PPAR c fl/fl; MMTV-Cre-) and mice lacking PPAR c in immune and epithelial cells (PPAR c fl/fl; MMTV2Cre+) in a C57BL/6 background. These mice were fed either control AIN-93G or high-fat diets (Table S1). Mice were either uninfected or challenged with cag PAI+ (wild-type) and isogenic cag PAI-mutant strains of H. pylori and a group of control non-colonized mice (n = 10/group) as described below. All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Virginia Tech and met or exceeded

Bacterial culture conditions
Helicobacter pylori was grown on Columbia blood agar (BD, Sparks, MD) supplemented with 7% lacked horse blood (Lampire biological laboratories, Piperscille, PA) plates. Cultures were incubated at 37uC for 4 days in an anaerobic jar (OXOID, UK) containing a campylobacter system for the generation of microaerophilic conditions composed of 5% oxygen, 10% carbon dioxide and 85% nitrogen.

Experimental H. pylori infection
Eight-week-old mice were challenged with cag PAI+ (wild-type) and isogenic cag PAI-mutant strains of H. pylori and a group of control non-colonized mice (n = 10/group). Specifically, we used strain 98-325 (wild-type, mouse-passed cag PAI+ European 26695) and strain 99-305 which has a knockout of the entire cag PAI made by transformation of 98-325 with a PCR product from a strain in which the chloramphenicol resistance cassette replaced the entire cag PAI [51]. The absence of cagA in the cag PAI mutant strain 99-305, and its expression in wild-type strain 98-325, as well as the presence of vacA in both strains, was confirmed by PCR (data not shown). H. pylori gastric infection with the strains was carried out by orogastric gavage. Briefly, freshly prepared aliquots (10 10 colony forming units) of the H. pylori strains in sterile Brucella broth were administered to mice three times on days 1, 3, and 5 of the study by oral gavage needle. This dose was used in all experiments. All uninfected mice (n = 10) were inoculated with the same volume of sterile Brucella broth alone. To increase the pH of the stomach and facilitate bacterial colonization, mice were treated with 5% Urea on the drinking water for 7 days starting the day of the first infection.

Assessment of body weight and glucose tolerance
All thirty mice were determined to be normoglycemic (fasting blood glucose levels lower than 250 mg/dl) and to have similar weights (2061.5 g) prior to the experimental challenge with H. pylori. Mice were weighed on a weekly basis and examined for clinical signs of disease by blinded observers. After a standard 12 h fast, glucose was determined on days 0, 7, 14, 35, 50, and 71 of the study for db/db mice and on days 0, 7, 14, 21, and 60 for mice with DIO. Briefly, blood was collected via the lateral tail vein and placed onto capillary blood collection tubes. Mice then were administered a glucose tolerance test by intraperitoneal injection of D-glucose (2 g/kg body weight) and blood samples collected prior to the injection (time 0) (corresponding to a baseline FBG level following a 12-h fast starting at 6 a.m.) and at 15, 60, and 90 minutes (db/db model) or 15, 45 and 90 minutes (DIO model) following the glucose injection.

Necropsy procedures
Mice were euthanized by CO 2 narcosis with secondary thoracotomy on day 71 of the study (db/db model) or day 60 of the study (DIO model). Blood was collected from the heart in heparinized tubes and centrifuged for 10 minutes at 5000 rpm. Plasma was collected and stored at 280uC for ELISA and RIA analyses. The abdominal white adipose tissue (WAT) and White adipose tissue fractionation WAT was fractionated into stromal vascular cells (SVCs) and adipocytes, as described [50,52]. Briefly, WAT was excised, weighed, minced into small (,10 mg) pieces and placed into digestion media consisting of DMEM (Mediatech, Herndon VA) supplemented with 2.5% HEPES (Mediatech) and 10 mg/mL fatty-acid free bovine serum albumin (FAB-poor BSA, Sigma), Liberase Blendzyme 3 (0.03 mg/mL, Roche) and DNase I (50 U/ mL, Qiagen, Valencia CA). Samples were incubated in a rotating 37uC water bath for 90 min, filtered through a 250 mm nylon mesh (Sefar America Inc., Depew NY) to remove undigested particles, and centrifuged at 4uC at 1,000 x g for 10 min. The pellet, consisting of SVCs, containing endothelial cells, pre-adipocytes, macrophages and T cells, was washed with DMEM and centrifuged at 4uC at 1,000 x g for 10 min. The supernatant was discarded and erythrocytes lysed by incubating the SVCs in 2 mL erythrocyte lysis buffer for 2 min before stopping the reaction with 9 mL phosphate-buffered saline (PBS). Cells then were re-centrifuged at 4uC at 1,000 x g for 10 min, suspended in 1 ml of PBS, and enumerated with a Z1 Single Particle Counter (Beckman Coulter, Fullerton CA). The SVCs were resuspended in FACS buffer (PBS, 1% normal goat serum, 0.2% sodium azide) at a concentration of 2610 6 cells/mL.

Quantitative real-time reverse transcriptase PCR
Total RNA was isolated from stomachs using the RNA isolation Minikit (Qiagen), according to the manufacturer's instructions, including a DNAase digestion step. Total RNA (1 mg) was used to generate complementary DNA (cDNA) template using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules CA). The total reaction volume was 20 mL with the reaction incubated as follows in an MJ MiniCycler: 5 min at 25uC, 30 min at 52u, 5 min at 85uC, hold at 4uC. PCR was performed on the cDNA using Taq DNA polymerase (Invitrogen), as described [53]. Each gene amplicon was purified with the MiniElute PCR Purification Kit (Qiagen) and quantitated on an agarose gel by using a DNA mass ladder (Promega). These purified amplicons were used to optimize realtime PCR conditions and to generate standard curves in the realtime PCR assay. Primer concentrations and annealing temperatures were optimized for the iCycleriQ system (Bio-Rad) for each set of primers using the system's gradient protocol. PCR efficiencies were maintained above 92% and correlation coefficients above 0.98 for each primer set (Table S2) during optimization and also during the real-time PCR of sample DNA.
Complementary DNA (cDNA) concentrations for genes of interest were examined by real-time quantitative PCR using an iCycler IQ System and the iQ SYBR green supermix (Bio-Rad). Standard curves were generated using 10-fold dilutions of purified amplicons starting at 5 pg of cDNA and used later to calculate the starting amount of target cDNA in the unknown samples. SYBR green I is a general double-stranded DNA intercalating dye and may therefore detect nonspecific products and primer/dimers in addition to the amplicon of interest. To determine the number of products synthesized during the real-time PCR, melting curve analysis was performed. Real-time PCR was used to measure the starting amount of nucleic acid of each unknown sample of cDNA on the same 96-well plate.  Statistics Data were analyzed by analysis of variance (ANOVA), performed by using the general linear model procedure of SAS [54] as previously described [55,56]. A probability value (P),0.05 was considered to be significant. Experiments 1 and 2 were analyzed as completely randomized studies, and experiment 3 was analyzed as a (26263) factorial arrangement within a completely randomized design. ANOVA was utilized to determine the main effects of the dietary treatment (low-fat vs. high-fat diet), mouse genotype (wild-type vs. PPAR c knockout), or the infection status (uninfected, infected with the wild-type strain or infected with the cag PAI mutant strain) and the 2-way and 3-way interactions between dietary treatment, mouse genotype, and infection status. When the model was significant, ANOVA was followed by Sheffe's post-hoc multiple comparison method and statistically significant (P,0.05) differences among treatment groups are depicted using different superscripts.

Gastric gene expression and plasma levels of leptin
We assayed expression of gastric leptin and circulating levels of leptin, ghrelin, and insulin in the db/db mouse model. The realtime qRT-PCR and ELISA results indicate that gastric leptin mRNA and plasma leptin protein concentrations were higher in the H. pylori-colonized mice, regardless of the cag PAI status, in comparison to the non-colonized group (Table 1). While mice challenged with the cag PAI mutant strain 99-305 had lower plasma levels of ghrelin and higher levels of insulin than the other groups, the differences were not statistically significant. Consistent with the findings in the genetically obese mouse model, in DIO mice gastric infection with the cag PAI mutant strain resulted in a 4-fold increase in gastric leptin mRNA levels compared to noncolonized mice ( Table 2). In addition, gastric infection with the cag PAI mutant strain suppressed gastric IL-6 mRNA expression, thus, indicating lower levels of gastric inflammation in mice colonized with the mutant strain.

Infiltration of macrophages and regulatory T cells into white adipose tissue (WAT)
Adipose tissue macrophages (ATM) are primary contributors to obesity-related inflammation and the derived insulin resistance [41], but their effector function is suppressed by CD4 + CD25 + Foxp3 + regulatory T cells (Treg) [57]. In addition to its role in metabolism, leptin has been recognized as an important immunoregulatory hormone and both T cells and macrophages express leptin receptors. We first isolated the stromal vascular fraction (SVF) of WAT, containing lymphocytes, macrophages, fibroblasts, endothelial cells, and pre-adipocytes, and then examined the macrophage and lymphocyte subsets within this fraction based on surface expression of glycoproteins F4/80 and CD11b. Mice infected with the cag PAI mutant 99-305 had a lower proportion of ATM and greater numbers of Tregs in the SVF compared to uninfected mice, or to mice infected with the cag-positive wild-type H. pylori strain 98-325 (Figure 1). This study provides evidence that gastric cag PAI-negative H. pylori colonization affects adipose tissue inflammatory cell populations.

Effect of H. pylori infection on body weight
Body weights of mice infected with the cag PAI mutant were consistently lower than mice in the other two groups from day 14 before equalizing by day 70, but differences only were significant on day 42 (data not shown). We further examined the effect of the cag PAI mutant strain 99-305 on obesity, by performing a follow-up study using the diet-induced obesity (DIO) mouse model and examining abdominal and subcutaneous fat accumulation. We found that gastric infection with the cag PAI mutant dramatically decreased visceral WAT accumulation in comparison to uninfected control mice (0.23 vs 0.46 grams, P,0.04).

Fasting blood glucose (FBG) and insulin concentrations
We determined the effect of H. pylori on glucose homeostasis by measuring fasting blood glucose concentrations following experimental infection in two models of obesity (Figure 2). In the db/db model, on days 35 and 50, fasting blood glucose levels were significantly higher in the uninfected mice (n = 10/group) compared to the H. pylori-infected mice, regardless of the strain of H. pylori utilized, with the lowest FBG concentrations in mice colonized with the cag PAI mutant strain 99-305, throughout the course of the study (Figure 2A). We confirmed the effect of infection with the cag PAI mutant strain 99-305 using the DIO model ( Figure 2B). Area under the curve (AUC) calculations show that mice infected with the cag PAI mutant strain 99-305 had the lowest glucose levels in both models ( Figure 2C&D). Together, these studies provide evidence of the beneficial effects of colonization with H. pylori on glycemic control, at least in the two mouse models used in this study.

Effect of H. pylori infection and PPAR c expression on Glucose tolerance
To determine whether gastric H. pylori infection modulates how the host initiates glucose homeostasis, we gave an intraperitoneal glucose challenge to experimental animals and evaluated the kinetics of plasma glucose from 0 to 90 minutes following glucose injection (Figure 3). In the db/db model, glucose levels fell toward normal levels more rapidly than in the H. pylori + mice (Panel 3A). In addition, the AUC calculations of GTT indicate that H. pylori + mice had lower levels than uninfected control mice (Panel 3C). In contrast, in the DIO model, levels were relatively similar (Panels 3B&D). The microarray results suggested that H. pylori infection modulated gastric PPAR c pathway expression. To validate this prediction and to characterize potential interactions between H. pylori infection, high-fat feeding and PPAR c expression we conducted a follow up study using wild-type and PPAR c null mice fed low-or high-fat diets. Mice fed regular AIN-93G diets showed no differences between strains or genotypes ( Figure 4A,B). Our data shows that glucose levels in wild-type mice fed high-fat diet, and infected with H. pylori 99-305 were more rapidly normalized than in the uninfected or H. pylori 98-325 groups ( Figure 4C). Of note, our data also demonstrates that the beneficial effects of strain 99-305 on glucose homeostasis were impaired in PPAR c null mice. Specifically, while wild-type mice challenged with strain 99-305 showed improved glucose normalization at 15, 45 and 60 min, PPAR c null mice exhibited a significant change only 15 min after glucose challenge ( Figure 4C and D).

Effect of H. pylori infection and PPAR c expression on WAT macrophage accumulation
To determine whether gastric H. pylori infection modulates the infiltration of immune cells in the WAT, we isolated stromal vascular cells and performed flow cytometric analysis of macrophage and T cell subsets. Our data demonstrates that the infiltration of F4/80+CD11b+ macrophages in WAT of wild-type mice fed high-fat diets infected with H. pylori strain 99-305 was significantly lower than in mice infected with H. pylori 98-325 or uninfected control mice ( Figure 5A). Of note, the beneficial effect of H. pylori 99-305 was reversed in mice lacking PPAR c in immune and epithelial cells ( Figure 5B).
Effect of H. pylori colonization on gastric PPAR cresponsive gene expression in mice PPAR c responsive genes (CD36 and FABP4) in stomachs of mice were assayed by real-time quantitative RT-PCR. Wild type mice on a high-fat diet infected with the H. pylori 99-305 showed an increased expression of CD36 and FABP4, suggesting a higher PPAR c activity when compared the uninfected and H. pylori 98-325 infected groups ( Figure 6).

Discussion
Recognition of the pathogenic role of H. pylori in the development of diseases of the distal stomach, such as gastric cancer and peptic ulceration [5,6] highlights its intimate interaction with host tissues. As such, from an evolutionary view, exploration of H. pylori roles in human gastric endocrine and immune functions is warranted. H. pylori is the dominant member of the gastric microbiota [58,59] and has co-evolved with our ancestors since before humans left Africa [4]. That most mammalian species studied have their own gastric Helicobacter species supports pre-human origins of gastric colonization by this genus. Accordingly, it might be expected that H. pylori plays roles as a commensal or symbiont in the regulation of gastric physiology [60]. Indeed, H. pylori affects gastric acid secretion [61] and modulates gastric acidity [62]. The endocrine function of the stomach also can be modulated by H. pylori; for example, gastric H. pylori colonization down-regulates ghrelin and somatostatin [29,63] secretion, and stimulates that of leptin [64,65] and gastrin [66].
The results of these studies provide in vivo evidence that gastric infection with a cag PAI-negative H. pylori strain, but not with an isogenic cag PAI-positive strain, ameliorates glucose tolerance possibly by activating PPAR c, modulating appetite-controlling hormones and suppressing inflammation. H. pylori-colonized human hosts always harbor strains that lack the cag PAI, since cag-positive infections also involve cag-negative strains [58] and both undergo adaptation processes and differentially interact with the host [67]. For instance, variation in acid susceptibility by CagA status may contribute to the differential colonization of gastric sites. The host-specific equilibrium that has been observed between cag-positive and negative strains may facilitate metabolic and mucosal immune homeostasis by tightly controlling the balance of effector versus regulatory responses.
Obesity-related insulin-resistance is associated with low-grade chronic inflammation [42]. Macrophage-specific gene expression is up-regulated in WAT of db/db mice [42] and inflammatory macrophage subsets infiltrate WAT following high-fat feeding in mice [50]; these contribute to enhanced inflammatory signaling, thereby causing insulin resistance and impairing glucose tolerance [68]. Herein, we provide preliminary evidence that the cag PAI mutant, but not the wild-type H. pylori strain, is associated with enhanced influx of Treg cells into WAT during obesity, which is consistent with a predominance of anti-inflammatory responses. H. pylori-induced gastritis is associated with recruitment of Treg cells into the antral mucosa, suggesting a putative role for this CD4 + T cell subset in both the persistence of H. pylori infection [69], and in modulating systemic inflammation. Infection with a cag-positive strain may involve co-infection with cag-negative strains [70] with potentially opposing effects on the host, which might help neutralize inflammatory damage and disease burden through immunoregulation.
Our results also show that mice infected with either of the H. pylori strains tested had increased gastric-derived leptin without affecting WAT leptin expression. These data are consistent with results of a recent clinical study in humans showing elevated plasma leptin concentrations in H. pylori-colonized subjects [71]. Changes in leptin levels alone could not explain reduced obesity in the present study, as mice infected with the cag PAI mutant strain, but not the wild-type H. pylori strain, had significantly reduced body weights on days 42 and 48 compared to uninfected mice, whereas the effects on leptin were observed in mice infected with either strain. The db/db mice lack the leptin receptor long isoform (ObRb), but express fully functional short isoforms (i.e., ObRa, ObRe, ObRc, ObRd and ObRf) that contribute to its full spectrum of in vivo effects. ObRa is functionally active in the colons of db/db, but not ob/ob mice [72]. Thus, in this study gastricderived leptin could contribute to H. pylori effects in metabolism, through the leptin receptor short isoforms in general, and ObRa in particular. However, the suppressive effects of leptin on feed intake would be abrogated due to the mutation in the ObRb receptor. In contrast, the C57BL/6 wild-type mice with DIO had a 4-fold increase in gastric leptin concentrations. In that model, leptin interacted with all isoforms of its receptor and we also observed improved obesity-related outcomes.
H. pylori infection improved fasting blood glucose levels in both mouse models of obesity and diabetes, although improvements were greater in mice infected with the cag PAI mutant strain. The differences in glucose tolerance observed between the wild-type European and the cag PAI mutant strain may be due to chronic gastric inflammation caused by infection with the cag + strain; through such inflammation the cag PAI-associated tissue responses may contribute to insulin resistance and counteract improvements attributable to changes in gastric-derived hormones. Interestingly, in a follow up study using wild-type and tissue-specific PPAR c null mice, we demonstrated that the loss of PPAR c in immune and epithelial cells impaired the ability of H. pylori infection to ameliorate glucose normalization during a GTT in mice fed highfat diets. Moreover, infection with the cag PAI mutant H. pylori strain resulted in upregulation of gastric PPAR c responsive genes (i.e., CD36 and FABP4), suggesting increased PPAR c activation in vivo. These in vivo findings illustrate the compounded effect of H. pylori colonization on gastric mucosa epithelial and immune cells, all of which express PPAR c, and therefore portray a comprehensive assessment of the effect of H. pylori infection on host response. Interestingly, two recent clinical reports suggest an association between PPAR c and H. pylori-related gastric carcinoma [73,74] thereby providing a molecular basis for the possible role of H. pylori in controlling gastric inflammation, carcinogenesis and metabolism.
Both the bacterial ecology of the stomach [59] and host physiology change in the absence of H. pylori, a bacterium that has co-evolved and co-adapted with humans and that is now disappearing in modern societies that are showing epidemic increases in obesity and diabetes. The present study demonstrates the importance of gastric H. pylori interactions in host control of body weight and glucose tolerance and suggests the importance of PPAR c as a central regulator of host-bacterial interactions. These data suggest that colonization by H. pylori strains lacking the cag PAI could provide partial protection against some metabolic disorders. Thus, if this theory holds true, the disappearance of H. pylori in developed countries may be a contributing factor to the epidemics of obesity and diabetes. Future studies will examine the mechanisms by which specific H. pylori strains modulate regulatory and effector pathways in the gastric mucosa, and their correlation with improvements of chronic inflammatory diseases.