IL-1RT1 signaling antagonizes IL-11 induced STAT3 dependent cardiac and antral stomach tumor development through myeloid cell enrichment.

IL-1 is key driver of gastric tumorigenesis and is a downstream target of IL-11 signaling. Recently, IL-1 cytokines, particularly IL-1β, have been flagged as therapeutic targets for gastric cancer treatment. Here, we assess the requirement for IL-1 signaling in gastric tumorigenesis. gp130757FF xIL-1RT1-/- mice were generated to determine the pathological consequence of ablated IL-1 signaling in the IL-11 dependent gp130757FF mouse model of gastric tumorigenesis. Gastric lesions in gp130757FF xIL-1RT1-/- mice were increased in incidence and size compared to gp130757FF mice. Proximal gastric lesions originated from the cardiac region and were associated with elevated STAT3 activation, loss of specialized gastric cells and a modulated immune response including increased expression of TNF-α and MDSC associated genes. Administration of IL-11 to IL-1RT1-/- mice showed similar changes to gp130757FF xIL-1RT1-/- mice. Spleens from IL-11 treated wildtype mice showed an enrichment of MDSC and gp130757FF xIL-1RT1-/- mice had increased MDSCs in the stomach compared to gp130757FF mice. Furthermore, crossing TNF-α-/- to gp130757FF mice resulted in reduced lesion size. We conclude that IL-1 signaling antagonizes IL-11/STAT3 mediated pathology and the genetic deletion of IL-1RT1 results in increased tumor burden. We provide evidence that a likely mechanism is due to IL-11/STAT3 dependent enrichment of MDSCs.


INTRODUCTION
Historically gastric cancers are categorized into 2 main types according to phenotypic and histological characteristics; diffuse carcinomas and intestinal-type gastric cancers (IGC) [1].IGC develop incrementally after the initial Helicobacter pylori (H.pylori) infection and acute inflammation, progressing through chronic inflammation, and metaplasia to neoplasia [2].IGC occur with persistent H. pylori infection, most prominently in the distal gastric antrum and also in the cardia.The cardia is a discrete region of glands distinct from the fundus or body of the stomach, located at the gastro-esophageal junction (GEJ) in humans [3], or the limiting ridge between the squamous esophagus and forestomach in mice [4].These glands are similar to antral glands in structure, being composed primarily of mucous cells, with other specialized cell types being absent.The normal function of cardiac glands is not well defined, nonetheless recent large scale retrospective population studies in humans have shown that cancers of the gastroesophageal junction are increasing in prevalence [5], and that patients with GEJ cancers have reduced life expectancy compared to those with antral IGC [6].
Numerous cytokines have enhanced expression during chronic H. pylori infection and subsequent gastric disease progression, in particular members of the IL-1 and IL-6 families.Elevated levels of these cytokines in the absence of other mitigating factors, including H. pylori, cause gastric pathology in murine models [4,7,8].Clinical studies have demonstrated a strong correlation between IL-1β polymorphisms and predisposition to gastric cancer development [8][9][10].In mice, IL-1β is increased following H. pylori infection [10], and transgenic over-expression of IL-1β using the H/K ATPase promoter caused gastric inflammation and dysplasia [8].More recently, the Epstein-Barr virus promoter/IL-1β transgenic mouse was shown to develop cardiac and esophageal pathology [4].The link between IL-1β expression and gastric cancer progression in humans and murine models, has resulted in IL-1 receptor blockade being suggested as a novel therapeutic target in the effort to combat gastric cancer progression [11,12].
IL-1β belongs to a large family of cytokines of which the best characterized are IL-1α, IL-1β, IL-18 and IL-33 [13].IL-1α and IL-1β bind to their shared receptor, the type 1 IL-1 receptor (IL-1RT1), which subsequently dimerizes with the IL-1R receptor accessory protein (IL-1RAP) resulting in activation of NF-ĸB [14].The functions of IL-1α and IL-1β are similar in many aspects [13], however IL-1α has the capacity to distinguish between necrosis and atrophy and can be released from damaged cells to promote an immune response [15].Less is known about the involvement of IL-1α in gastric cancer progression, however IL-1α expression is increased in H. pylori associated pathology [16] and elevated IL-1α in gastric tumors has been associated with liver metastasis [17].
The IL-6 family member IL-11 is a pleiotrophic cytokine [7], and has elevated expression in association with gastric cancer development [18,19].IL-11 signals via the IL-11Rα/gp130 receptor complex to activate genes involved in proliferation, angiogenesis, inflammation and inhibition of apoptosis [20,21].gp130 757FF mice contain a knock-in mutation at Y757 of gp130, where a tyrosine has been substituted with phenylalanine, preventing both the phosphatase SHP2 and the negative regulator of STAT3, SOCS3 from binding and consequently resulting in chronic hyper-activation of STAT3 [21].gp130 757FF mice spontaneously develop distal (antral) stomach tumors, which phenocopy chronic H. pylori-induced pathology, particularly the key initiating event, pan-gastritis including an abundance of myeloid derived cells [20,21].Furthermore, gp130 757FF pathology is absolutely dependent on IL-11 signaling [19].IL-11 expression is increased in multiple mouse models of progressive gastric pathology [19], and in human gastric tumors [19,22].Additionally, IL-11 causes profound atrophy in the stomach [7], with a prominent downstream target being IL-1β [8].
The aim of our study was to dissect the individual contributions of oncogenic signaling mediated by gp130 and IL-1RT1 ligands in IL-11/STAT3 mediated gastric cancer development.In addition we evaluated the suitability of IL-1 as novel therapeutic target for gastric cancer.Here we show that IL-1RT1-mediated signaling is dispensable for antral tumor development in the gp130 757FF mouse, as well as IL-11-induced fundic atrophy and metaplasia.In fact, in the absence of IL-1RT1 signaling gastric pathology develops more rapidly and extensively in both the antrum and gastric cardia suggesting that IL-1RT1 signaling may minimize tumor burden by antagonizing IL-11/STAT3 mediated pathology.
We also isolated E-cadherin positive epithelial cells from wildtype and gp130 757FF mice and assessed IL-1α and IL-1β mRNA expression however we could not detect a signal for either cytokine (data not shown) suggesting the epithelium does not significantly contribute to increased IL-1α and IL-1β in gp130 757FF gastric lesions.

The loss of IL-1RT1 increases gp130 757FF gastric tumor burden
In order to assess the contribution of IL-1 signaling to tumor burden directly, gp130 757FF mice were crossed with IL-1RT1 -/-(receptor utilized by both IL-1α and IL-1β) mice.Stomachs were dissected and opened along the lesser curvature (Fig. 2A).As expected gp130 757FF mice develop tumor lesions in the antral stomach averaging 44.3 ± 2.3mm 2 .Surprisingly, gp130 757FF xIL-1RT1 -/-mice develop considerably larger antral lesions cumulatively measuring 60.6 ± 5.6mm 2 (Fig. 2B&2Cii).This data demonstrates that IL-1 signaling is not required for gp130 757FF tumorigenesis and that an absence of IL-1 signaling enhances antral tumor growth.
Histological assessment of gp130 757FF xIL-1RT1 -/- stomachs supports a cardiac gland origin for the proximal stomach lesions.In wildtype mice the cardia is typically restricted to a single gland directly adjacent to the forestomach (arrow in Fig. 3Ai&ii).The cells within this gland are cuboidal epithelial cells and resemble the mucous cells of antral glands, reinforced by the presence of acidic carbohydrates evident with Alcian blue staining (Fig. 3A iii).Histologically, parietal, chief and mucous neck cells are absent from the cardiac glands.Examination of the proximal stomach from a gp130 757FF xIL1RT1 -/- mouse demonstrates that the most significantly affected glands are the cardiac glands adjacent to the forestomach.These cardiac glands expand and become hyperplastic with associated inflammatory cells.Directly adjacent to the abnormal cardia are normal fundic glands (Fig. 3Aiv&v).With larger cardiac lesions the origin of the lesion is no longer as apparent since the entire proximal stomach is affected and proximal lesions become continuous with distal lesions (Fig. 3Bi&ii).Characteristics of the proximal lesions include near complete atrophy, foveolar hyperplasia and some dysplasia (Fig. 3Ci-iii), accompanied by a mix of mucous cells (Fig. 3Civ-vii) and moderate inflammation (Fig. 3Cviii&ix).
Previously, in a study assessing the effects of maximal gp130/STAT3 activation on the gastric mucosa, we reported that systemically administered IL-11 (20μg/ day) caused gastric atrophy and mucus metaplasia through STAT3 activation, and coincident with elevated IL-1β expression [7].Others have shown that overexpression of IL-1β in the mouse stomach results in tumor induction [4,8].Here we administered a significantly lower and more physiologically relevant dose of IL-11 (1μg/day) to wildtype and IL-1RT1 -/-mice to determine if IL-1RT1 signaling was required for IL-11 induced changes to the gastric mucosa.Histological assessment of the cardiac and fundic glands following administration of low concentrations of IL-11 demonstrated no obvious changes in either wildtype or IL-1RT1 -/-mice (Fig. 3D).Normal cardiac glands stained positive for acidic mucins are indicated in each image (Fig. 3Di-iv).

IL-1RT1 signaling does not affect IL-11 signaling in the stomach
To determine the requirement of IL-1 signaling for the pathological effects of IL-11 following its systemic administration, we measured the pSTAT3/STAT3 ratio in the proximal stomach of mice treated with 1μg IL-11/ day delivered by mini-osmotic pumps (MOPs).Wildtype (1.62 ± 0.20 fold) and IL-1RT1 -/-mice (1.54 ± 0.11 fold) had similar increases in STAT3 phosphorylation following IL-11 administration, suggesting that IL-1 signaling is not required for IL-11 signaling in the stomach (Fig. 4Di).We also tested the activation of STAT1, ERK1/2 and AKT.STAT1 activation could not be detected in any group, and the ratio of ERK1/2 and AKT phosphorylation compared to total protein was not altered (4Dii&iii).As IL-11 administration may promote endogenous expression of gp130 ligands, thereby confounding its effects, we measured the mRNA expression of endogenous gastric IL-6 and IL-11.IL-6 was decreased in IL-11 treated IL-1RT1 -/-mice compared to IL-11 treated wildtype mice (-3.5 ± 1.0 fold; Fig. 4E).The expression of IL-11 was not changed in IL-11 treated wildtype or IL-1RT1 -/-mice (Fig. 4F).These data show that the loss of IL-1 signaling does not influence IL-11-mediated STAT3 activation.However, consistent with previous reports [4], IL-1RT1 signaling is required for IL-6 expression during STAT3 induced gastric atrophy.

gp130 757FF xIL-1RT1 -/-cardiac lesions show a loss of differentiated gastric cell markers
Since gp130 757FF xIL-1RT1 -/-cardiac lesions showed extensive atrophic changes we compared the loss of specialized gastric cell types by assessing mRNA expression markers of differentiated fundic cell populations.gp130 757FF xIL-1RT1 -/-cardiac lesions showed a significant reduction in mRNA expression of markers for parietal cells (H/K ATPase α, H/K ATPase β & CCKBR), chief cells (Mist1), surface mucous cells (Muc5AC) and progenitor cells (BMP4&Lgr5; Supp Fig. 2).The metaplastic SPEM lineage (TFF2) was also reduced in gp130 757FF xIL-1RT1 -/-cardiac lesions however not significantly (Supp Fig. 2).Administration of low dose IL-11 to wildtype mice did not alter the expression of proximal stomach gastric cell markers, however IL-11 treatment of IL-1RT1 -/-mice led to reduced expression of markers of differentiation in a similar pattern to gp130 757FF xIL-1RT1 -/-mice (Supp Fig. 2).This data supports previous findings which demonstrate IL-11 to induce gastric atrophy [7], a process which appears to be accelerated in the absence of IL-1RT1 signaling as demonstrated by a decrease of mRNA markers of gastric epithelial cells (Supp Fig. 2).
To further assessed the effects of TNF-α on the gp130 757FF gastric tumor phenotype, gp130 757FF mice were crossed with TNF-α -/-mice.Macroscopic assessment of lesion size demonstrated that gp130 757FF xTNF-α -/- mice had considerably smaller distal tumors compared to gp130 FF mice (38.91 ± 5.46 and 70.16 ± 6.27 mm 2 respectively; Fig. 6Bi-iii).These data show that TNF-α promotes gp130 757FF tumor development, potentially through the recruitment of MDSCs that are able to restrict the anti-tumor response and are thus permissive for tumor growth.

DISCUSSION
IL-1α and IL-1β mRNA expression is upregulated in antral tumors of gp130 757FF mice, consistent with the view that IL-1RT1 ligands may contribute to the development of gastric antral tumors in this STAT3-dependent model, and may serve as novel therapeutic targets for gastric cancer.To test this directly gp130 757FF mice were crossed with IL-1RT1 -/-mice.Unexpectedly, in the absence of IL-1RT1 signaling there was a significant increase in the size of gp130 757FF gastric antral tumors, demonstrating that not only is activation of IL-1 signaling dispensable for tumorigenesis in the distal stomach of gp130 757FF mice, but that patent IL-1 signaling pathways can act to restrict tumor growth in the presence of cytokine-driven STAT3mediated gene transcription.
The negative influence of IL-1RT1 signaling was also observed in the gp130 757FF xIL-1RT1 -/-proximal stomach where a significant increase in both tumor incidence and size compared to gp130 757FF mice was observed.These proximal tumors originated from the gastric cardia at the limiting ridge near the gastro-esophageal and fore-stomach junction.gp130 757FF xIL-1RT1 -/-cardiac lesions were associated with elevated STAT3 activation, a modulated immune response, including an enrichment of MDSCs, and loss of specialized gastric cells.The clear positive correlation between STAT3 activation and cardiac lesions indicates that STAT3 activation is causative for cardiac tumor development in this model.Furthermore, increased STAT3 activation correlated with increased IL-11 but not IL-6 mRNA expression, suggesting that IL-11 is the main ligand driving gp130 activation in cardiac lesions, as we have previously demonstrated for the primary antral lesions in the gp130 757FF mouse model [19].Additionally, at this advanced stage of pathology the recruitment of immunocytes which express STAT3 inducing genes may also contribute to STAT3 activation.This alternative pathway is supported by our analysis of the mRNA expression of IL-23 and amphiregulin, both of which signal via STAT3 [29,30] and which were increased in gp130 757FF xIL-1RT1 -/-cardiac lesions.Nonetheless, IL-11 is absolutely required for gp130 757FF gastric tumorigenesis and therefore must be considered the primary initiator of cardiac lesions.
The role of IL-1RT1 signaling during IL-11induced STAT3 activation without tumor inducing gp130 mutations was also assessed.This was achieved by continuous, systemic, low dose IL-11 administration.IL-1RT1 -/-mice treated with IL-11 showed immune and gastric cell marker changes similar to those observed in gp130 757FF xIL-1RT1 -/-cardiac lesions, whereas wildtype mice treated with IL-11 did not.This demonstrates that IL-11 can promote gastric atrophy, consistent with our previous findings [7] and that IL-1 signaling antagonizes IL-11-dependent tumorigenesis.
Both wildtype and IL-1RT1 -/-mice treated with IL-11 had increased STAT3 activation in the proximal stomach and unchanged endogenous IL-11 expression.IL-6 expression on the other hand, was markedly reduced in IL-1RT1 -/-mice administered IL-11, suggesting that IL-6 expression is dependent on intact IL-1RT1 signaling.Quante et al (2012) [4] suggest that enhanced IL-6-induced STAT3 activation is crucial for the cardiac/ esophageal pathology that develops in IL-1β transgenic mice.However they did not measure IL-11 expression, a well-documented inducer of STAT3 phosphorylation whose expression is absolutely required in the gp130 757FF mouse [19] and is strongly upregulated in several other models of gastric tumorigenesis [19], including the cardiac lesions in gp130 757FF xIL-1RT1 -/-mice.
In general we observed the same cardiac pathology and gene expression pattern in mouse genetic models which have either exaggerated IL-1RT1 signaling [4], or ablated IL-1RT1 signaling (gp130 757FF xIL-1RT1 -/-).Therefore it seems that the common thread that links the cardiac pathology in both models is not intact IL-1RT1 signaling per se, but augmented expression of IL-6 family cytokines which hyperactivate gp130/STAT3 leading to MDSC recruitment.Together these outcomes highlight the importance of gp130 ligands and dysfunctional IL-1RT1 cytokines in driving upper gastric pathology.Supporting evidence comes from individuals carrying polymorphisms in the IL-1α, IL-1β and IL-1RN gene cluster that lead to elevated expression of IL-1 cytokines, and who are actually protected from gastro-eosphageal reflux disease and Barrett's esophagitis [31][32][33], suggesting that a delicate balance is at play, with expression of IL-1 cytokines in the proximal stomach and esophagus under some circumstances leading to pathology, and in others being protective.These findings in conjunction with the observed detrimental effects of ablated IL-1 signaling in IL-11/STAT3 mediated gastric pathology demonstrated here, suggest that IL-1 receptor blockade, or immunoneutralisation of ligands may not be suitable targets for preventing gastric cancer progression.Furthermore inhibiting these cytokines in humans may result in more aggressive tumors and possibly enhanced metastasis.
The observation that loss of IL-1RT1 signaling in the gp130 757FF mouse produces more pathology was unexpected considering that high expressing IL-1β polymorphic variants correlate with increased gastric cancer risk in Caucasians, and transgenic overexpression of IL-1β in mice results in severe gastric and esophageal pathology [4,8].However, while much of the current literature has focused on the pro-tumorigenic effects of IL-1β, particularly in gastric cancer, IL-1α, which was preferentially expressed in gp130 757FF xIL-1RT1 -/-cardiac lesions at the expense of IL-1β, has been poorly studied in comparison.Although these two cytokines share many functions, IL-1α has a novel role as an alarmin.Alarmins describe a class of multifunctional cytokines released by necrotic cells in response to infection or injury to promote an immune responses [34].As such, IL-1α is expressed constitutively in the nuclei of cells to be released and activated by cellular damage, thereby initiating an immediate inflammatory response [15,35,36].In mice, the loss of IL-1α, but not IL-1β, results in decreased survival rates in response to lethal endotoxemia [37], and impaired immune function during infection [38][39][40].Furthermore, the loss of IL-1 signaling in colitis in mice results in more severe colonic pathology suggesting IL-1 has the capacity to protect against mucosal injury [41,42].Many studies identify IL-1α as being a critical component of the early phase recruitment of immunocytes particularly macrophages and neutrophils [43][44][45].We suggest that in our model the absence of patent IL-1α alarmin function in the stomach due to IL-1RT1 deletion results in inappropriate activation of cellular infiltrate, ultimately resulting in chronic damage, and enhanced tumorigenesis.
The current dogma supports the view that IL-1β promotes enrichment of MDSCs in the stomach thus balancing the alarmin activity of IL-1α [46], and can lead to gastric tumour formation [8].Additionally, MDSC numbers are known to be an independent risk factor for disease progression in gastric and esophageal cancer [47,48].MDSCs can promote tumor immune tolerance through suppression of T cell responses leading to tumor expansion and disease progression [49,50].In the gastrointestinal tract, MDSC-induced tumor immune tolerance has been demonstrated to promote gastric and esophageal cancers [4,8,47,48], suggesting that enrichment of MDSCs directly correlates to gastric and esophageal disease outcomes.Here, we demonstrate that IL-11/STAT3 causes an enrichment of MDSCs independent of IL-RT1 signaling.This finding is consistent with published data from the lung and pancreas [51,52].Furthermore, IL-6 signaling, a potent STAT3 inducer, is required for IL-1βinduced gastric and esophageal pathology [4,8], indicating that STAT3 signaling is necessary for MDSC expansion and explains how both overexpression and ablated IL-1 signaling can induce cardiac pathology through STAT3 dependent MDSC enrichment.
An additional consequence of a deficiency in IL-1RT1 signaling in the gp130 757FF stomach is elevated TNFα, also known to promote the enrichment of MDSC populations though expression of S100A8 and S100A9 [52].One mechanism by which MDSCs may promote tumor growth is through expression of arginase-1, which in turn diminishes T cell mediated anti-tumor responses [48].Here we show elevated S100A8 and 9 as well as arginase-1, suggesting enhanced MDSC activity in stomach tumors deficient in IL-1RT1 signaling.Moreover the absence of TNFα in gp130 757FF mice results in smaller tumors, likely as a result of reduced numbers of MDSCs and an exaggerated pro-inflammatory environment.
Here we have demonstrated that IL-1RT1 -/- signaling is not required for IL-11/STAT3 mediated pathology, with genetic depletion of IL-1RT1 -/-resulting in a more severe tumor phenotype by promoting an antiinflammatory, pro-tumorigenic environment.We propose a mechanism involving IL-11-dependent recruitment of tumor promoting MDSCs to the stomach.Furthermore, these data provide additional evidence that gp130-induced STAT3 activation is required for cardiac and esophageal pathology, and may provide the basis for as a future therapeutic target for such pathologies.

Mice
All mice were on a C57Bl/6 background and were used at 12-14 weeks of age.Transgenic mice were genotyped by multiplex PCR as previously described [7,20].Ethical approval (#664A) was obtained from the animal ethics committee of the Murdoch Children's Research Institute.

Tissue preparation
Mouse stomachs were prepared and analysed as previously described [21].Briefly, the stomach was bisected, antrum and fundus from one half were frozen for TRIzol extraction, and the other half fixed in 4% paraformaldehyde in PBS.When obvious proximal (cardiac) stomach lesions were evident, the proximal stomach was further divided into affected and unaffected tissues.Spleens were also collected from MOP mice; one half was frozen for TRIzol extraction and the other fixed in 4% paraformaldehyde in PBS.

Macrophage isolation and analysis
To harvest peritoneal macrophages, 5 mL of HBSS (Sigma) containing 10 units/mL of heparin (Sigma) was injected into the peritoneal cavity (n = 5/group).Extracted cells were harvested, washed, and resuspended in complete RPMI then plated onto 6-well culture plates.Plates were incubated at 37°C for 10 minutes to allow macrophages to adhere.Cultures were then washed to remove nonmacrophage cells.Macrophages were incubated overnight at 37°C.The next day, macrophage cultures were treated with either saline or 100 ng LPS/mL in fresh media.Cells were harvest using TRIzol reagent (Life Technologies).

Quantitative RT-PCR
RNA was harvested using TRIzol reagent (Life Technologies).RNA (3μg) (n ≥ 10 animals/group) was reverse transcribed using Moloney murine leukemia virus reverse transcriptase (Promega) primed with oligo (dT).Quantitative RT-PCR (QRTPCR) primers were designed using PRIMER EXPRESS (Applied Biosystems) (Supplementary Table 1).SYBR green chemistry was used with rL32 as the internal reference gene.QRTPCR conditions were 95°C for 10 min, 40 cycles of 95°C for 15 sec and 60°C for 15 sec (Applied Biosystems AB7500).Results were analyzed using sequence detector software, relative fold differences were determined using the ΔΔCt method.

Immunoblotting
Proteins (n ≥ 10 animals/group) were prepared using TRIzol (Life Technologies) and 20μg of extract was subjected to SDS PAGE.Membranes were incubated with specific antibodies (Supplementary Table 2), peroxideconjugated secondary antibody and visualized by enhanced chemiluminescence (Amersham).Quantification was done using Quantity 1 software (Bio-Rad Laboratories) and ratios of phosphorylated: total protein determined from duplicate membranes.

IL-1β ELISA assay
TRIzol extracted proteins were diluted to 0.1% SDS using sterile PBS and IL-1β sandwich ELISA was performed as previously described [53,54].

Quantitative morphometry
Macroscopic: Images were captured on a Coolpix 4500 digital camera (Nikon Instruments, Melville, NY).Quantitative analysis of stomach lesions was done using ImageJ software for Windows v1.38 [55].Measurements were converted to millimetres after comparison with a calibrated graticule.
Microscopic: Images of H&E stomachs were taken along the length of the stomach using a Coolpix 4500 digital camera attached to a light microscope.Lengths and areas were manually traced on these images using ImageJ software (see above).Measurements were converted to millimetres after comparison with a calibrated graticule.
Spleen cell isolation: One third of the spleen was made into a single cell suspension.Red blood cells were lysed with Ammonium-Tris Chloride buffer for 5 minutes at room temperature.Cell suspensions were stained with fluorescently labeled antibodies similar to that of the stomach.
FACS analysis: Dead, autofluorescent and aggregated cells were gated out of on the basis of FSC, SSC and propidium iodide staining.Live cells were gated based on FSC v SSC plot and the number of events was recorded for the different cell types (Supplementary Table 3).The number of events for each cell type was compared to the total number of all live cell events and proportions of cell types were compared between treatment groups.

Statistical analysis
All data were expressed as mean ± SEM and statistical analysis was performed by one-way analysis of variance (ANOVA) and the appropriate parametric or nonparametric statistical test using Sigmastat (Jandel Scientific).All p values were derived and comparison was performed between the wildtype control group and specific treatment groups unless indicated.p values ≤ 0.05 were considered statistically significant.

Figure 2 :
Figure 2: Macroscopic analysis of pathology in 12 week old gp130 757FF xIL-1RT1 -/-mice compared to gp130 757FF mice.(A) A graphical representation of an intact mouse stomach (i) which has been dissected along the lesser curvature and pinned onto balsa (ii).(B) Macroscopic images of a gp130 757FF mouse stomach (i) and a gp130 757FF xIL-1RT1 -/-mouse stomach (ii).(C) Macroscopic analysis of lesion area in the proximal stomach (i) and distal (ii) and microscopic analysis of distal lesion height (iii).p-values are presented for statistically significant changes (p < 0.05).