Effects of EGFR Inhibitor on Helicobacter pylori Induced Gastric Epithelial Pathology in Vivo

Helicobacter pylori transactivates the Epidermal Growth Factor Receptor (EGFR) and predisposes to gastric cancer development in humans and animal models. To examine the importance of EGFR signalling to gastric pathology, this study investigated whether treatment of Mongolian gerbils with a selective EGFR tyrosine kinase inhibitor, EKB-569, altered gastric pathology in chronic H. pylori infection. Gerbils were infected with H. pylori and six weeks later received either EKB-569-supplemented, or control diet, for 32 weeks prior to sacrifice. EKB-569-treated H. pylori-infected gerbils had no difference in H. pylori colonisation or inflammation scores compared to infected animals on control diet, but showed significantly less corpus atrophy, mucous metaplasia and submucosal glandular herniations along with markedly reduced antral and corpus epithelial proliferation to apoptosis ratios. EKB-569-treated infected gerbils had significantly decreased abundance of Cox-2, Adam17 and Egfr gastric transcripts relative to infected animals on control diet. EGFR inhibition by EKB-569 therefore reduced the severity of pre-neoplastic gastric pathology in chronically H. pylori-infected gerbils. EKB-569 increased gastric epithelial apoptosis in H. pylori-infected gerbils which counteracted some of the consequences of increased gastric epithelial cell proliferation. Similar chemopreventative strategies may be useful in humans who are at high risk of developing H. pylori- induced gastric adenocarcinoma.

The aims of this study were to investigate the effects of a selective EGFR tyrosine kinase inhibitor, EKB-569 [30] on H. pylori-induced gastric pathology, epithelial hyperproliferative responses and expression of genes in the EGFR triple membrane passing signalling (TMPS) cascade which are essential for EGFR transactivation [31]. Our previous longitudinal studies on pathology, epithelial proliferation and apoptosis in H. pylori SS1 strain infected Mongolian gerbils identified a significant progression to corpus atrophy and reduction in gastric epithelial cell apoptosis at 36 weeks post-infection [10]. Corpus atrophy is a recognised pre-cursor condition for gastric cancer [1][2][3]. Based on these earlier kinetic studies [10], the effects of EKB-569 treatment in H. pylori SS1 strain infected Mongolian gerbils were examined at 38 weeks post-infection to assess the effect of treatment on the development of the pre-neoplastic lesion corpus atrophy and associated changes in epithelial cell apoptosis, proliferation and gene expression.

EKB-569 Effects on H. pylori-Induced Epithelial Responses in Vitro
Previous in vitro studies using "In Cell Western" analysis identified that EKB-569 significantly inhibited H. pylori-stimulated ERK1/2 phosphorylation (pERK1/2) in A-431 cells [16]. Inhibitory effects were not specific to strains with a functional cag PAI. ERK1/2 phosphorylation in A-431 cells was examined as a readout of EGFR signalling rather than phosphorylated EGFR due to cross reactivity of pEGFR antibodies with H. pylori [16]. Initial in vitro studies determined whether EKB-569 inhibited SS1-induced pERK1/2 in A431 cells. At 100nm and 1,000nm EKB-569 inhibited circa 70% SS1-induced pERK1/2 relative to untreated SS1-stimulated cells (Table 1). Table 1. Inhibition of ERK1/2 phosphorylation induced by H. pylori SS1 strain in A-431 epithelial cells by EKB-569. A-431 cells were co-incubated with H. pylori strain SS1 for 180 min with, or without, 60 min EKB-569 pre-incubation. Unstimulated A-431 cells were similarly pre-incubated with, and without, EKB-569. ERK1/2 phosphorylation in H. pylori stimulated and unstimulated cells was quantified by "In Cell Western" analysis of the 700/800 nm Relative Response (RR) [16]. Values represent net values after subtraction of the effect of EKB-569 on the endogenous ERK1/2 phosphorylation in unstimulated cells. The significance between control H. pylori SS1 and EKB-569 treated groups was analysed by paired t-test (* p < 0.05; ** p < 0.001, Values are mean ± SEM of 4 independent experiments.

H. pylori Infection of Mongolian Gerbils
The effects of EKB-569 treatment on gastric pathology and corpus atrophy were examined at 38 weeks post-infection. At 38 weeks post-inoculation all gerbils on control diet (n = 7) or EKB-569 diet (n = 8) were H. pylori positive by microbial culture. Additionally, all inoculated gerbils on EKB-569 and control diet were classified histologically as H. pylori infected. No control gerbils were positive for H. pylori by histology or microbial culture (n = 10 per group). There was no significant difference in H. pylori antral density in EKB-569 treated gerbils (density grade mean ± SEM = 1.38 ± 0.20) compared to gerbils on control diet (1.43 ± 0.26), or in the corpus mucosa (control and EKB-569 diet, 1.0 ± 0.0).

Gastric Pathology in EKB-569 Treated and Control H. pylori-Infected Gerbils
In H. pylori-infected gerbils, chronic and active inflammation and atrophy were evident in the antrum at 38 weeks post-infection, but there was no significant difference in scores between infected gerbils on control diet and EKB-569 diets (data not shown). In the corpus mucosa, chronic and active inflammatory scores in H. pylori-infected gerbils did not differ between EKB-569 treated and untreated gerbils ( Figure 1A,B). In contrast, corpus atrophy ( Figure 1C) and corpus mucous metaplasia ( Figure 1D) in H. pylori-infected gerbils was significantly lower (p < 0.05) in EKB-569 treated gerbils. Uninfected control gerbils on normal diet and EKB-569 diet had histologically normal antral and corpus mucosa. (D) mucous metaplasia scores. Open circles gerbils on control diet, closed triangles gerbils on EKB-569 diet, n = 7-8 per group. Comparisons between groups were made using stratified Chi squared for trend. Bars represent median scores.

EGFR Inhibitor Reduces Phosphorylated Erk Positive Gastric Epithelial Cells in H. pylori-Infected Gerbils
Immunohistochemistry was used to assess phosphorylation of Erk1/2 in the gastric mucosa in the Mongolian gerbils. At 38 weeks post-infection H. pylori infected gerbils on control diet had significantly increased (p = 0.01) phosphorylated Erk1/2 gastric epithelial cells relative to the uninfected controls ( Figure 3A,B). Both cytoplasmic, and in some instances, nuclear phosphorylated Erk positive epithelial cells were evident mainly in the upper parts of the gastric mucosa in H. pylori infected gerbils on the control diet ( Figure 3A). However, phospho-Erk positivity was also evident in gastric epithelial cells in the sub-mucosal glandular herniations in infected gerbils on control diet ( Figure 3C). Treatment of H. pylori infected gerbils with EKB-569 resulted in a significant reduction in phospho-ERK positive epithelial cells compared to H. pylori infected gerbils on control diet ( Figure 3D).
Phospho-Erk1/2 positive mononuclear cells in the lamina propria were evident in H. pylori infected gerbils both on the control diet and EKB-569 diet. In uninfected gerbils on control diet and EKB-569 diet there were very few phospho-Erk positive gastric epithelial or lamina propria cells ( Figure 3B,E,F). The number of phospho-Erk positive gastric epithelial cells per high-power field (×40) in uninfected and H. pylori infected gerbils on control and EKB-569 diet is shown in Figure 3F. The results indicate that EKB-569 significantly reduces H. pylori induced gastric phospho-Erk signalling in gastric epithelial cells in vivo (p < 0.002) ( Figure 3F) as well as in vitro (Table 1).

EGFR Inhibitor Modifies Gastric Epithelial Cell Proliferation Apoptosis Ratios in H. pylori-Infected Gerbils
In uninfected animals no differences in epithelial proliferation or apoptosis were observed between gerbils on control or EKB-569 diet ( Figure 4A-D). At 38 weeks post-infection there was a significant increase in antral (p < 0.001) and corpus (p < 0.001) epithelial cell proliferation in EKB-569 treated and untreated gerbils compared to uninfected controls ( Figures 4A-B). In infected gerbils the corpus epithelial proliferation index in EKB-569 treated gerbils was lower than that of infected gerbils on control diet ( Figure 4B). Antral and corpus epithelial apoptosis was significantly increased in H. pylori-infected gerbils treated with EKB-569 compared to H. pylori positive gerbils on control diet (p < 0.01) and uninfected control groups (p < 0.001) ( Figure 4C,D).
The ratio of gastric epithelial cell proliferation to apoptosis in EKB-569 treated control and H. pylori infected gerbils is shown in Figure 3E,F. In uninfected gerbils, EKB-569 treatment had no effect on the proliferation to apoptosis ratio. In H. pylori infected gerbils on control diet there was a significant increase in antral epithelial cell proliferation to apoptosis ratio compared to uninfected controls (p < 0.02) and the H. pylori infected group treated with EKB-569 (p < 0.01). EKB-569 treated infected gerbils had an antral proliferation to apoptosis ratio comparable to uninfected control groups ( Figure 4E). In the corpus the gastric epithelial proliferation to apoptosis ratio was similarly significantly increased in H. pylori infected gerbils on control diet relative to uninfected controls (p < 0.01) and the H. pylori infected group treated with EKB-569 (p < 0.02) ( Figure 4F). The EKB-569 treated infected gerbils had a corpus proliferation to apoptosis ratio comparable to uninfected untreated controls, but was greater than the uninfected EKB-569 treated group (p = 0.05).

Sequence Analysis of Gerbil Transcripts
To facilitate analysis of the effects of EKB-569 on gastric expression of genes involved in H. pylori induced EGFR transactivation, RT-PCR products produced from transcripts coding for Adam17, Egfr and Hb-Egf were analysed by agarose gel electrophoresis, DNA bands were excised and gerbil DNA sequenced. The sequences were aligned with respective rodent and human sequences and showed a high degree of homology ( Figure S1). The 213 base pair nucleotide encoding for Mongolian gerbil Adam17 (EMBL Accession number HF679116) was 95%, 95%, 95% and 94% identical to mouse, rat, Chinese hamster and human orthologues, respectively ( Figure S1A). The 216 base pair nucleotide encoding Mongolian gerbil Egfr (EMBL Accession number HF679117) was 92%, 91% and 80% identical to mouse, rat and human orthologues respectively ( Figure S1B). The 200 base pair nucleotide encoding Mongolian gerbil Hb-egf (EMBL Accession number HF679118) was 93%, 93%, 90% and 79% identical to mouse, rat, Chinese hamsters and human orthologues, respectively (Figure S1C).

Effects of EGFR Inhibitor on H. pylori Induced Gastric Gene Expression in Gerbils
As Ifn-γ transcripts in the gastric mucosa of H. pylori-infected gerbils have previously been shown to correlate with the grade of chronic inflammation [10], Ifn-γ transcripts were compared in H. pylori-infected EKB-569 treated and untreated gerbils. Transcripts were quantified using the 2 −ΔΔCt method [32] using Gapdh as a reference gene. No significant difference was evident in Ifn-γ transcript abundance in H. pylori-infected gerbils on EKB-569 or control diet (data not shown) concurring with their similar chronic inflammatory scores (Figure 1). The miRNAs miR-155 and miR-146a are increased in human gastric mucosa in H. pylori infection [33,34]. miRNAs were quantified using the 2 −ΔΔCt method [32] using U6 snRNA as a reference gene. Quantitative PCR showed transcript abundances of miR-146a and miR-155 were also significantly increased (p < 0.01) in H. pylori SS1-infected gerbils compared to uninfected controls ( Figure 5A,B). There was no significant difference in gastric miR-146a and miR-155 transcript abundance in infected gerbils on EKB-569 or control diet. Analysis of Adam-17, Egfr and Cox-2 in the gastric mucosa of uninfected control gerbils showed that EKB-569 had no significant effect on transcript abundance of these genes ( Figure 6A-C). In contrast H. pylori infection resulted in significant increases in gastric mucosal Adam17 (p < 0.01), Egfr (p < 0.05) and Cox-2 (p < 0.01) transcript abundance in gerbils on control diet ( Figure 6A-C). EKB-569 treatment of H. pylori-infected gerbils significantly reduced the abundance of Adam17 (p < 0.01), Cox-2 (p < 0.01) and decreased Egfr gastric mucosal transcripts ( Figure 5A-C). HB-EGF expression is upregulated in cultured human gastric epithelial cells by H. pylori [14,35] and HB-EGF is also present in parietal cells [36]. EKB-569 treatment of uninfected gerbils resulted in increased Hb-egf transcripts (p < 0.05) ( Figure 6D). No increase in Hb-egf transcripts was observed in H. pylori-infected gerbils on control diet relative to uninfected controls on control diet. However in H. pylori-infected gerbils a significant decrease (p < 0.01) was evident in animals treated with EKB-569 ( Figure 6D). Hatched bars infected and control groups treated with EKB-569 diet, unhatched bars infected and control groups on control diet. Group size n = 7-10 gerbils. Results are expressed as mean ± SEM. * p < 0.05; ** p < 0.01 Mann-Whitney Test.

Discussion
H. pylori can directly transactivate the EGFR on epithelial cells [14][15][16] and peptides such as gastrin [37], chemokines [38] and molecules such as PGE 2 [39] which are upregulated by H. pylori will also, via their specific G coupled protein receptors, induce EGFR transactivation. Thus, in vivo, multiple pathways induced by H. pylori may contribute to EGFR transactivation via the TMPS cascade, which involves extracellular cleavage of membrane bound EGFR ligands such as proHB-EGF and amphiregulin by ADAM metalloproteases [4].
EKB-569 is a potent selective irreversible inhibitor of EGFR tyrosine kinase activity both in vitro and in vivo in rodents at clinically relevant doses [30,40]. EKB-569 also has reduced activity against ErbB-2 (HER2/neu) [30] which is also activated by H. pylori [41]. EKB-569 administration in dietary formulation blocks intestinal polyp formation in APC Min/+ mice [30] and combination therapy with cyclooxygenase inhibitors results in synergistic prevention of polyp development [30]. This latter study only investigated polyp formation and did not evaluate any effects of EKB-569 on epithelial kinetics and intestinal gene expression.
In the present study EKB-569 significantly reduced H. pylori phosphorylation of ERK1/2 in gastric epithelial cells both in vitro and in vivo. In addition, nuclear phospho-Erk was mainly observed in H. pylori infected gerbils on control diet. Previous studies indicate nuclear accumulation of Erk2 depends on phosphorylation state and dimerization [42]. Phospho-Erk positive mononulcear cells in the lamina propria were present in infected gerbils on both control diet and EKB-569 diet. Although EGFR is strongly expressed on the basal lateral membranes of gastrointestinal epithelial cells [43], EGFR is also present in myeloid cells [44] and fibroblasts [45] which are increased in H. pylori infection [46]. The lack of specific reagents for gerbils precluded detailed analysis of the lamina propria specific mononuclear cell populations. The potential effects of EKB-569 on the function of lamina propria mononuclear cell populations should be considered in future studies given the importance of epithelial-stromal interactions in disease pathogenesis [46].
In the present study in vivo EKB-569 treatment of gerbils during weeks 6 to 38 post-infection had no effect on density of H. pylori colonisation or gastric inflammation determined histologically and by assessment of gastric Ifnγ, miR-146a and miR-155 transcripts, respectively. Gastric miR-146a and miR-155 transcripts, which were significantly increased in H. pylori infected gerbils on both EKB-569 and control diets, negatively regulate the inflammatory response to H. pylori [33,34]. In contrast, in H. pylori infected gerbils, EKB-569 treatment was associated with significantly reduced corpus atrophy (the pre-neoplastic lesion end point), mucous metaplasia and glandular submucosal herniations and a marked reduction in antral and corpus epithelial to apoptosis ratios. Consistent with previous studies with H. pylori SS1 strain [10], no gerbils in this study had dysplasia or cancer.
EKB-569 treatment of infected gerbils significantly increased gastric epithelial apoptosis relative to infected gerbils on control diet at 38 weeks post-infection. Enhanced gastric epithelial cell apoptosis and the reduced proliferation to apoptosis ratio may account for the significant reduction in submucosal herniations, the latter being a proliferative response considered to be a first-step in the development of adenocarcinoma in this model. Whether enhanced apoptosis contributes to the significant reduction in corpus atrophy and a diminished requirement for mucous metaplasia requires further investigation. Earlier studies on murine airway epithelium have shown EKB-569 interrupts viral induced anti-apoptotic pathways and partially inhibits goblet cell metaplasia [47]. Previous studies have shown a significant reduction in gastric epithelial cell apoptosis in gerbils at 38 weeks post-infection with H. pylori SS1 strain relative to 12 weeks post-infection [10]. The H. pylori CagA protein can inhibit vacuolating cytotoxin induced apoptosis in epithelial cell lines [48]. However as the SS1 strain does not have a functional cag PAI [49], epithelial CagA translocation is unlikely to be involved in the decline in epithelial apoptosis in untreated infected gerbils.
The increased gastric epithelial apoptosis in EKB-569 treated infected gerbils was associated with a significant decrease in Cox-2, Adam17 and Egfr gastric transcripts. Gastric cyclooxygenase-2 is increased in H. pylori infection [50] and elevated levels of PGE 2 will transactivate the EGFR [39] decreasing apoptosis. Reduction of Cox-2 gastric transcripts in EKB-569 treated gerbils is in line with previous studies showing that upregulation of COX-2 protein in human [51] and murine [52] gastric epithelial cell lines by H. pylori is dependent on EGFR transactivation. Furthermore, pharmacological inhibition of Cox-2 upregulation in H. pylori stimulated murine gastric epithelial cells increases apoptosis [52]. Clinically in H. pylori infected patients with gastric ulcers the use of non-steroidal anti-inflammatory drugs reduces the risk of gastric cancer [53]. The synergistic effects of cycloxygenase inhibitors and EKB-569 on polyp formation in APC Min/+ mice [30] may similarly result from increased apoptosis.
Surprisingly no significant increase in gastric Hb-egf transcripts was observed in untreated gerbils with H. pylori infection relative to uninfected controls, although a significant reduction of Hb-egf transcript abundance was evident in infected gerbils with EKB-569 treatment. Although antral mucosa was used to avoid Hb-egf positive parietal cells [36], possible parietal cells in the transition zone in two control gerbils with high Hb-egf transcripts may have precluded the expected increase in Hb-egf transcripts with H. pylori infection. In vitro studies with human MKN-28 gastric cancer cells have shown H. pylori induced upregulation of COX-2 and HB-EGF, but not ADAM17, is reduced by EKB-569 treatment [54]. EGF stimulated upregulation of COX-2 and HB-EGF, but not ADAM17, in human MKN-28 gastric epithelial cells are also downregulated by EKB-569 treatment [54].
The membrane metalloprotease ADAM17 has a critical role in cleaving proHB-EGF on gastric epithelial cells following H. pylori stimulation resulting in EGFR transactivation by mature HB-EGF [14,18]. In vitro studies indicate that ADAM17 is required for the anti-apoptotic effect of H. pylori [18]. The significant decrease in Egfr transcripts in the gastric mucosa of infected gerbils with EKB-569 treatment probably reflects an overall reduction in gastric epithelial cells. These in vivo observations in gerbils concur with previous studies in an immortalised gastric epithelial cell line in which EGFR knockdown by siRNA enhanced H. pylori stimulated apoptosis [18].
In gastric cancer the expression of ADAM10, ADAM15, ADAM17 and ADAM 20 transcripts is markedly increased [19]. In addition, increases in EGF-related peptides HB-EGF and amphiregulin are evident in patients with H. pylori infection and/or gastric cancer [20,22]. Upregulation of these key components for EGFR transactivation by H. pylori could promote autocrine signaling loops promoting epithelial hyperproliferation and block apoptosis.
Although no significant reduction in gastric epithelial proliferation was identified with EKB-569 treatment in infected gerbils, in vivo rodent studies with other carcinogens such as asbestos, which also transactivates the EGFR [55], indicate bronchial hyperproliferative responses are dependent on EGFR activation [56]. Asbestos induced inflammation alone in transgenic mice with mutant EGFR was not associated with epithelial hyperproliferative responses nor proto-oncogene activation [56]. In H. pylori SS1-infected gerbils gastric epithelial proliferation, but not apoptosis, correlates with active and chronic inflammatory scores [10]. The lack of reduction in epithelial cell proliferation with EKB-569 treatment in the present study is thus consistent with absence of changes in gastric inflammation in this model. Studies in immune deficient transgenic mice also indicate the important role of gastric inflammation in development of epithelial hyperproliferative responses to gastric Helicobacter infection [57]. The lack of effect of EKB-569 on antral epithelial proliferation in infected gerbils suggests multiple pathways may contribute to the epithelial hyperproliferative response.

H. pylori Culture
H. pylori SS1 strain was grown on blood agar base with 7% (v/v) horse blood under microaerobic conditions at 37 °C. For in vitro experiments two day cultures were harvested into antibiotic-free RPMI 1640 medium and suspended at a concentration of 2.5 × 10 7 organisms/mL. For in vivo experiments two day SS1 cultures were harvested into sterile tryptose soya broth and used immediately for inoculation into gerbils.

Infection of Mongolian Gerbils with H. pylori
Male Mongolian gerbils aged 6-8 weeks old (supplied by MGS/Sea, Seac Yoshimoto, Fukuoka, Japan and bred at the University of Leeds) were inoculated three times by oral gavage with H. pylori SS1 (>10 8 CFU) over a 5 day period. Inoculated and control gerbils were maintained under a 12 h light/dark cycle, temperature 19-23 °C and humidity 45%-55%. After six weeks inoculated and control animals (n = 8-10) received per os 10 mgs/kg/day/gerbil EKB-569 in dietary formulation or control. Previous rodent studies [30] have demonstrated the efficacy of this dose in long term inhibition of EGFR activity. All experimental procedures were approved by the local ethics committee of Leeds University and the UK Home Office.

Histological and Microbial Analysis of H. pylori Infection
Gerbils were sacrificed at 38 weeks post-inoculation. One hour prior to sacrifice inoculated gerbils and controls received an intra-peritoneal injection of bromodeoxyuridine (BrdU) (50 mg/kg). At sacrifice, gastric tissue was taken for microbial culture, histology and snap frozen in liquid nitrogen and stored at −-70 °C for subsequent RNA extraction. H. pylori culture from gastric mucosa was on selective plates as previously described [12].
Gastric pathology was graded on haematoxylin and eosin stained sections by a pathologist in a blinded fashion as previously described [10,12]. The entire field of four sections of gastric mucosa was evaluated histologically to give one score for antral and one score for corpus mucosa per gerbil. Gerbils were scored for active and chronic inflammation, atrophy, and mucous metaplasia. All parameters were graded on a scale of 0-3, with 0 being histologically normal, 1-mild, 2-moderate and 3-severe abnormality. In addition, the extent of herniations below the muscularis mucosa was scored on a scale 0-3, with a score of zero being no herniations, 1-mild, 2-moderate and 3 severe submucosal herniations. Density of H. pylori in antrum and corpus was assessed on Giemsa stained sections as previously described [10]. Gastric epithelial cell proliferation was assessed by BrdU immunohistochemistry as previously described [10][11][12]. Epithelial cell proliferation was expressed as a labelling index (LI%) ([percentage of stained cells/total cells per gastric gland] ×100).
Apoptotic epithelial cells were identified immunohistochemically with a rabbit polyclonal antibody against active caspase 3 (Cell Signalling Technology, Danvers, MA, USA). Glandular epithelial apoptosis was expressed as an apoptotic index (AI%) ([percentage of stained cells /total cells per gastric gland] ×100) as previously described [10,11]. Quantitative assessments were undertaken on a Nikon E1000 microscope (Nikon Inc., Melville, NY, USA). To evaluate changes in epithelial cell proliferation and apoptosis, a score for epithelial cell kinetics in individual gerbils was determined by dividing the epithelial proliferation index (LI%) by the apoptotic index (AI%) [7].

Sequence Analysis of Mongolian Gerbil Genes
Primers specific for Adam17, Egfr and Hb-Egf transcripts were designed by cross species PCR after aligning respective murine, rat and human sequences and identifying regions of homology. PCR products were sequenced as previously described [27].

Quantitative Polymerase Chain Reaction (qPCR)
Gastric mucosal RNA was extracted in Trizol (Invitrogen, Paisley, UK), DNase treated (Applied Biosystems, Warrington, UK) and reverse-transcribed with SuperScript II Reverse Transcriptase (Invitrogen, Paisley, UK) and random hexamers (Bioline, London, UK). Real-time qPCR amplification was performed using gene-specific oligonucleotide primers for Adam17, Cox-2, Egfr, Hb-egf, Ifn-γ, and Gapdh (Table 2) and SYBR Green PCR Master Mix (Applied Biosystems). Quantitative PCR amplification consisted of an initial denaturing step at 95 °C for 10 min, 40 cycles of 95 °C for 15 s and 60 °C for 1min. Reactions were performed in a 7500 Real Time PCR System (Applied Biosystems) and quantified using 2 −ΔΔCt method [32] using Gapdh as a reference gene.

Quantitative PCR Analysis of miRNAs Expression in Gerbil Gastric Mucosa
miRNAs were reverse-transcribed from extracted gastric RNA using the miRNome microRNA Profilers QuantiMir™ kit (System Biosciences; Mountain View, CA, USA) following the manufacturer's instructions. Quantitative PCR for miR-146a and miR-155 was performed by real-time PCR, using U6 snRNA as reference gene using forward primers detailed in Table 2 and the 3' universal reverse primer (System Biosciences). Quantitative PCR was as above. miRNAs were quantified using the 2 −ΔΔCt method [32] using U6 snRNA as a reference gene.

Statistics
Results are presented as means ± S.E.M. Comparisons between groups were made using either Mann-Whitney U test, paired t test, stratified Chi squared for trend or Fishers exact test. A p value of less than <0.05 was considered significant.

Conclusions
In conclusion, this study demonstrates that treatment of H. pylori infected gerbils with a selective EGFR inhibitor improves gastric corpus pathology and precancerous lesions and substantially decreases both the antral and corpus epithelial to apoptosis ratios to those of uninfected gerbils. A significant increase in apoptosis in infected gerbils treated with the EFGR inhibitor was associated with reduced mucosal transcripts of Cox-2, Adam17 and Egfr. These studies indicate for the first time a pathogenic role of H. pylori EGFR transactivation in vivo during long term chronic H. pylori infection. Targeting pre-malignant proliferative signalling pathways in non-malignant conditions with tyrosine kinase inhibitors is considered to have therapeutic potential for some conditions [59]. Whether such an approach with EGFR inhibitors would have a chemopreventative role in subjects who are at high risk of developing gastric cancer, who have failed H. pylori eradication therapy, requires consideration.

Acknowledgments
The study was undertaken with financial support of Yorkshire Cancer Research and funding under the Sixth Research Framework Programme of the European Union, Project INCA (LSHC-CT-2005-018704).

Conflicts of Interest
None of the authors have a conflict of interest. EKB-569 was provided by Wyeth Research Chemical Sciences who have had no role in study design, data analysis or preparation of the manuscript.