Abstract

Helicobacter pylori infection is the main risk factor for intestinal metaplasia (IM) and gastric cancer development. IM is a pre-neoplastic lesion, induced by the transcription factor CDX2, where the gastric mucosa is converted to an intestinal phenotype. We previously demonstrated that key elements of the bone morphogenetic protein (BMP) pathway co-localize with CDX2 in IM and upregulate CDX2 expression in gastric cell lines. These observations, together with the hypothesis that CDX2 could be repressed by SOX2, led us to test whether H. pylori , through BMPs, SOX2 and CDX2 could participate in a molecular network critical for the development of IM. AGS cells with and without SMAD4 knock-down were co-cultured with H. pylori or BMP2 to assess the expression of BMP pathway members as well as CDX2 and SOX2 by qPCR and western blot. Proximity ligation assay (PLA) was also performed to evaluate SMAD proteins interaction. Immunohistochemistry and western blot were performed in gastric samples from mice infected with Helicobacter spp. to measure Smad4, pSmad1/5/8, Cdx2 and Sox2 expression in vivo . Increased expression and activity of the BMP pathway accompanied by CDX2 upregulation and SOX2 downregulation were observed in AGS cells co-cultured with H. pylori or BMP2. These effects were impaired by downregulation of the BMP pathway. Finally, infected mice present BMP pathway upregulation, focal Cdx2 expression and decreased Sox2. These results provide a novel link between H. pylori infection and the BMP pathway in the regulation of intestinal and gastric-specific genes that might be relevant for gastric IM.

Introduction

Helicobacter pylori infection is the major risk factor for gastric cancer development. Helicobacter pylori , a gram-negative, spiral-shaped microaerophylic bacterium, triggers a stepwise sequence of alterations of the gastric mucosa starting with superficial gastritis, which can progress to chronic gastritis, atrophic gastritis, intestinal metaplasia (IM), dysplasia and ultimately gastric carcinoma ( 1 ). Helicobacter pylori infection causes severe damage to the gastric mucosa and IM appears in this context as a pre-neoplastic regenerative process consisting of a switch of the gastric mucosa to an intestinal phenotype. IM is the most relevant pre-neoplastic lesion of the stomach affecting about 30% of the individuals infected with H. pylori and confers a significantly increased risk for gastric cancer development. The key molecular mediator of this differentiation switch is the transcription factor CDX2, which is a homeobox protein involved in intestinal differentiation both in normal and in aberrant locations ( 2–5 ). Under normal conditions, CDX2 expression is restricted to the intestine, but it is ectopically expressed in IM lesions, not only of the stomach, but also of the oesophagus and gall bladder, among other locations. In these ectopic settings, it appears in the context of chronic inflammation/regeneration, like in gastric IM. Cdx2 null mice are not viable, whereas Cdx2 + / mice develop polyp-like lesions in the intestine with loss of Cdx2 expression and appearance of gastric differentiation ( 6–8 ). Conversely, forced expression of Cdx2 in the stomach of transgenic mice leads to extensive IM, with subsequent progression to gastric cancer ( 9–11 ). CDX2 regulation in gastric IM has not been fully uncovered but we have demonstrated that key elements of the bone morphogenetic protein (BMP) pathway co-localize with CDX2 in IM and upregulate CDX2 expression in gastric cell lines ( 12 ). Moreover, we and Manzo et al. observed an upregulation of CDX2 expression induced by H. pylori in a co-culture model with gastric cell lines ( 13 , 14) .

BMPs constitute the largest subfamily of the transforming growth factor-beta superfamily of growth factors and exert pleiotropic biological effects, ranging from regulation of early developmental processes to organogenesis ( 15 ). BMPs are extracellular proteins that initiate signalling through binding to specific transmembrane serine/threonine kinase receptors. Upon ligand binding, type II receptor kinases phosphorylate type I receptors, which in turn phosphorylate the intracellular signal transducers, SMAD proteins. Upon phosphorylation, SMAD1, -5 and/or -8 form a complex with SMAD4 and translocate to the nucleus, where they act as transcriptional regulators ( 16–18 ). In the intestine, the BMP signalling pathway, through BMP2 and BMP4, is fundamental for the maintenance of differentiation and architecture of the intestinal epithelium as demonstrated by several mouse models with impairment of the BMP pathway ( 19–22 ). Also in humans, mutations of the BMP pathway generate juvenile polyposis with loss of intestinal differentiation ( 23–25 ).

Different studies have shown that, in addition to the positive regulation, CDX2 expression could be repressed and SOX2 has emerged as a likely candidate for that role. Benahmed et al. ( 26 ) showed that Sox2 negatively regulated the Cdx2 promoter by hampering the action of other transcription factors in an intestinal cell line. Moreover, Asonuma et al . ( 27 ) showed that SOX2 expression is negatively affected by H. pylori and that SOX2 downregulation leads to an upregulation of CDX2 expression in a gastric carcinoma cell line. SOX2 is the sex-determining region Y-box 2 gene, a member of the high mobility group domain proteins and is a crucial transcription factor for the maintenance of cellular pluripotency ( 28 , 29) . On the other hand, Sox2 is expressed and participates in the development of the foregut-derived organs, such as oesophagus and stomach, and is absent from the hindgut-derived intestine ( 30 ). Likewise, in adults, Sox2 was shown to be expressed in the stomach and was absent from the intestine ( 31 , 32) . These studies have launched SOX2 as a putative gastric transcription factor and this was reinforced by the demonstration of its involvement in the regulation of the stomach-specific genes, pepsinogen and Muc5ac ( 33–35 ).

Since H. pylori is the main trigger for the development of gastric IM, we aimed at studying if H. pylori infection affects the expression and activity of the BMP pathway and if both factors modulate CDX2 and SOX2 expression, hypothesizing that H. pylori , through BMPs, SOX2 and CDX2 could participate in a molecular network critical for the development of IM.

Material and methods

Cell culture

Human gastric carcinoma cell line AGS (ATCC) was maintained in RPMI 1640 (Gibco, Invitrogen) supplemented with 10% fetal bovine serum (Gibco, Invitrogen, Carlsbad, CA) and 1% antibiotics (10U/ml penicillin and 10 µg/ml streptomycin; Gibco, Invitrogen, Carlsbad, CA) at 37°C in a humidified 5% CO 2 incubator.

The AGS cells with SMAD4 knock-down (AGS-SMAD4i) and respective scrambled control (AGS-Sc) cell lines were maintained in selective puromycin-containing (5 µg/mL) standard medium ( 12 ).

Co-culture of AGS cell line with Helicobacter pylori

Helicobacter pylori strains 26695, containing the virulence-associated cag pathogenicity island (cagPAI) and Tx30a, lacking the cag pathogenicity island, were grown for 48h in selective medium (Pylori-Gelose, BioMérieux, Marcy l’Ètoile, France) at 37°C under microaerophylic conditions (Genbox microaerophylic, BioMérieux). One day prior to co-cultures, AGS cells were seeded (5 × 10 5 ) in six-well plates in RPMI 1640 medium supplemented with 10% fetal bovine serum without antibiotics. The AGS-SMAD4i and AGS-Sc cell lines were seeded 72h prior to co-cultures in standard medium supplemented with puromycin. One day prior to co-cultures, the medium was replaced by RPMI 1640 supplemented with 10% fetal bovine serum without antibiotics. All co-cultures were performed for 8h at multiplicities of infection (MOI) of 1:100 or 1:200.

BMP pathway activation

For BMP pathway activation experiments, AGS cells were treated for 24h with BMP2 (R&D Systems, Minneapolis, MN) added to the culture medium, at a concentration of 50 or 100ng/mL. Vehicle solution [4mM HCl, 0.1% bovine serum albumin (BSA)] was used as the negative control.

Protein extraction and western blot

Whole-cell extracts were obtained by resuspension of cell pellets in RIPA buffer (50mM Tris–HCl pH 7.4, 150mM NaCl, 2mM EDTA, 1% NP-40, 0.1% sodium dodecyl sulphate) in the presence of complete protease inhibitors cocktail (Roche, Indianopolis, IN). Quantification of total protein was determined by bicinchoninic acid protein assay (Pierce, Rockford, IL). Protein extraction from gastric samples of mice was performed by tissue disruption, in lysis buffer, using a rotor–stator homogenizer (5 pulses of a 20s homogenization intercalated with 5 pulses of resting at room temperature) and was then passed 5 times through a 20-gauge needle attached to a sterile plastic syringe. Protein was precipitated with four volumes of ice-cold acetone and incubated on ice for 30min. Samples were then centrifuged at full speed (4000rpm) and the supernatant was discarded. Pellets were then washed with ice-cold ethanol and left to ‘air-dry’. Pellets were resuspended in a 2D buffer (4% w/v CHAPS, 6M urea, 2M thiourea, 10mM Tris–HCl pH 8.5) and kept at 4°C overnight. Lysates were then centrifuged at 14 000 × g at 4°C for 5min, the supernatant was recovered and total protein was quantified using the 2D Quant Kit (Amersham, GE Healthcare, UK). In this study, 50–60 µg of total protein extracts were subjected to standard sodium dodecyl sulphate–polyacrylamide gel electrophoresis, transferred to a nitrocellulose membrane (Amersham, GE Healthcare, UK) and blotted with primary antibodies overnight at 4°C: mouse monoclonal anti-CDX2 (1:500, Biogenex), mouse monoclonal anti-SOX2 (1:4500, Sigma-Aldrich), mouse monoclonal anti-SMAD4 (sc-7966, 1:1000, Santa Cruz Biotechnology), rabbit polyclonal anti-pSMAD1/5/8 (1:1000, Cell Signalling), rabbit monoclonal anti-e-cadherin (1:1000, Cell Signalling), mouse monoclonal anticytokeratin (AE1/AE3, 1:2000, Zymed) and goat polyclonal anti-β-actin (1:8000, Santa Cruz Biotechnology) in 5% BSA in tris-buffered saline 0.01% Tween-20. Peroxidase-conjugated secondary antibodies were used and developed with the enhanced chemiluminescence detection kit (Amersham, GE Healthcare, UK). Quantification of the western blots was performed using the Quantity One software (BioRad, CA). Each experiment was performed at least twice and a representative result is shown.

RNA extraction and real-time PCR

Total RNA was extracted using TRI Reagent (Sigma, St. Louis, MO) and converted to cDNA using the SuperScript® II Reverse Transcriptase (Invitrogen, Carslbad, CA). CDX2 (5ʹ-TTC ACT ACA GTC GCT ACA TCA CC-3ʹ and 5ʹ- TTG TTG ATT TTC CTC TCC TTT GC- 3ʹ), SOX2 (5ʹ-AAC GGC TCG CCC ACC TAC AGC-3ʹ and 5ʹ-AGT GGG AGG AAG AGG TAA CC-3ʹ), BMP2 (5ʹ-CTC AGG TCA GCC GGG CTC A-3ʹ and 5ʹ-GTT CTT CCA AAG ATT CTT CAT GG-3ʹ) and 18S (5ʹ-CGC GCG CTA GAG GTG AAA TTC-3ʹ and 5ʹ-CAT TCT TGG CAA ATG CTT TCG-3ʹ) were amplified with SYBR Green (Applied Biosystems, Foster City, CA) in a fluorescence reader ABI Prism 7500. The levels of 18S were used for normalization and relative mRNA levels were calculated. Each experiment was carried out in triplicates at least twice. The results are expressed as mean ± SD of representative triplicates.

Proximity ligation assay

Proximity ligation assays (PLA) were performed using the DuoLink® II Fluorescence Kit (Olink® Bioscience, Uppsala, Sweden) according to the manufacturer’s instructions. Briefly, after co-culture with both H. pylori strains at a multiplicity of infection of 100:1 cells were recovered, fixed with methanol for 5min and frozen at −20°C. Cells were then incubated at 37°C for 30min with a blocking solution in a humidity chamber after which the primary antibodies (SMAD4, 1:50 and pSMAD1/5/8, 1:75) were added for overnight incubation at 4°C. In the following day, PLA probes were hybridized for 1h at 37°C, ligation was performed for 30min at 37°C and amplification was carried out for 100min at 37°C to produce rolling circle products. These products were visualized with fluorescently labelled oligonucleotides and the sections counterstained using Duolink II Mounting Medium with 4ʹ,6-diamidino-2-phenylindole. Samples were examined under a Zeiss Imager. Z1 Axio fluorescence microscope equipped with 4ʹ,6-diamidino-2-phenylindole and Texas Red filters. PLA products are seen as fluorescent dots. Images were acquired using a Zeiss Axio cam MRm and the AxioVision Rel. 4.8 software. The resulting images were modified using ImageJ as follows: background with radius two was subtracted from the red channel of the RGB images. The result was intensity scaled to suit printing demands. PLA products were quantified using BlobFinder v3.2 (Olink® Bioscience).

Mice infection with Helicobacter pylori and Helicobacter felis

Helicobacter pylori Sydney strain (SS1) was cultured on H. pylori selective agar, Wilkins–Chalgren agar supplemented with 10% defibrinated horse blood, vancomycin (10mg/l), cefsulodin (2mg/l), trimethoprim lactate (5mg/l) and fungizone (1mg/l; Biogerm, Maia, Portugal), and was incubated at 37°C for 24h under microaerophylic conditions. Helicobacter felis strain was cultured on H. pylori non-selective agar, Wilkins–Chalgren agar supplemented with 10% defibrinated horse blood and 2U of Vitox, and was incubated at 37°C for 48h under microaerophylic conditions. Colonies were tested for urease, catalase and oxidase activities and observed under the microscope by Gram stain. Pathogen-free male C57Bl/6 5-week-old mice were used in compliance with guidelines and a protocol was approved by the Animal Care and Use Committee of the Direcção Geral de Veterinária. Mice were subjected to fasting for 10h and inoculated intragastrically with 0.1ml of H. pylori SS1 or H. felis cell suspension containing 10 8 colony-forming units/ml on 3 consecutive days. Eight weeks after the infection, mice were tested for the presence of the bacteria by 13 C–urea breath test as described in Santos et al. ( 36 ) and also by immunohistochemistry.

Immunohistochemistry

The mice were euthanized by cervical dislocation at 6, 18 and 27 weeks post infection, the stomachs were harvested and dissected along the greater curvature, fixed and embedded in paraffin. Paraffin-embedded samples were serially sectioned at 4 µm, mounted on gelatin-coated slides, dried overnight at 37°C and deparaffinized with clear rite (Thermo Scientific Richard-Allan). Antigen retrieval was performed by boiling the slides in 10mM citric acid at pH 6.0, in a microwave oven for 20min. After cooling, slides were incubated with 3% hydrogen peroxide in methanol for 15min, followed by blocking with the non-immune serum for 30min (Dako) diluted 1:5 in 10% BSA (v/v). Excess normal serum was removed and slides were incubated overnight at 4°C with one of the following primary antibodies: mouse monoclonal anti-SOX2 (0.4 µg/mL, Sigma-Aldrich), mouse monoclonal anti-CDX2 (1:50, Biogenex), rabbit polyclonal anti-pSMAD1/5/8 (1:75, Cell Signalling) and mouse monoclonal anti-SMAD4 (1:50, Santa Cruz Biotechnology) diluted in 5% BSA (v/v). Primary antibody was visualized with biotinylated secondary antibody and an avidin/biotin detection system (Vectastain ABC kit, Vector Laboratories, Burlingame, CA) following the protocol provided by the manufacturer. Slides were developed with diaminobenzidene (Sigma Aldrich) and counterstained with Mayer’s haematoxylin, dehydrated and mounted (Thermo Scientific Richard-Allan).

Results

Helicobacter pylori regulates the BMP pathway, CDX2 and SOX2

We assessed whether H. pylori activates the BMP pathway and thereafter modulates the expression of CDX2 and SOX2, forming a regulatory network that could be involved in the onset of gastric IM. To test that, we co-cultured AGS cells with two H. pylori strains, 26695 and Tx30, at different multiplicities of infection. We studied the expression of BMP2 by real-time PCR and the expression of SMAD4 and the phosphorylated form of SMAD1/5/8 (pSMAD1/5/8), which is generally accepted as the readout of an active BMP pathway, by western blot. A significantly increased expression of BMP2, SMAD4 and pSMAD1/5/8 was observed upon infection of AGS cells with both H. pylori strains ( Figure 1A and 1B ). In addition to characterizing expression, PLA that detects protein interactions was used to see whether interaction of pSMAD/1/5/8 with SMAD4 was increased upon H. pylori infection. Furthermore, subcellular localization of the complexes was also assessed by this method ( 37 ). The results obtained showed increased interaction between these proteins in cells infected with both H. pylori strains, evidenced by the presence of a significantly higher number of dots both in the cytoplasms and in the nucleae ( Figure 2 ). This indicates that the pathway becomes more active upon H. pylori infection.

Fig. 1.

BMP pathway regulation by Helicobacter pylori . (A) Fold increase in BMP2 mRNA expression in AGS cells upon co-culture with H. pylori Tx30a and 26695 strains, for 8h at multiplicities of infection of 100:1 and 200:1. The values obtained with non-infected cells are referred to as 1. BMP2 mRNA levels are normalized with the respective 18S mRNA. (B) Western blot for SMAD4 and pSMAD1/5/8 in the same conditions described above. Non-infected cells were used as control and β-actin was used as a loading control.

Fig. 1.

BMP pathway regulation by Helicobacter pylori . (A) Fold increase in BMP2 mRNA expression in AGS cells upon co-culture with H. pylori Tx30a and 26695 strains, for 8h at multiplicities of infection of 100:1 and 200:1. The values obtained with non-infected cells are referred to as 1. BMP2 mRNA levels are normalized with the respective 18S mRNA. (B) Western blot for SMAD4 and pSMAD1/5/8 in the same conditions described above. Non-infected cells were used as control and β-actin was used as a loading control.

Fig. 2.

Interaction between SMAD4 and pSMAD1/5/8 upon co-culture with Helicobacter pylori . SMAD4 and pSMAD1/5/8 interactions were determined by PLA in non-infected AGS cells (A) and in AGS cells infected with H. pylori 26695 (B) and Tx30a (C) for 3h at a multiplicity of infection of 100:1. (D) Average number of red fluorescent dots in the cytoplasm and in the nucleus of non-infected cells ( N = 265 cells), cells infected with 26695 ( N = 319 cells) and cells infected with Tx30a ( N = 261 cells). *Average number of red dots per nucleus and per cytoplasm was significantly different with P = 0.00001 using Student’s t -test. Original magnifications: (A, B, C) 400×.

Fig. 2.

Interaction between SMAD4 and pSMAD1/5/8 upon co-culture with Helicobacter pylori . SMAD4 and pSMAD1/5/8 interactions were determined by PLA in non-infected AGS cells (A) and in AGS cells infected with H. pylori 26695 (B) and Tx30a (C) for 3h at a multiplicity of infection of 100:1. (D) Average number of red fluorescent dots in the cytoplasm and in the nucleus of non-infected cells ( N = 265 cells), cells infected with 26695 ( N = 319 cells) and cells infected with Tx30a ( N = 261 cells). *Average number of red dots per nucleus and per cytoplasm was significantly different with P = 0.00001 using Student’s t -test. Original magnifications: (A, B, C) 400×.

In the same cells, CDX2 and SOX2 expression was assessed by real-time PCR and western blot showing that CDX2 was upregulated, whereas SOX2 was downregulated ( Figure 3 ).

Fig. 3.

CDX2 and SOX2 regulation by Helicobacter pylori . (A and B) Fold increase in CDX2 and SOX2 mRNA expression in AGS cells upon co-culture with H. pylori Tx30a and 26695 strains, for 8h at multiplicities of infection of 100:1 and 200:1. The values obtained with non-infected cells are referred to as 1. CDX2 and SOX2 mRNA levels are normalized with the respective 18S mRNA. (C) Western blot for CDX2 and SOX2 expression in AGS cells upon co-culture with H. pylori strains as previously at a multiplicity of infection of 100:1. Non-infected cells were used as control and β-actin was used as a loading control.

Fig. 3.

CDX2 and SOX2 regulation by Helicobacter pylori . (A and B) Fold increase in CDX2 and SOX2 mRNA expression in AGS cells upon co-culture with H. pylori Tx30a and 26695 strains, for 8h at multiplicities of infection of 100:1 and 200:1. The values obtained with non-infected cells are referred to as 1. CDX2 and SOX2 mRNA levels are normalized with the respective 18S mRNA. (C) Western blot for CDX2 and SOX2 expression in AGS cells upon co-culture with H. pylori strains as previously at a multiplicity of infection of 100:1. Non-infected cells were used as control and β-actin was used as a loading control.

The BMP pathway regulates CDX2 and SOX2

We assessed next whether the BMP pathway regulated SOX2 expression, using upregulation of CDX2 as an internal control ( 12 ). To test that, we cultured AGS cells, shown previously to have an active BMP pathway, in the presence of BMP2 in the culture medium ( 12 ). We observed that CDX2 and SOX2 expression levels were significantly altered, in an inverse manner, upon addition of BMP2 to the culture medium. CDX2 expression was upregulated, as expected, whereas SOX2 was downregulated ( Figure 4A ). To further confirm the inverse regulation of CDX2 and SOX2 by the BMP pathway, we characterized the expression of these two proteins in AGS cells with stable SMAD4 knock-down using shRNAs (AGS-SMAD4i) and in the respective scrambled control (AGS-Sc; 12 ). In accordance with the previous results, we observed a significant upregulation of SOX2 in these cells, concomitant with a downregulation of both SMAD4 and CDX2 ( Figure 4B ).

Fig. 4.

CDX2 and SOX2 regulation by the BMP pathway. (A) Western blot for CDX2 and SOX2 expression in AGS cells upon activation of the BMP pathway by the addition of BMP2 to the culture medium. Vehicle-treated cells were used as control. (B) Western blot for SMAD4, CDX2 and SOX2 in AGS cells with SMAD4 knock-down (SMAD4i) and respective scrambled control (Sc). β-actin was used as a loading control.

Fig. 4.

CDX2 and SOX2 regulation by the BMP pathway. (A) Western blot for CDX2 and SOX2 expression in AGS cells upon activation of the BMP pathway by the addition of BMP2 to the culture medium. Vehicle-treated cells were used as control. (B) Western blot for SMAD4, CDX2 and SOX2 in AGS cells with SMAD4 knock-down (SMAD4i) and respective scrambled control (Sc). β-actin was used as a loading control.

Role of the BMP pathway in Helicobacter pylori -induced SOX2 and CDX2 expression

To evaluate if H. pylori regulates CDX2 and SOX2 expression through the BMP pathway, we infected AGS-SMAD4i and AGS-Sc with the H. pylori strains used previously. The results obtained showed impairment of CDX2 upregulation by H. pylori in cells knock-down for SMAD4, as opposed to the negative control ( Figure 5 ). On the contrary, SOX2 continues to be downregulated by H. pylori in cells with SMAD4 knock-down ( Figure 5 ).

Fig. 5.

Role of the BMP pathway in Helicobacter pylori -induced SOX2 and CDX2 expression. Western blot for SMAD4, CDX2 and SOX2 in AGS cells with SMAD4 knockdown (SMAD4i) co-cultured with H. pylori strains Tx30a (A) and 26695 (B) for 8h at a multiplicity of infection of 100:1. Scrambled-transfected and non-infected cells were used as controls. β-Actin was used as a loading control.

Fig. 5.

Role of the BMP pathway in Helicobacter pylori -induced SOX2 and CDX2 expression. Western blot for SMAD4, CDX2 and SOX2 in AGS cells with SMAD4 knockdown (SMAD4i) co-cultured with H. pylori strains Tx30a (A) and 26695 (B) for 8h at a multiplicity of infection of 100:1. Scrambled-transfected and non-infected cells were used as controls. β-Actin was used as a loading control.

Mice infected with Helicobacter spp. have CDX2 de novo expression, SOX2 repression and BMP pathway activation

To challenge our hypothesis in vivo , we used a mouse model, C57Bl/6 mice, infected with two Helicobacter species, H. pylori (strain SS1) and H. felis. We analysed, by immunohistochemistry, the expression of CDX2, SOX2, SMAD4 and pSMAD1/5/8 in mice infected with both Helicobacter spp. for 6, 18 and 27 weeks and in non-infected controls. A total of 12 mice infected with Helicobacter spp. were characterized, two infected for 6 weeks, two infected for 18 weeks and two infected for 27 weeks, with each strain. Two non-infected controls were analysed for each time-point. CDX2 de novo expression was detected in 7 out of 12 infected mice and in none of the controls ( Figure 6A ). CDX2 expression was focal and was not accompanied by morphological alterations suggestive of an intestinal phenotype. SOX2 expression was downregulated in all infected mice compared with controls ( Figure 6A ). Finally, expression of SMAD4 and pSMAD1/5/8 was detected by immunohistochemistry in both controls and infected mice, with an increased and nuclear localized expression in the latter ( Figure 6A ).

Fig. 6.

BMP pathway activation and SOX2 and CDX2 expression in mice infected with Helicobacter spp. (A) Immunohistochemistry for Smad4, pSmad1/5/8, Sox2 and Cdx2. a: Smad4 staining in most of the nuclei in the neck region of the gastric antrum mucosa of non-infected mouse; b and c: Smad4 staining both in the foveolar, glandular and neck regions of Helicobacter felis - and Helicobacter pylori -infected mice, respectively. d–i: pSmad1/5/8 staining the nuclei in the foveolar, glandular and neck region of non-infected (d,e), H. felis infected (f,g) and H. pylori infected (h,i) mice. j: Sox2 nuclear staining in foveolar, glandular and neck regions of gastric antrum mucosa of non-infected mouse; k and l: Weak staining for Sox2 in foveolar, glandular and neck region of H. felis- and H. pylori -infected mice, respectively. m: CDX2 staining negative in the normal gastric mucosa; n and o: Focal CDX2 de novo expression in half a gland in the antrum of H. felis infected mouse; p and q: Focal whole gland CDX2 de novo expression in the antrum mucosa of H. pylori infected mouse. Original magnifications: (a,b,c,d,f,h,j,k,l,m,n,p) 100×; (e,g,i,o,q) 400×. (B) Western blot for pSmad1/5/8 in protein extracts of gastric mucosa from mice infected with H. pylori and H. felis at 18 and 27 weeks post infection. E-cadherin was used as loading control of the epithelium compartment. Quantification of pSMAD1/5/8 normalized to their respective e-cadherin is depicted in the lower panel, where the white bars represent the non-infected controls and the grey and black bars correspond to the H. felis- and H. pylori -infected mice, respectively. In this representation, all values were normalized to the first no-infection control.

Fig. 6.

BMP pathway activation and SOX2 and CDX2 expression in mice infected with Helicobacter spp. (A) Immunohistochemistry for Smad4, pSmad1/5/8, Sox2 and Cdx2. a: Smad4 staining in most of the nuclei in the neck region of the gastric antrum mucosa of non-infected mouse; b and c: Smad4 staining both in the foveolar, glandular and neck regions of Helicobacter felis - and Helicobacter pylori -infected mice, respectively. d–i: pSmad1/5/8 staining the nuclei in the foveolar, glandular and neck region of non-infected (d,e), H. felis infected (f,g) and H. pylori infected (h,i) mice. j: Sox2 nuclear staining in foveolar, glandular and neck regions of gastric antrum mucosa of non-infected mouse; k and l: Weak staining for Sox2 in foveolar, glandular and neck region of H. felis- and H. pylori -infected mice, respectively. m: CDX2 staining negative in the normal gastric mucosa; n and o: Focal CDX2 de novo expression in half a gland in the antrum of H. felis infected mouse; p and q: Focal whole gland CDX2 de novo expression in the antrum mucosa of H. pylori infected mouse. Original magnifications: (a,b,c,d,f,h,j,k,l,m,n,p) 100×; (e,g,i,o,q) 400×. (B) Western blot for pSmad1/5/8 in protein extracts of gastric mucosa from mice infected with H. pylori and H. felis at 18 and 27 weeks post infection. E-cadherin was used as loading control of the epithelium compartment. Quantification of pSMAD1/5/8 normalized to their respective e-cadherin is depicted in the lower panel, where the white bars represent the non-infected controls and the grey and black bars correspond to the H. felis- and H. pylori -infected mice, respectively. In this representation, all values were normalized to the first no-infection control.

In order to confirm increased expression and activity of the BMP pathway in mice infected with Helicobacter spp., we performed western blot for pSMAD1/5/8 with proteins extracted from the stomach of four control mice and in two mice infected with each Helicobacter spp. for 18 and 27 weeks. We observed increased pSMAD1/5/8 expression in seven out of the eight infected mice ( Figure 6B ). Two epithelial markers were used as loading controls of the epithelium compartment, e-cadherin and cytokeratins (data not shown) with similar results.

Discussion

In this study, we showed for the first time that the BMP pathway is upregulated by H. pylori . Furthermore, we confirmed CDX2 and identified SOX2 as a novel target of this pathway as well as of H. pylori and demonstrated its downregulation concomitantly with CDX2 upregulation. These results provide novel information that contributes to understand the molecular events that precede the development of gastric IM, reinforcing the role of the BMP pathway in the whole process.

A frequent outcome of H. pylori infection of the gastric mucosa is the aberrant expression of CDX2 and the consequent development of gastric IM, a lesion that results from the transdifferentiation of the gastric mucosa to an intestinal phenotype that predisposes to cancer ( 1 , 2) . In a previous study, we obtained evidence for the involvement of BMPs in this process, since key elements of this pathway, in particular the activated (phosphorylated) form of the receptor-regulated SMADs, pSMAD1/5/8, were overexpressed in IM lesions and CDX2 was upregulated by the BMPs in AGS cells ( 12 ). In the chain of events leading to gastric IM, BMP pathway activation would presumably appear following H. pylori infection. Concordantly, in this study, we show for the first time that H. pylori upregulates this pathway, demonstrated by the increased expression of BMP2, SMAD4 and pSMAD1/5/8 and by the increased interaction between SMAD4 and pSMAD1/5/8 determined by PLA in in vitro co-culture experiments. Most compelling, the results obtained in vivo lend support to this hypothesis, since overexpression of pSmad1/5/8, the hallmark of an active pathway, was observed upon infection of the mouse gastric mucosa with two Helicobacter strains and was accompanied by de novo CDX2 expression. In accordance with our results, an upregulation of BMP2 was described previously upon H. pylori infection of MKN45 gastric cell line, detected by microarray analysis ( 39 ). However, this cell line does not have an active BMP pathway due to the lack of SMAD4 expression, and therefore, we did not use it further ( 12 ). Our results are also in accordance with a previous observation of an influx of BMP-expressing inflammatory cells to the stomach following infection with H. pylori , in human tissues ( 40 ). In our study, however, the inflammation developed in mice stomachs was mild, which may explain the also mild activation of the pathway. The BMP pathway was also shown to be involved in the onset of Barrett’s oesophagus, which is a lesion similar to gastric IM and also involving CDX2 ( 41 , 42) . Two studies have shown that BMPs and pSMAD1/5/8 were overexpressed in different models of oesophagitis and Barrett’s oesophagus, with only one of the studies showing association with CDX2 expression ( 42 ). The clues that can be taken from the Barrett’s oesophagus model reinforce the importance of the BMP pathway in regenerating oesophageal in addition to gastric mucosa after injury, through a transdifferentiation mechanism, and suggest that a certain threshold of BMP activation is probably needed in order to induce intestinal differentiation. In addition, it has also been observed that the BMP pathway is recruited to regenerate the respiratory tract after acute injury, recapitulating the role it has during lung development ( 43 ).

Here, we reinforce the previously identified BMP–CDX2 interaction, which we show, in the current study, to be initiated by H. pylori . Most interestingly, Cdx2 de novo expression was detected in the gastric mucosa of mice infected both with H. pylori and H. felis , in discrete foci and in the absence of morphological alterations resembling IM. To the best of our knowledge, this is the first description of Cdx2 de novo expression in the gastric mucosa of mice upon infection with Helicobacter spp. The presence of CDX2 in partial or whole glands of the gastric antrum without further intestinal differentiation was also observed in humans by others and by us ( 44–47 ). In humans, this was considered a reversible expression, induced by H. pylori infection and not related with IM development. Our interpretation for both human and now mice focal Cdx2 expression induced by H. pylori is that it occurs in cells already committed to terminal gastric differentiation and therefore not capable of initiating a metaplasia process. On the contrary, we speculate that CDX2 expression will lead to gastric IM if occurring in gastric stem cells.

Mice are the most convenient experimental models to study Helicobacter spp. infection but it is well known that they also offer limitations since they do not completely mimic the progress of the human disease, developing instead a relatively mild gastritis that does not evolve to IM and rarely develop gastric cancer. Now that we have shown aberrant Cdx2 expression in the mice gastric mucosa upon Helicobacter spp. infection, it will be interesting to find the missing gaps for IM and gastric cancer development in the animal model ( 48 , 49) .

In addition to CDX2, we identified SOX2 as a novel target of the regulation by H. pylori and by the BMP pathway. SOX2 is emerging as a gastric transcription factor that could be further and cooperatively involved in CDX2 regulation and in IM onset ( 32 ). At present, the function of SOX2 in the stomach and in IM is not as clear as the function of CDX2 in intestinal differentiation and IM development. SOX2 is mostly recognized by its involvement in the maintenance of embryonic stem cell pluripotency and is included in the ‘cocktail’ of genes that are able to induce pluripotency in differentiated cells ( 29 ). During embryonic development, it is clear that Sox2 is necessary for the foregut differentiation in mouse, whereas in chick embryo, it was demonstrated that SOX2 expression ends in the boundary of the intestine, where CDX2 expression begins, thus being mutually exclusive ( 30 , 50) . The expression and function of SOX2 in adult tissues is not so well characterized but recently it was shown that it is expressed in the stomach, among other epithelial tissues, in cells with stemness properties critical for normal tissue regeneration. SOX2 expression was not detected in the small intestine and colon, in accordance with previous reports ( 32 ). What we show here is that SOX2 expression is strongly downregulated both by H. pylori and the BMP pathway, suggesting that activation of an intestinal differentiation program occurs concomitantly with the silencing of a gastric differentiation one, induced or controlled by SOX2. We could not clarify, however, if the BMP pathway acts as a mediator of H. pylori in this regulation. Our results suggest that it is not, but other issues need to be taken into account, namely that SMAD4i cells still have some SMAD4, which might be enough to transduce this signal. In accordance with our results, it has been shown that the BMP pathway also represses Sox2 expression in the respiratory tract ( 51 ). Altogether, our results show that H. pylori , through the BMP pathway upregulates an intestinal differentiation program and downregulates a gastric differentiation one, which will eventually lead to the onset of IM.

Funding

Fundação para a Ciência e a Tecnologia (FCT)-Programa Operacional Ciência e Inovação 2010 do Quadro Comunitário de Apoio III and FEDER (Project PTDC/SAU-OBD/64490/2006). IPATIMUP—Associate Laboratory of the Portuguese Ministry of Science, Technology and Higher Education—is partially supported by FCT. FCT (SFRH/BD/63300/2009 to V.C.) and (SFRH/BPD/68276/2010 to R.B.)

Acknowledgements

The authors wish to thank Dr Paula Chaves and Dr António S. Guerreiro for creating the opportunity to use the mouse experimental model and Rui Ferreira for technical assistance.

References

1.
Correa
P.
1992
Human gastric carcinogenesis: a multistep and multifactorial process–First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention.
Cancer Res.
 
52
6735
6740
2.
Uemura
N.
et al
2001
Helicobacter pylori infection and the development of gastric cancer.
N. Engl. J. Med.
 
345
784
789
3.
Suh
E.
et al
1996
An intestine-specific homeobox gene regulates proliferation and differentiation.
Mol. Cell. Biol.
 
16
619
625
4.
Mesquita
P.
et al
2006
Metaplasia–a transdifferentiation process that facilitates cancer development: the model of gastric intestinal metaplasia.
Crit. Rev. Oncog.
 
12
3
26
5.
Almeida
R.
et al
2003
Expression of intestine-specific transcription factors, CDX1 and CDX2, in intestinal metaplasia and gastric carcinomas.
J. Pathol.
 
199
36
40
6.
Eda
A.
et al
2003
Aberrant expression of CDX2 in Barrett’s epithelium and inflammatory esophageal mucosa.
J. Gastroenterol.
 
38
14
22
7.
Osawa
H.
et al
2004
Aberrant expression of CDX2 in the metaplastic epithelium and inflammatory mucosa of the gallbladder.
Am. J. Surg. Pathol.
 
28
1253
1254
8.
Beck
F.
et al
1999
Reprogramming of intestinal differentiation and intercalary regeneration in Cdx2 mutant mice.
Proc. Natl. Acad. Sci. U.S.A.
 
96
7318
7323
9.
Silberg
D.G.
et al
2002
Cdx2 ectopic expression induces gastric intestinal metaplasia in transgenic mice.
Gastroenterology
 
122
689
696
10.
Mutoh
H.
et al
2002
Conversion of gastric mucosa to intestinal metaplasia in Cdx2-expressing transgenic mice.
Biochem. Biophys. Res. Commun.
 
294
470
479
11.
Mutoh
H.
et al
2004
Development of gastric carcinoma from intestinal metaplasia in Cdx2-transgenic mice.
Cancer Res.
 
64
7740
7747
12.
Barros
R.
et al
2008
Key elements of the BMP/SMAD pathway co-localize with CDX2 in intestinal metaplasia and regulate CDX2 expression in human gastric cell lines.
J. Pathol.
 
215
411
420
13.
Barros
R.
et al
2009
CDX2 expression is induced by Helicobacter pylori in AGS cells.
Scand. J. Gastroenterol.
 
44
124
125
14.
Manzo
B.A.
et al
2006
Helicobacter pylori regulates the expression of inhibitors of DNA binding (Id) proteins by gastric epithelial cells.
Microbes Infect.
 
8
1064
1074
15.
von Bubnoff
A.
et al
2001
Intracellular BMP signaling regulation in vertebrates: pathway or network?
Dev. Biol.
 
239
1
14
16.
Chen
Y.G.
et al
1998
Determinants of specificity in TGF-beta signal transduction.
Genes Dev.
 
12
2144
2152
17.
Miyazono
K.
2000
Positive and negative regulation of TGF-beta signaling.
J. Cell. Sci.
 
113 (Pt 7)
1101
1109
18.
Attisano
L.
et al
2002
Signal transduction by the TGF-beta superfamily.
Science
 
296
1646
1647
19.
Takaku
K.
et al
1999
Gastric and duodenal polyps in Smad4 (Dpc4) knockout mice.
Cancer Res.
 
59
6113
6117
20.
Haramis
A.P.
et al
2004
De novo crypt formation and juvenile polyposis on BMP inhibition in mouse intestine.
Science
 
303
1684
1686
21.
Hardwick
J.C.
et al
2004
Bone morphogenetic protein 2 is expressed by, and acts upon, mature epithelial cells in the colon.
Gastroenterology
 
126
111
121
22.
Auclair
B.A.
et al
2007
Bone morphogenetic protein signaling is essential for terminal differentiation of the intestinal secretory cell lineage.
Gastroenterology
 
133
887
896
23.
Howe
J.R.
et al
1998
Mutations in the SMAD4/DPC4 gene in juvenile polyposis.
Science
 
280
1086
1088
24.
Sayed
M.G.
et al
2002
Germline SMAD4 or BMPR1A mutations and phenotype of juvenile polyposis.
Ann. Surg. Oncol.
 
9
901
906
25.
Barros
R.
et al
2009
Juvenile polyps have gastric differentiation with MUC5AC expression and downregulation of CDX2 and SMAD4.
Histochem. Cell Biol.
 
131
765
772
26.
Benahmed
F.
et al
2008
Multiple regulatory regions control the complex expression pattern of the mouse Cdx2 homeobox gene.
Gastroenterology
 
135
1238
1247, 1247.e1
27.
Asonuma
S.
et al
2009
Helicobacter pylori induces gastric mucosal intestinal metaplasia through the inhibition of interleukin-4-mediated HMG box protein Sox2 expression.
Am. J. Physiol. Gastrointest. Liver Physiol.
 
297
G312
G322
28.
Boiani
M.
et al
2005
Regulatory networks in embryo-derived pluripotent stem cells.
Nat. Rev. Mol. Cell Biol.
 
6
872
884
29.
Takahashi
K.
et al
2006
Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors.
Cell
 
126
663
676
30.
Que
J.
et al
2007
Multiple dose-dependent roles for Sox2 in the patterning and differentiation of anterior foregut endoderm.
Development
 
134
2521
2531
31.
Tsukamoto
T.
et al
2004
Down-regulation of a gastric transcription factor, Sox2, and ectopic expression of intestinal homeobox genes, Cdx1 and Cdx2: inverse correlation during progression from gastric/intestinal-mixed to complete intestinal metaplasia.
J. Cancer Res. Clin. Oncol.
 
130
135
145
32.
Arnold
K.
et al
2011
Sox2(+) adult stem and progenitor cells are important for tissue regeneration and survival of mice.
Cell Stem Cell
 
9
317
329
33.
Tani
Y.
et al
2007
Transcription factor SOX2 up-regulates stomach-specific pepsinogen A gene expression.
J. Cancer Res. Clin. Oncol.
 
133
263
269
34.
Park
E.T.
et al
2008
Aberrant expression of SOX2 upregulates MUC5AC gastric foveolar mucin in mucinous cancers of the colorectum and related lesions.
Int. J. Cancer
 
122
1253
1260
35.
Mutoh
H.
et al
2011
Sox2 expression is maintained while gastric phenotype is completely lost in Cdx2-induced intestinal metaplastic mucosa.
Differentiation
 
81
92
98
36.
Santos
A.M.
et al
2011
Role of 13C-urea breath test in experimental model of Helicobacter pylori infection in mice.
Helicobacter
 
16
320
326
37.
Söderberg
O.
et al
2006
Direct observation of individual endogenous protein complexes in situ by proximity ligation.
Nat. Methods
 
3
995
1000
38.
Walsh
D.W.
et al
2010
Extracellular BMP-antagonist regulation in development and disease: tied up in knots.
Trends Cell Biol.
 
20
244
256
39.
Marcos
N.T.
et al
2008
Helicobacter pylori induces beta3GnT5 in human gastric cell lines, modulating expression of the SabA ligand sialyl-Lewis x.
J. Clin. Invest.
 
118
2325
2336
40.
Bleuming
S.A.
et al
2006
Altered bone morphogenetic protein signalling in the Helicobacter pylori -infected stomach.
J. Pathol.
 
209
190
197
41.
Milano
F.
et al
2007
Bone morphogenetic protein 4 expressed in esophagitis induces a columnar phenotype in esophageal squamous cells.
Gastroenterology
 
132
2412
2421
42.
Castillo
D.
et al
2012
Activation of the BMP4 pathway and early expression of CDX2 characterize non-specialized columnar metaplasia in a human model of Barrett’s esophagus.
J. Gastrointest. Surg.
 
16
227
37; discussion 237
43.
Masterson
J.C.
et al
2011
Bone morphogenetic protein signalling in airway epithelial cells during regeneration.
Cell. Signal.
 
23
398
406
44.
Vauhkonen
M.
et al
2008
Helicobacter pylori infection induces a reversible expression of the CDX2 transcription factor protein in human gastric epithelium.
Scand. J. Gastroenterol.
 
43
915
921
45.
Bornschein
J.
et al
2009
The combined presence of H. pylori infection and gastro-oesophageal reflux disease leads to an up-regulation of CDX2 gene expression in antrum and cardia.
J. Clin. Pathol.
 
62
254
259
46.
Barros
R.
et al
2010
Relevance of high virulence Helicobacter pylori strains and futility of CDX2 expression for predicting intestinal metaplasia after eradication of infection.
Scand. J. Gastroenterol.
 
45
828
834
47.
Peleteiro
B.
et al
2011
Association between environmental factors and CDX2 expression in gastric cancer patients.
Eur. J. Cancer Prev.
  , 21, 423–431.doi:10.1097/CEJ.0b013e32834ef1aa
48.
Marchetti
M.
et al
1995
Development of a mouse model of Helicobacter pylori infection that mimics human disease.
Science
 
267
1655
1658
49.
Lee
A.
et al
1997
A standardized mouse model of Helicobacter pylori infection: introducing the Sydney strain.
Gastroenterology
 
112
1386
1397
50.
Ishii
Y.
et al
1998
Region-specific expression of chicken Sox2 in the developing gut and lung epithelium: regulation by epithelial-mesenchymal interactions.
Dev. Dyn.
 
213
464
475
51.
Domyan
E.T.
et al
2011
Signaling through BMP receptors promotes respiratory identity in the foregut via repression of Sox2.
Development
 
138
971
981

Abbreviations:

    Abbreviations:
  • BMP

    bone morphogenetic protein

  • BSA

    bovine serum albumin

  • IM

    intestinal metaplasia

  • PLA

    proximity ligation assay