IL-33-mediated mast cell activation promotes gastric cancer through macrophage mobilization

Eissmann, Moritz F.; Dijkstra, Christine; Jarnicki, Andrew; Phesse, Toby; Brunnberg, Jamina; Poh, Ashleigh R.; Etemadi, Nima; Tsantikos, Evelyn; Thiem, Stefan; Huntington, Nicholas D.; Hibbs, Margaret L.; Boussioutas, Alex; Grimbaldeston, Michele A.; Buchert, Michael; O’Donoghue, Robert J. J.; Masson, Frederick; Ernst, Matthias

doi:10.1038/s41467-019-10676-1

Download PDF

Article
Open access
Published: 21 June 2019

IL-33-mediated mast cell activation promotes gastric cancer through macrophage mobilization

Moritz F. Eissmann¹,
Christine Dijkstra¹,
Andrew Jarnicki¹⁰,
Toby Phesse ORCID: orcid.org/0000-0001-9568-4916¹^nAff6,
Jamina Brunnberg¹^nAff7,
Ashleigh R. Poh¹,
Nima Etemadi ORCID: orcid.org/0000-0002-7137-1373¹^nAff8,
Evelyn Tsantikos²,
Stefan Thiem¹,
Nicholas D. Huntington³,
Margaret L. Hibbs²,
Alex Boussioutas⁴,
Michele A. Grimbaldeston⁵^nAff9,
Michael Buchert¹,
Robert J. J. O’Donoghue¹^nAff10,
Frederick Masson¹^nAff11 &
…
Matthias Ernst ORCID: orcid.org/0000-0002-6399-1177¹

Nature Communications volume 10, Article number: 2735 (2019) Cite this article

10k Accesses
120 Citations
5 Altmetric
Metrics details

Subjects

Abstract

The contribution of mast cells in the microenvironment of solid malignancies remains controversial. Here we functionally assess the impact of tumor-adjacent, submucosal mast cell accumulation in murine and human intestinal-type gastric cancer. We find that genetic ablation or therapeutic inactivation of mast cells suppresses accumulation of tumor-associated macrophages, reduces tumor cell proliferation and angiogenesis, and diminishes tumor burden. Mast cells are activated by interleukin (IL)-33, an alarmin produced by the tumor epithelium in response to the inflammatory cytokine IL-11, which is required for the growth of gastric cancers in mice. Accordingly, ablation of the cognate IL-33 receptor St2 limits tumor growth, and reduces mast cell-dependent production and release of the macrophage-attracting factors Csf2, Ccl3, and Il6. Conversely, genetic or therapeutic macrophage depletion reduces tumor burden without affecting mast cell abundance. Therefore, tumor-derived IL-33 sustains a mast cell and macrophage-dependent signaling cascade that is amenable for the treatment of gastric cancer.

A single-cell atlas enables mapping of homeostatic cellular shifts in the adult human breast

Article Open access 28 March 2024

Austin D. Reed, Sara Pensa, … Walid T. Khaled

Human lung cancer harbors spatially organized stem-immunity hubs associated with response to immunotherapy

Article 19 March 2024

Jonathan H. Chen, Linda T. Nieman, … Nir Hacohen

SOX17 enables immune evasion of early colorectal adenomas and cancers

Article 28 February 2024

Norihiro Goto, Peter M. K. Westcott, … Ömer H. Yilmaz

Introduction

The interactions between cancer cells and their microenvironment can result in tumor progression as well as suppression and/or eradication of cancers¹. Besides extracellular matrix and stromal cells, the tumor microenvironment is composed of immune cells of the adaptive and innate immune system, with the latter comprising neutrophils, macrophages, mast cells, myeloid-derived suppressor cells, dendritic cells, innate lymphocytes, and natural killer cells. Often T-lymphocytes and tumor-associated macrophages (TAMs) account for the most abundant immune cell populations infiltrating established tumors^2,3.

Mast cells are long-lived secretory cells of hematopoietic origin that function as sentinels by responding to changes in their environment⁴. They respond to extrinsic signals through a multitude of cell surface receptors to secrete histamine and proteases from prestored sources in cytoplasmic granules⁵, alongside newly synthesized inflammatory mediators⁴. Accordingly, mast cells can be part of innate and adaptive immune responses and therefore contribute to various pathophysiological conditions⁶. In cancer, the presence of mast cells, both at the tumor periphery and in the tumor core can correlate with disease progression, increased metastasis and reduced survival of patients with melanoma⁷, prostate cancer⁸, pancreatic adenocarcinoma⁹, squamous cell carcinoma¹⁰, and gastric cancer^11,12. However in other solid malignancies, mast cells have been associated with more favorable outcomes^13,14,15. The conflicting nature of these correlative findings are reminiscent of contradictory functional observations in mouse models claiming that mast cells are required for pancreatic islets tumorigenesis¹⁶, while pancreatic adenocarcinoma occurs in a tumor microenvironment devoid of mast cells¹⁷.

Interleukin (IL)-33 is a danger-associated signal that can serve as a molecular alarmin when released upon necroptotic and necrotic cell death including death of cancer cells^18,19,20,21. However, cell death-independent IL-33 release can also occur^22,23. IL-33 signals through the heterodimeric ST2 receptor, encoded by Il1rl1 gene, which is constitutively expressed on the surface of some innate immune cells including mast cells²⁴, innate lymphoid cells type 2 (ILC2)²⁵, and regulatory T-cells (Treg)^26,27. IL-33/ST2 signaling is involved in triggering innate immune responses upon parasite and viral infections, and has been identified as an important mast cell activating factor^24,28 in the context of allergy²⁹. Furthermore, elevated IL-33 expression was associated with poor outcomes in patients with gliomas³⁰, ovarian³¹, as well as head and neck cancers³². However, predicting the outcome of IL-33/ST2 signaling in malignancies remains uncertain with both tumor promoting as well as tumor restricting activities being reported in knockout mouse models^33,34,35,36.

Here, we employ preclinically validated mouse models of gastric cancer and corresponding patient biopsies to functionally elucidate the role of mast cells during gastric tumorigenesis. Our genetic analysis reveals a linear signaling axis initiated by tumor epithelial-derived IL-33 that activates mast cells to produce a chemotactic cytokine expression signature. These factors promote the accumulation of TAMs, which in turn sustain tumor angiogenesis and growth in mice. In gastric cancer patients, this mast cell activation signature, alongside markers for tumor-associated macrophages, correlates with decreased patient survival. Our findings delineate an IL-33/mast cell/macrophage axis, which affords a clinical opportunity for the treatment of gastric cancer.

Results

Increased mast cell density in human intestinal-type gastric cancer and in corresponding mouse models

In order to characterize the role of mast cells in gastric cancer, we initially investigated the mast cell frequency in gp130^FF mice, a preclinically validated model for spontaneously occurring intestinal-type gastric cancer^37,38. Gastric tumors in gp130^FF mice, which harbor a knock in germline mutation in the shared IL-6/IL-11 receptor subunit gp130, arise from excessive IL-6/IL-11 dependent STAT3 activity; these tumors remain associated with chronic inflammation and immune cell infiltration. We used toluidine blue stains to quantify mast cells, and observed profound accumulation of mast cells in the submucosa of the gastric antrum in gp130^FF mice compared to wild-type mice. Importantly, mast cells in the submucosa adjacent to gastric tumors were more abundant than the mast cells in the normal antrum-associated submucosa of gp130^FF mice (Fig. 1a, b). In addition, mast cell numbers in gp130^FF; Stat3^+/− compound mutant mice, which neither develop gastric tumors nor the other pan-inflammatory changes observed in their Stat3-proficient gp130^FF; Stat3^+/+ littermates^39,40, remained comparable to those in wild-type mice.

To ascertain that submucosal mast cell accumulation correlated with tumorigenesis independently of the nature of the oncogenic driver mutation, we also assessed mast cell distribution in Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^fl/fl mice. In this model, conditional activation of a pan-glandular gastric epithelium-specific Cre-recombinase induces aberrant activation of phosphatidylinositol 3-kinase (PI3K), a pathway frequently mutated in human gastric cancer. Akin to our observations in gp130^FF mice, mast cells also accumulated in the submucosa of tumors in Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^fl/fl mice (Supplementary Fig. 1a, b). Strikingly, toluidine blue, alcian blue/safranin O, and May–Grunwald–Giemsa stains for granular content all consistently revealed that mast cells localized to the submucosal layers of the antrum and body of the stomach, rather than within the tumors (Supplementary Fig. 1c).

We next assessed the stromal abundance of mast cells in human biopsies collected from patients with either histopathologically classified diffuse or intestinal-type gastric cancer, or patients with preneoplastic intestinal metaplasia or chronic gastritis. Similar to our observation in mice, we observed a profound submucosal accumulation of toluidine-blue positive mast cells in intestinal-type gastric cancer. Strikingly, mast cells numbers were not elevated in the submucosa of patients with gastritis or intestinal metaplasia (Fig. 1c, d). Collectively, our results show increased mast cell abundance in the tumor-adjacent submucosa of intestinal-type gastric adenocarcinoma in human and mice, irrespective of the nature of the oncogenic driver mutation(s).

Mast cell deficiency diminishes gp130 ^FF-driven tumor growth

Having established a reproducible correlation between submucosal mast cell density and gastric cancer in humans and mice, we next determined the functional role of mast cells during tumor initiation and progression. For this we exploited the gp130^FF model because of its high disease concordance with the human GC subtype, reproducible penetrance, onset, and progression of gastric tumorigenesis.

In order to establish whether submucosal mast cell accumulation preceded tumorigenesis, we analyzed stomach sections of 4-week-old gp130^FF mice prior to onset of tumor formation. We detected increased mast cell density in the gastric submucosa of these tumor-free gp130^FF mice when compared to the submucosa of age-matched WT littermates (Supplementary Fig. 2a). To functionally determine whether mast cells indeed promote gastric tumor growth, we crossed gp130^FF mice with mast cell-deficient C57BL/6 c-Kit^W-sh/W-sh mice carrying an inversion of the 5′-regulatory region of the c-Kit gene that results in hypomorphic expression of the corresponding stem cell factor receptor protein. Accordingly, c-Kit^W-sh/W-sh mice exhibit a substantial reduction of mast cells in all tissues^41,42,43, but do not suffer from anemia, sterility and lethality associated with complete c-Kit deficiency⁴¹. Likewise, we did not observe differences in the peripheral blood composition between gp130^FF and gp130^FF; c-Kit^W-sh/W-sh compound mutants and we confirmed that EpCAM positive tumor cells lack c-Kit expression (Supplementary Figs. 2b, c). However, gp130^FF; c-Kit^W-sh/W-sh compound mice had significantly smaller and fewer tumors than their age and sex-matched gp130^FF littermates (Fig. 2a–c). This reduction in tumor burden coincided with reduced tumor cell proliferation rather than increased apoptosis, and was associated with a reduction of CD31⁺ endothelial cells (Fig. 2d and Supplementary Fig. 2d).

To exclude that the anticancer effect of the c-Kit^W-sh hypomorphic allele was not in part mediated by hematopoietic cells other than mast cells, we generated mast cell-deficient gp130^FF; Cpa3-Cre; Mcl1^fl/fl gastric cancer mice, where mast cell-specific carboxypeptidase A3 (Cpa3) promoter driven Cre recombinase activity leads to the deletion of Mcl1 prosurvival gene. As a consequence Cpa3-Cre; Mcl1^fl/fl mice retain less than 10% mast cells and have reduced numbers of basophils, while all other hematopoietic cell populations remain unaffected⁴⁴. We confirmed that gp130^FF; Cpa3-Cre; Mcl1^fl/fl mutant mice lack mast cells in their stomachs, while their gp130^FF; Cpa3-Cre; Mcl1^+/+ littermates display normal mast cell density (Supplementary Fig. 2e). Importantly, mast cell-deficient gp130^FF;Cpa3-Cre; Mcl1^fl/fl mice had significantly reduced tumor mass and tumor number compared to their mast cell-proficient controls (Fig. 2e, f), and this observation coincided with reduced angiogenic vessel density in the tumors of gp130^FF; Cpa3-Cre; Mcl1^fl/fl mice (Fig. 2g).

To assess whether therapeutic mast cell manipulation could reduce the burden of established tumors, we exploited sodium cromoglycate (cromolyn) as a blocking agent for mast cell degranulation in patients. We treated tumor-bearing gp130^FF mice for 6 weeks with cromolyn, which significantly decreased tumor burden when compared to vehicle-treated mice (Supplementary Fig. 2f and Fig. 2h). Akin to our observations genetic observations in gp130^FF; c-Kit^W-sh/W-sh and gp130^FF; Cpa3-Cre; Mcl1^fl/fl mice, cromolyn treatment of gp130^FF mice also reduced macrophage accumulation, tumor angiogenesis and proliferation (Supplementary Fig. 2g). These findings demonstrate the importance of mast cells and their released products for gastric tumor growth and suggest that mast cells promote tumor proliferation through mechanisms involving angiogenesis.

In the gp130^FF mutant mice all cells harbor the gp130^F mutation. When expressed, the gp130^F mutant protein increases Stat3 signaling in response to IL-6 family cytokines. Because mast cells express the gp130 coreceptor and can respond to IL-6 family cytokines^45,46, we next excluded the possibility that the gp130^F allele in mast cells may augment their tumor-promoting ability. For this, we reconstituted the hematopoietic compartment of 6-week-old lethally irradiated gp130^FF hosts with the bone marrow from WT (gp130^+/+) or gp130^FF mice (Supplementary Fig. 2h). Indeed, WT → gp130^FF and gp130^FF→ gp130^FF bone marrow chimeras showed not only similar mast cell frequencies in the gastric submucosa but also comparable tumor burden, and macrophage infiltration in the submucosa and tumor core (Supplementary Fig. 2i). These results indicate that the systemic presence of the gp130^FF mutation did not skew hematopoietic cells towards a tumor-promoting role, nor did it increase the mast cell abundance within the gastric tumor microenvironment.

Macrophages are reduced in gastric tumors in a mast cell-dependent manner

Because we observed mast cells outside of the tumor cores, we surmised that tumor-promoting effect of mast cells might occur indirectly by affecting the composition and/or function of tumor infiltrating immune cells. We therefore enumerated tumor-free and tumor-containing stomach sections collected from gp130^FF and mast cell-deficient gp130^FF; c-Kit^W-sh/W-sh mice for the F4/80 macrophage, CD3 T lymphocyte and B220 B-lymphocyte markers. While we observed a 30% reduction in macrophage frequency in the submucosa and cores of tumors from gp130^FF; c-Kit^W-sh/W-sh mice, the lymphocyte (CD3⁺), Treg (Foxp3⁺), and CD8⁺ T cell frequencies remained unchanged (Fig. 3a, b). Reduced macrophage abundance within the tumor core and submucosa was also observed in the mast cell-deficient gp130^FF; Cpa3-Cre; Mcl1^fl/fl mice (Fig. 3c) and cromolyn-treated gp130^FF mice (Supplementary Fig. 2g). Because genetic or pharmacological interference with mast cells not only inhibited tumor growth but concomitantly also reduced the abundance of tumor-associated macrophages, we postulate that a hierarchical mast cell—macrophage relationship underpins the growth of gastric cancer.

Macrophage infiltration coincides with gastric tumorigenesis

To ascertain that the accumulation of tumor-associated macrophages was a common occurrence for intestinal-type gastric cancers, we assessed macrophages density in the submucosa and mucosa of tumor-bearing gp130^FF and Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^fl/fl mice. We noted an increased abundance of macrophages in the submucosa and the mucosa of either model (Fig. 3d, e and Supplementary 3a, b), consistent with the increased abundance of submucosal mast cells in both models (Fig. 1a, b and Supplementary Fig. 1a, b). To establish whether the increased macrophage accumulation occurs prior to tumor formation, akin to our observations with mast cells, we analyzed naïve gp130^FF mice at 4 weeks of age. In these mice macrophage numbers were increased in the submucosal layers, but remained similar between the precancerous mucosa of gp130^FF and of WT mice (Supplementary Fig. 3c, d). Collectively, these findings positively correlate macrophage numbers with gastric tumor growth and suggest a potential macrophage contribution to tumor formation and progression.

Macrophages are required for gastric tumorigenesis in gp130 ^FF mice

Macrophages have been established as drivers of tumor growth and cancer therapy targets for several tumor entities^47,48. To formally investigate a functional contribution of macrophages to gastric tumorigenesis, we generated macrophage-deficient gp130^FF; Csf1r^−/− mice, because expression of the Csf1r gene and its corresponding receptor protein for macrophage colony-stimulating factor is required for maturation of bone marrow-derived macrophages (BMDM) from their monocytic precursors. Strikingly, the few surviving gp130^FF; Csf1r^−/− mice remained completely free of gastric tumors (Fig. 4a, b). The submucosal mast cell density was not altered in the macrophage-deficient animals (Fig. 4c). We next confirmed that this phenotype reversion was not related to the runted phenotype of gp130^FF; Csf1r^−/− mice associated with their feeding requirement for mashed chow to overcome the Csf1R deficiency-dependent absence of teeth. We therefore depleted macrophages in adult tumor-bearing gp130^FF mice by systemic administration of clodronate-loaded liposomes (clodrosomes) for 6 consecutive weeks and observed a significant reduction of tumor burden when compared to the vehicle control cohort (Fig. 4d). Although toluidine-staining confirmed that clodrosome treatment did not reduce the abundance of submucosal mast cells, gastric tumors and adjacent submucosa from the clodrosome cohort revealed reduced density of CD31-positive microvessels (Fig. 4e and Supplementary Fig. 4a).

The reduced tumor-burden and microvessel density observed in the clodronate cohort were similar to our findings from the mast cell depletion experiments. We next aimed to exclude that they relate to off-target-effects associated with bisphosphonate (i.e., clodronate) administration. We therefore treated tumor-bearing gp130^FF mice with the small molecule Csfr1/c-kit/Flt3 kinase inhibitor PLX3397 (Pexidartinib). We supplied PLX3397 in the food to tumor-bearing gp130^FF mice for 4 weeks (acute therapeutic) and continued a cohort for a further 4 weeks without PLX3397 treatment (follow-up) (Supplementary Fig. 4b). Acute PLX3397 treatment revealed a significant reduction in tumor burden, although this antitumoral effect was reversible as the tumor burden increased once PLX3397 treatment ceased in the follow-up cohort (Fig. 5a). In the acute treatment cohort, we confirmed by immunohistochemical staining a significant reduction of F4/80-positive macrophages accompanied by reduced density of CD31-positive angiogenic vessels and increased staining for tumor-hypoxia (Fig. 5b and Supplementary Fig. 4c). However, owing to the specificity of PLX3397, we also observed a decrease of tumor-adjacent submucosal mast cells. All these parameters reverted back to their pretreatment state following a drug-free period in the follow-up cohort (Supplementary Fig. 4c).

Based on our proposed role for macrophages to mediate the tumor promoting effect of mast cells, we hypothesized that the tumor-associated macrophages in the gastric lesions of gp130^FF mice should have a tumor-promoting alternatively activated endotype. We therefore compared expression of macrophage endotype markers between naïve BMDM and tumor-associated macrophages, and used standard in vitro polarization protocols for BMDM to generate a reference population for alternative activation. For the tumor-associated macrophage populations, this analysis revealed high expression of the alternative activation markers Arg1, Fizz1, and Mrc1, which coincided with reduced mRNA levels of the classically activated marker Nos2 (Fig. 5c). Consistent with the pronounced presence of microvasculature in tumors of these gp130^FF mice, we also recorded elevated expression of angiogenic Vegfa in the tumor-associated macrophage population when compared to the naïve BMDM population from the same mice (Fig. 5c). Given the alternative activated endotype of tumor-associated macrophages, our data collectively suggests that their abundance is likely to be a rate-limiting factor for the establishment of tumor microvasculature and the growth of gastric tumors.

Tumor-derived IL-33 activates mast cell secretion and expression of macrophage-attracting chemokines

Although our observations thus far suggested a functional mast cell–macrophage relationship, it remained to determine the molecular network underpinning this hierarchy including how its activation could occur. We therefore focused on IL-33 as a potential tumor-derived activator of mast cells, because IL-33 was reported to serve as an alarmin that is generated and released by necrotic cells, and that triggers mast cell degranulation⁴⁹. Indeed, we detected increased expression of Il33 in the EpCam⁺/CD45⁻ epithelial components of the tumors in gp130^FF mice with a reciprocal expression of the Il1rl1 gene in the hematopoietic CD45⁺/EpCam⁻ compartment of the antrum, which encodes the cognate St2 receptor subunit for IL-33 (Fig. 6a). In line with the reported storage of IL-33 protein in vesicles, we detected increased levels of Il33 mRNA and IL-33 protein in whole tumor tissue (Supplementary Figs. 5a, b). Immunofluorescence staining revealed that IL-33 protein localized most prominently to the peripheral regions of the epithelial components of the tumors (Fig. 6b) coinciding with the location of the highest level of cellular turn-over³⁷.

To assess whether tumor-associated mast cells could respond to IL-33, we FACS purified the c-Kit⁺/F_cεR1⁺ mast cells from the submucosal antrum of WT and gp130^FF mice using stomachs from gp130^FF; c-Kit^W-sh/W-sh mice as a specificity control for our purification strategy (Supplementary Fig. 5c). We confirmed the homogeneity of the c-Kit⁺/F_cεR1⁺ mast cell population by toluidine blue staining (Supplementary Fig. 5d) and showed prominent expression of the St2 receptor in the c-Kit⁺/F_cεR1⁺ mast cells isolated from the stomach of gp130^FF mice (Supplementary Fig. 5e). We then stimulated this cell population for 3 h with IL-33 and detected by the multiplex cytokine assay increased amounts of IL-6 and IL-13 and to lesser extent of IL-1α, IL-4, alongside elevated secretion of the macrophage growth and attracting factors Gm-csf, Mcp1, Mip-1α, and Mip-1β (Supplementary Fig. 5f and Fig. 6c). Likewise, qPCR analysis of the IL-33 stimulated mast cells also revealed increased expression of Ccl2 (encoding Mcp1 protein), Ccl3 (encoding Mip-1α protein), and Ccl7 (Fig. 6d), which collectively exert chemotactic activity towards macrophages and neutrophils^50,51,52. Also, IL-33-stimulation induced elevated expression of angiogenic growth factor gene Vegfa (Fig. 6d). In order to confirm the chemotactic activity of the mast cell produced factors on gp130^FF macrophages, we performed migration assays with BMDM. Indeed, Mcp1 and Mip-1α alone induced migration significantly (Supplementary Fig. 5g).

Because gp130^FF-driven gastric tumorigenesis has an absolute requirement for IL-11 signaling³⁸, we predicted that IL-11 could stimulate IL-33 expression in tumor-derived epithelium. To test this, we exposed epithelial organoids grown from gp130^FF-tumors to recombinant IL-11 in vitro. While organoid growth and morphology were not altered (Fig. 6e and Supplementary Fig. 5h), we observed a marked increase in Il33 expression upon IL-11 stimulation, which was comparable to the extent of induction of the bona fide IL-11/Stat3-target gene Socs3 (Fig. 6f). Additionally, we conducted a chromatin immunoprecipitation (ChIP) experiment on DNA from gastric tumors of gp130^FF mice 60 min after acute systemic administration of IL-11. The ChIP experiment revealed a Stat3-specific binding peak on chromosome 19, which maps to the 5′-region of the Il33 gene and which contains a Stat3-binding site motif (Supplementary Fig. 5i, j). Collectively, these data indicate that tumor-derived IL-11 induces Il33 gene expression in gastric tumor cells, which triggers the secretion as well as the de novo synthesis of cytokines and chemokines by mast cells to produce a tumorigenic immune environment.

IL-33/St2 signaling deficiency decreases gp130 ^FF-mediated gastric tumor growth

To validate a functional contribution of IL-33 signaling as a master regulator of mast cell activation in the gastric tumor microenvironment, we generated IL-33 signaling deficient gp130^FF; St2^−/− mice. Compared to their St2^+/+ and St2^+/⁻-proficient compound mutant littermates, gp130^FF; St2⁻^/− mice harbored a significantly lower overall tumor burden at 100 days of age (Fig. 7a), which arose from reduced tumor growth rather than reduced tumor incidence (Fig. 7b). These observations coincided with a decrease in mast cell numbers in the tumor-adjacent submucosal layers and reduced density of F4/80⁺ macrophages in the tumors of gp130^FF; St2^−/⁻ mice as well as reduced abundance of CD31-positive microvessels (Fig. 7c).

We next confirmed that c-Kit⁺/F_cεR1⁺ mast cells isolated from gp130^FF; St2^−/− mice showed significantly decreased expression of the macrophage recruiting and inflammatory mediators Csf2 (encoding Gm-csf), Ccl3, and Il6 (Fig. 7d). Because IL-33 signaling is involved in expansion and activation of ILC2 and Tregs, which could also afford a mechanism for IL-33 to promote gastric tumorigenesis in gp130^FF mice, we assessed the relative abundance of ILC2 and Tregs among all cells harboring the pan-hematopoietic surface marker CD45. We observed comparable frequency of these cells between unaffected antrum and tumors of gp130^FF and gp130^FF; St2^−/− mice. Moreover, less than 30% of all tumor-associated ILC2 cells, and less than 15% of all Tregs, expressed the St2 receptor (Fig. 7e, f).

To investigate the involvement of mast cells in the antitumoral effect of St2-deficiency, we established bone marrow-derived mast cells (BMMC) from gp130^FF; ST2^−/− and gp130^FF; St2^+/+ mice, confirmed their purity (Supplementary Figs. 6a–c) and performed adaptive transfer of BMMC into gp130^FF; St2^−/− hosts (Supplementary Fig. 6d). Reconstitution with St2⁺ wild-type BMMC (gp130^FF; St2^+/+) increased tumor burden compared to littermates injected with gp130^FF; St2^−/− BMMCs (Fig. 7g and Supplementary Fig. 6e). At the time point of tumor analysis 8% of the gastric submucosal mast cells stained positive for St2 expression and therefore represent the transplanted BMMC mast cells, which homed to the gastric tumor site and compete with the endogenous ST2^−/− mast cell compartment (Supplementary Fig. 6f).

These observations further support a mechanism by which IL-33 promotes gastric tumorigenesis through mast cell activation rather than by promoting ILC2 or Treg accumulation.

IL-33/mast cells-activity gene expression signature predicts poor survival for human intestinal-type gastric cancer

In order to translate our preclinical findings in mice to a relevant setting in humans, we hypothesized that the extent of mast cell activation might provide retrospective prediction of patient survival. Based on the mast cell activation gene expression signature derived from our in vitro mast cell stimulation with IL-33, we analyzed a publically available data set (DErrico Gastric data set; GEO: GSE13911) for differential expression of the mast cell signature comprising CCL2, CCL3, CCL4, IL1a, IL4, IL6, IL13, CSF2, and CCL7. With the exception of IL-4 and IL-13, we found that all other factors were at least 1.6-fold increase in the cancer samples (p ≤ 0.05) (Fig. 8a). We therefore assigned the differentially expressed factors (i.e., CCL2, CCL3, CCL4, IL1a, IL6, CSF2, and CCL7) to an IL-33/mast cell activation gene expression signature for a Kaplan–Meier survival analysis (KMplot.com)⁵³. Remarkably, we found a lower overall survival for intestinal-type gastric cancer patients with high expression of the entire signature (HR = 1.61; logRank P = 0.043) (Fig. 8b). Similar to our IL-33/mast cell activation gene expression signature, a trend for better survival probability was found for high expression of a classical mast cell marker gene signature consisting of KIT, FCER1G, and HDC (encoding Histidine Decarboxylase) (HR = 1.42; logRank P = 0.067) (Supplementary Fig. 7a) as well as for ST2-receptor encoding gene IL1RL1 (HR = 1.57; logRank P = 0.049) (Supplementary Fig. 7b). In accordance with our mast cell–macrophage hierarchy model, Kaplan–Meier survival analysis for macrophages revealed a survival disadvantage for intestinal-type gastric cancer patients with high expression of an alternatively activated macrophage gene signature (Fig. 8c). On the other hand, high expression of a classically activated gene signature correlates with a higher survival probability (Fig. 8d). Our preclinical data argue strongly for a tumor promoting role of a signaling axis emanating from IL-11-induced, tumor cell-derived IL-33 and the subsequent hierarchical activation cascade of mast cells and tumor-associated macrophages. Collectively, these innate immune cells enable the establishment of the microvasculature required for gastric tumor growth and may provide therapeutically actionable targets (Fig. 8e).

Discussion

Although conflicting data indicate a dichotomous role of mast cells to either restrict or support tumorigenesis, compelling evidence implicates mast cells in promoting progression and metastasis in many solid malignancies. Retrospective studies on human gastric cancers have suggested an involvement of tumor-adjacent submucosal mast cells to late stage disease⁵⁴ and metastasis formation¹². However, the molecular mechanism by which submucosal mast cells contribute to gastric tumorigenesis, and indeed early stage tumorigenesis, is not understood. Here we demonstrate that submucosal accumulation of mast cells in two independent murine models of intestinal-type gastric cancer functionally contributes to tumorigenesis and the maintenance of established tumors and that this coincides with the presence of alternatively activated protumorigenic tumor-associated macrophages. We provide evidence that tumor cell-derived IL-33 stands on the apex of a cascade by which mast cells, through recruitment of tumor-associated macrophages and their support of a vascular network, ensures the growth and maintenance of gastric tumors. Indeed, ablating mast cells or macrophages in tumor-bearing mice either individually with cromolyn or clodrosomes, or together with PLX3397, were associated with a vascular collapse and tumor hypoxia. Importantly, we support the relevance of our functional findings in mice with a correlation in humans where a mast cell activation signature associates with poor patient outcome for intestinal-type gastric cancer.

Within the tumor microenvironment, myeloid cells are among the most important innate immune cells that promote tumor formation and restrict the effect of many forms of therapy, most notable immunotherapy⁵⁵. In particular the plasticity of tumor-associated macrophages has attracted interest as therapeutic opportunities to limit the angiogenic and immune suppressive functions conferred by their alternatively activated endotype³. On the other hand, mast cells have been proposed to be an abundant source of VEGF, TGFβ, and other angiogenic factors, as well as matrix metalloproteinase 9 and related stromal remodeling enzymes⁵⁶. Indeed, mast cells are one of the earliest cell types recruited into the tumor microenvironment^11,57, and progression of gastric cancer correlates with both the accumulation of chymase-positive mast cells and increased microvascular density^11,58. Functionally, these observations are mirrored by an angiogenic switch proposed to be orchestrated by mast cells during early stages of pancreatic cancer in mice required for the extensive vascular elaboration to prevent hypoxia during expansion of tumors¹⁶. Surprisingly, the authors in this early study had argued against a role for macrophages in mast cell-dependent induction of early tumor angiogenesis, based on their observations that macrophage recruitment to pancreatic islets remained unaffected by mast cell deficiency. Indeed, our observations that IL-33 result in the induction of Vegfa expression in isolated mast cells supports this view and this is corroborated by reduced microvessel density in tumors of gp130^FF; St2^−/− mice. However, our observations presented here following either genetic ablation of mast cells or macrophages, or their respective pharmacological inhibition by cromolyn and clodronate or PLX3397 strongly suggest that the endothelial cell support provided by mast cells also depends in part on the recruitment of macrophages.

Because we functionally link tumor growth to the presence of mast cells in two independent genetic models, it is highly unlikely that this arises from independent serendipitous effects of the two models other than mast cells. Thus, it is highly likely that the incompletely understood mechanisms by which cromolyn blocks mast cell degranulation mediated the anti-tumor effect observed here, rather than cromolyn also affecting monocytes, the production of the neutrophil product myeloperoxidase, and circulating growth⁵⁹. Akin to tumor-associated macrophages being induced by tumor cells to adopt an alternatively activated endotype optimized for wound healing, neoplastic epithelia appears to also corrupt the wound healing function of mast cells to facilitate tumor growth⁵⁵. Accordingly, tumor cell-specific overexpression of the master-regulator c-Myc not only results in epithelial hyper-proliferation (when coinciding with increased expression of BH3 survival proteins) but also in expression of chemo-attracting CCL2, CCL3, and CCL5 and concomitant accumulation of mast cells¹⁶. Here we identify a second mechanism that depends on the IL-1 cytokine family member IL-33, which acts as a chief regulator of innate immunity and inflammation, and maintains epithelial barrier functions of the intestine. Indeed, IL-33 serves a bona fide alarmin and is released by damaged epithelium as an endogenous danger signal to activate innate immune responses²⁸. Accordingly, mast cells and ILC2s respond through the IL-33/ST2 axis to the damaged intestinal mucosa arising from acute injury or infection. Therefore mast cells are found at sites of epithelial recovery⁶⁰, and ILC2 cells release IL-13 to stimulate the intestinal stem cell compartment, respectively⁶¹. Intriguingly, our experimental lines of evidence from tumors of gp130^FF; St2^−/− mice show that their reduced tumor burden correlates with decreased mast cell frequency rather than a change in ILC2 abundance. While not ruling out minor contributions by ILC2, regulatory T-cells or other IL-33 responsive cells, the increased tumor mass in gp130^FF; St2^−/− mice following adoptive transfer of St2-proficient mast cells, provides definitive functional evidence that mast cells act as the major cell population through which IL-33 promotes tumor growth. Accordingly tumor growth of gp130^FF mice was susceptible to the highly mast cell-selective cellular deficiency in gp130^FF; Cpa3-Cre; Mcl1^fl/fl mice, or to cromolyn-dependent inhibition of mast cell degranulation. Meanwhile, and consistent with the capacity of IL-33 to promote tissue repair and wound healing, IL-33 can also mediate metaplasia of the gastric epithelium⁶².

Although epithelial cells of barrier tissues exposed to the environment are major sources of IL-33, and IL-33 expression is further increased during inflammation and in the tumor epithelium⁶³, the exact nature of the factors that induce IL-33 expression and its release remain unclear. For instance TNF, IL-1β, TGFβ, prostaglandins, lipopolysaccharides, and other pathogen-associated molecular patterns have all been suggested as drivers for IL-33 expression during intestinal adenomatous polyposis^64,65,66. Here, we provide evidence for IL-33 expression in gastric tumor epithelium to also be stimulated by an IL-11/Stat3 signaling cascade, which we and others have identified as an absolute requirement for effective growth of gastrointestinal tumors. However, at this stage we can only speculate that subepithelial myofibroblasts and possibly tumor-associated endothelial cells, which have been proposed as sources of IL-11⁶⁷, may also contribute to epithelial expression of IL-33. Furthermore, because mast cells express gp130, we could not formally exclude the possibility that the presence of the hypermorphic gp130^F allele in mast cells may promote their proliferation and survival⁴⁶, as this receptor allele augments Stat3 signaling in response to IL-6 family cytokines. However, we noted similar submucosal mast cell accumulation adjacent to tumors in gp130^FF and in Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^fl/fl mice, where the latter cohorts express WT gp130. Likewise, reconstitution of gp130^FF mice with WT bone marrow did not affect mast cell numbers and tumor burden, as would be expected if the gp130^F allele augmented mast cell activity.

Therapeutic inhibition of IL-33 signaling is gaining interest with anti-ST2 antibodies being trialed for asthma. Given the dichotomous activity of IL-33 in tumor biology, timing of anti-IL-33 signaling therapy will be crucial. In the context of gastric cancer, mast cells may not only promote metastasis but also modulate the immunosuppressive tumor environment through the release of IL-17⁶⁸. Thus, IL-33 signaling provides a target to restrict the tumor promoting activities of the myeloid compartment and may ultimately enable rational combination therapies to alleviate the activity by which myeloid cells limit the antitumor immune response. The linear signaling cascade identified here comprising IL-11/IL-33/mast cells/macrophages/tumor cells should provide complementary molecular and cellular targets for the development of improved cancer therapies.

Methods

Study approval

All animal studies were conducted in accordance with all relevant ethical regulations for animal testing and research including the Australian code for the care and use of animals for scientific purposes. All animal studies were approved by the Animal Ethics Committee of the Ludwig Institute for Cancer Research, the Walter and Eliza Hall Institute of Medical Research, or Austin Health.

We have complied with all relevant ethical regulations for work with human participants. Collection and usage of human gastric cancer tissues was approved by the Peter MacCullum Cancer Center Ethics Committee and informed consent was obtained from all subjects.

Mice

Knockin mice (gp130^Y757F/Y757F, alias gp130^FF mice), the compound mutants strains gp130^FF; Stat3^+/−^39,40 and Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^lox/+ as well as wild-type control mice were bred on a mixed C57B6 × 129/Sv background. gp130^FF; c-Kit^W-sh/W-sh⁴¹, gp130^FF; Csf1r^−/−⁶⁹, and gp130^FF; St2^−/− mice⁷⁰, knockout of the Il-33 receptor (encoding gene name: Il1rl1), gp130^FF; Cpa3-Cre; Mcl1^lox/lox⁴⁴ (mast cell-deficient gp130^FF compound mutant mice) were maintained on a C57B6 background.

Tg(Tff1-CreERT2); Pik3ca^H1047R/+; Pten^lox/+ compound mutant mice were established, where mutant allele induction is induced by the Tff1-CreERT2 transgene and is specific for the gastric epithelium of the stomach⁷¹.

gp130^FF mice used in the PLX-treatment experiment were propagated on a C57B6 background.

Cohoused, age- and gender-matched littermates were utilized for all experiments. All strains were housed under specific pathogen–free conditions.

Human gastric cancer tissue

Human gastric cancer tissue micro arrays (TMA) were used to identify toluidine blue-stained mast cells in gastric disease. TMA were established previously⁷². In short, formalin-fixed paraffin-embedded gastric tissue samples we selected, pathology was confirmed by independent pathology review and 0.6 mm punches were reembedded to generate the TMA. Usage of human gastric cancer tissues was approved by the Peter MacCullum Cancer Center Ethics Committee and informed consent was obtained from all subjects.

Inhibitors and treatment regimes

Cromolyn sodium salt (referred to as cromolyn; Sigma–Aldrich, Cat# C0399) is a clinically approved mast cell-degranulation inhibitor⁷³. Cromolyn was dissolved in phosphate buffered saline (PBS) and administered at a concentration of 75 mg/kg body weight via the intraperitoneal route (100 µl) three times per week for 6 consecutive weeks. Control cohorts received PBS.

Clodronate, formulated as liposome-loaded clodrosomes (containing 5 mg clodronate per 1 ml of clodrosome suspension), was injected intraperitoneal as 50 µl clodrosomes suspension twice per week for 6 consecutive weeks. Liposomes containing suspension was injected as placebo in control mice.

The csf1r/c-kit/Flt3-specific inhibitor PLX3397 (Plexxicon), supplied at 800 mg PLX3397 per kg of chow, was given to mice for 4 consecutive weeks represents an estimated dose of 100 mg per kg body weight daily. Vehicle cohorts received unmanipulated chow. For the follow-up cohorts, mice were kept for 4 weeks on unmanipulated chow following their 4-week-treatment period.

Tissue preparation and processing

Tumors and adjacent antral tissues were resected and weighed and then snap-frozen for RNA or protein isolation. Entire stomachs were removed and fixed in 10% neutral buffer formalin for histological analysis³⁹.

Histological and immunohistochemical analysis

Hematoxylin-Eosin, toluidine blue, Alcian blue, Safranin O, and May-Grünwald-Giemsa staining of formalin-fixed paraffin-embedded stomach slides were performed according to theory and practice of histological technique from JD Bancroft⁷⁴.

For in vivo assessment of proliferation by anti-BrdU staining, tissues were collected 2 h after i.p. injection of 50 mg/kg BrdU (Amersham Biosciences, GE Healthcare). Stainings for apoptosis (Cell Death Detection Kit, Roche) and tissue hypoxia (60 mg/kg HP1 i.p. injection 30 min prior to tissue collection, detection with Hypoxyprobe-1 Kit, from Hypoxyprobe Inc., USA) were performed according to manufacturers’ instructions.

All other immunostaining procedures were conducted as follows. For antigen-retrieval, paraffin-embedded sections were either heated in citrate buffer in a microwave pressure cooker (pH 6 for 15 min) or were incubated in 0.1% trypsin, 3% acetic acid solution at 37° for 10 min. Sections were then blocked in 10% (v/v) normal goat serum for 1 h at room temperature. Primary antibodies were diluted in 10% (v/v) normal goat serum and incubated overnight at 4 ^oC in a humidified chamber. Biotin-labeled secondary antibodies from the Avidin Biotin Complex ABC-kit (Vector Laboratories) were used according to the manufacturer’s instructions. Visualization was achieved using 3,3-Diaminobenzine (DAB, DAKO). Images were generated with Aperio ImageScope v11.2.0.780 software. Primary antibodies used: rabbit anti-mouse CD31 (1:200 dilution, Abcam, Cat# 28364, RRID: AB_726362), rat anti-mouse F4/80 (1:200 dilution, Abcam, Cat# 6640, RRID: AB_1140040), rat anti-mouse B220 (1:200 dilution, BD Biosciences, Cat# 550286, RRID:AB_393581), rabbit anti-mouse CD3 (1:200 dilution, Abcam, Cat# ab5690, RRID:AB_305055), rat anti-mouse CD8 (1:150 dilution, eBioscience, Cat# 14–0808–82, RRID: AB_2572861), and rat anti-mouse Foxp3 (1:100 dilution, eBioscience, Cat# 12–5773–80, RRID: AB_465935).

Immunofluorescence staining

Paraffin-embedded stomach sections were dewaxed, rehydrated and antigen-retrieval was performed by heating in EDTA buffer (pH9) for 15 min in microwave pressure cooker. Section were incubated for 1 h at room temperature with goat anti-mouse IL-33 (1:150 dilution, R&D Systems, Cat# AF3626, RRID:AB_884269) and then for 1 h with AlexaFluor 568-conjugated donkey anti-goat secondary antibody (Molecular Probes Cat# A-11057, RRID:AB_142581). After counter staining with spectral DAPI (1:200 dilution, PerkinElmer, Cat# FP1490) for 5 min, sections were mounted with Vectashield mounting solution (VECTORlabs, Cat# H-1400). Imaging was preformed with a Vectra 3.0 Automated Quantitative Pathology Imaging System (PerkinElmer, Cat# CLS142338) and representative images were produced using Phenochart^TM v1.0.4 software (PerkinElmer).

Quantification

For quantification of mast cell numbers at least three fields of view (10x ocular) of three toluidine blue section per mouse were counted, submucosal area was measured and data represented as mast cells number/area tissue. Similarly, for quantification of CD3, BrdU, and ApopTag staining, either entire tumors or three fields of view of three sections were counted and analyzed comparing to tissue area.

F4/80, CD31, B220, and Hypoxia immune-staining was quantified using Metamorph software (Molecular Devices) or Fiji scripts (ImageJ, htps://Fiji.sc/) determining percentage positive area per area tissue.

Isolation of gastric epithelial cells, immune cells, and flow cytometric analysis

Glandular stomachs or antral tumors were dissected, cut into very small pieces and incubated at 37 °C for 30 min in Ca²⁺- and Mg²⁺-free HBSS medium plus 2.5% FCS and 1 mM EDTA with gentle shaking. Then samples were vortexed for 30 s and the supernatants containing intraepithelial lymphocytes were separated from the tissue fragments and kept on ice. The remaining tissue samples were further digested in Collagenase/Dispase (Roche) and DNase I (Roche) in Ca²⁺- and Mg²⁺-free HBSS medium plus 2% FCS for 45 min at 37 °C under continuous rotation. Samples were vortexed for 30 s once during incubation and once after the incubation. Afterwards, cell suspensions from both incubations were pooled, filtered and washed in PBS plus 5% FCS for analysis by flow cytometry.

Single cell suspensions were stained for cell surface markers (listed below) and cell viability was controlled with propidium iodide (ThermoFisher Scientific, Cat# P3566) or SYTOXBlue (ThermoFisher Scientific, Cat# S34857) staining. Cell sorting was performed with an Aria II cell sorter (BD Bioscience). All cell-type specific flow cytometric gating strategies are presented as a supplementary figure (Supplementary Fig. 8a–e). Mast cells sorting: from CD45⁺EpCam⁻ cell population the CD11b⁻ cells were selected, then finally the c-Kit⁺FceRI⁺ population represents the mast cells. Macrophage sorting: from CD45⁺EpCam⁻ cells, the F4/80^HighCD11b⁺ population was selected and back-gated to confirm that the selected macrophage population was Ly6C⁻Ly6G⁻.

The following fluorochrome-conjugated antibodies were used for flow cytometric cell sorting and analysis: CD16/CD32 (1/100 dilution, clone 93, Cat# 14-0161-86), EpCAM-FITC (1/400 dilution, clone 9CA,Cat# 11-5791-82), Ly6c-eF450 (1/300 dilution, HK1.4, Cat# 48-5932-82), F4/80-PE-Cy7 (1/400 dilution, BM8, Cat# 25-4801-82), St2-PerCP-eFluor710 (1:200 dilution, clone RMST2-2, Cat# 46-9335-82), Foxp3-PE (1/200 dilution, clone FJK-165, Cat# 12-5773-82), Gata3-PE (1/200 dilution, clone TWAJ, Cat# 12-9966-42), FceR1-PE-Cy7 (1/300, clone 36951, Cat# 25-5898-82), and CD3-PE-Cy7 (1/1000, 145-2C11, Cat# 25-0031-82) all from Ebioscience; CD11b-PE (1/400 dilution, M1/70, Cat# 553311), Ly6g (1/300 dilution, 1A8, Cat# 560602) from BD Pharmingen and CD45.2-A700 (1/400 dilution, clone S450-15-2), CD4 (1/50 dilution, clone GK 1.5), CD11b-PB (1/400 dilution, M1/70), and c-Kit-APC (1/200, ACK-2) from WEHI monoclonal antibody facility.

Isolation and stimulation of BMDM

Bone marrow was collected from the femur and tibia of mice by flushing with sterile PBS. Cells were washed with PBS twice and were filtered through a 100 mm sieve. Resulting cell suspensions were cultured in DMEM supplemented with 10% (v/v) FCS and L929 conditioned medium for 7 days with fresh media changes every second day until fully differentiated into BMDM. After reseeding of cells, alternative macrophage polarization was conducted by stimulation with IL-4 (20 ng/ml, Preprotech) and IL-13 (20 ng/ml, Preprotech) for one day.

Cytokine bead array

A total of 50,000 isolated gastric gp130^FF mast cells were cultured in 96-well plates for 3 h in 30 ng/mL IL-33. Supernatants were collected, diluted 1:2 in assay buffer and assayed for cytokine concentration using the Bio-Rad Bioplex cytokine bead assay (Bio-Rad Mouse 23-Plex Panel M60009RDPD) strictly according to manufacturer’s instructions.

BMMC adaptive transfer assay

Mast cells were generated from bone marrow from either gp130^FF; ST2^+/+ or gp130^FF; ST2^−/− mice. Bone marrow was extracted and cells were maintained in IL-3-containing culture medium for several weeks until >95% pure BMMC cultures were established. Purity of BMMCs (FceR1⁺, c-Kit⁺) was determined via flow cytometry. For the BMMC transplantation assay, syngeneic BMMC from the same colony as the host mice were used. Either gp130^FF; ST2^+/+ or gp130^FF; ST2^−/− BMMCs were injected into the tail vein of gp130^FF; ST2⁻^/− mice. Each mouse received 10⁷ BMMC at 5, 8, and 11 weeks of age and final tumor analysis was performed at 14 weeks of age.

Gastric organoid assay

Antral tumors from gp130^FF mice were used to establish tumor-epithelial organoids. Organoids were established and maintained in IntestiCult^TM Organoid Growth Medium (StemCell Technologies) according to manufacturer's protocols. Established organoids were stimulated with 100 ng/ml IL-11 or PBS control for 4 h and then processed for gene expression analysis via qPCR. For assessment of organoid growth, PBS or IL-11 stimulated organoids were monitored over 4 days.

RNA isolation and quantitative RT-PCR

Total RNA was extracted from frozen tissue samples using Trizol^®Reagent (life technologies, Cat# 15596026) and cDNA was prepared from 2 µg RNA using the High capacity cDNA Reverse Transcription kit (Applied Biosystems, Cat# 4368813) according to the manufacturer’s protocol.

From isolated gastric mast cells and from FACS sorted macrophages, RNA was extracted using the RNeasy Plus Micro Kit (QIAGEN, Cat# 74034) and cDNA synthesis was performed with the ThermoScript^TM RT-PCR System (Invitrogen, Cat# 11146-024) according to the manufacturer’s instructions.

Quantitative RT-PCR analyses were performed in technical triplicates with SensiMix SYBR kit (Bioline, Cat# QT605–20) using the ViiA™ 7 Real Time PCR System (life technologies). Further details and the sequences of the used oligonucleotides are described in the supplementary methods and Supplementary Table S1.

Kaplan–Meier survival analysis

All Kaplan–Meier survival analysis was performed with KMplot (KMplot.com)⁵³. Data from 179 intestinal-type gastric cancer patient were analyzed and following settings were applied: probes = all probe sets, Laurens classification = intestinal-type, Data sets included: GSE14210, GSE15459, GSE22377, GSE29272, and GSE51105, but excluding GSE62254 (as recommended; due to markedly different survival and shifted expression compared to the other data sets). All other settings were kept as default.

Statistics

Unless otherwise stated, data are presented as mean values ± standard error of the mean (M ± SEM). Exact n values are depicted for every data set in the figure legends and they are always true biological replicates, not technical replicates. All data sets were tested for normality with Shapiro–Wilk normality test. Comparisons between two mean values were performed by two-tailed unpaired Student’s t-test using Prism6 software (GraphPad Software, California, USA). When data sets had significantly different standard deviations according to Prism’s F-test of variances, Welch correction was applied. T-test t values and corresponding degree of freedom (df) are given in figure legends. For comparisons of more than two groups one-way ANOVA test was performed with multiple comparison correction (either Tukey or Bonferroni). ANOVA F values are depicted in each figure legend as F (DFn, Dfd), where DFn is the df nominator and Dfd the df denominator. P values of less than 0.05 were considered statistically significant.

Sample size determination

G*power3.1 software was used to estimate the minimal required sample sizes. Based on previous mean values and SD from previous studies using the gp130^FF mouse model, we calculated a minimal sample size for eight mice per group, when the mean difference is 40% and the SD 20% (unpaired two-tailed t-test; α = 0.05, Power = 0.95). Due to difference in availability and frequency of target genotypes born, the number of mice analyzed between cohorts differed. Sample size for subsequent analysis of tissue section staining was limited to number available, as tissues were required for several downstream processes including FACS analysis, RNA and protein expression in addition to histological and immunohistochemical analysis.

Randomization

No formal randomization procedure was performed. However, for genetic mouse experiments, littermates were used to compare different target genotype mice. For in vivo treatment experiments, different treatment groups were equally distributed between mouse litters and mouse cages as well as female and male ratio were kept at 1:1.

Data exclusion

No data was excluded after commencement of quantification. In a few cases, tissue slides were excluded from downstream analysis due to strong background staining or absence of tumor tissue. These slides were excluded from analysis before commencement of quantification and without knowledge of genotypes/treatment details (aka blinded).

Blinding

Assessors were blinded for determining tumor masses and numbers and for quantification of histo- and immunohistochemical staining. Blinding was achieved through labeling minimized to mouse IDs. Genotypes and treatment were only linked back to the mouse ID after completion of quantification.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The Stat3 chromatin immunoprecipitation-sequencing (ChIP-Seq) data have been deposited in the NCBI gene expression omnibus (GEO) under the accession code GSE48285. The source data underlying Figs. 1b, d, 2b–h, 3b–e, 4b–e, 5a–c, 6a, c, d, f, 7a–g, and 8a and Supplementary Figs 1b, 2a, b, g, i, 3d, 5a, b, g, h, and 6e, f are provided as a source data file. All the other data supporting the findings of this study are available within the article and its supplementary information files and from the corresponding author upon reasonable request. A reporting summary for this article is available as a Supplementary Information file.

References

Joyce, J. A. Therapeutic targeting of the tumor microenvironment. Cancer Cell 7, 513–520 (2005).
Article CAS Google Scholar
Fridman, W. H., Pagès, F., Sautès-Fridman, C. & Galon, J. The immune contexture in human tumours: impact on clinical outcome. Nat. Rev. Cancer 12, 298–306 (2012).
Article CAS Google Scholar
Qian, B.-Z. & Pollard, J. W. Macrophage diversity enhances tumor progression and metastasis. Cell 141, 39–51 (2010).
Article CAS Google Scholar
Galli, S. J., Nakae, S. & Tsai, M. Mast cells in the development of adaptive immune responses. Nat. Immunol. 6, 135–142 (2005).
Article CAS Google Scholar
Blank, U. & Rivera, J. The ins and outs of IgE-dependent mast-cell exocytosis. Trends Immunol. 25, 266–273 (2004).
Article CAS Google Scholar
Galli, S. J., Grimbaldeston, M. & Tsai, M. Immunomodulatory mast cells: negative, as well as positive, regulators of innate and acquired immunity. Nat. Rev. Immunol. 8, 478–486 (2008).
Article CAS Google Scholar
Ribatti, D. et al. Tumor vascularity and tryptase-positive mast cells correlate with a poor prognosis in melanoma. Eur. J. Clin. Investig. 33, 420–425 (2003).
Article CAS Google Scholar
Nonomura, N. et al. Decreased number of mast cells infiltrating into needle biopsy specimens leads to a better prognosis of prostate cancer. Br. J. Cancer 97, 952–956 (2007).
Article CAS Google Scholar
Strouch, M. J. et al. Crosstalk between mast cells and pancreatic cancer cells contributes to pancreatic tumor progression. Clin. Cancer Res. 16, 2257 (2010).
Article CAS Google Scholar
Elpek, G. et al. The prognostic relevance of angiogenesis and mast cells in squamous cell carcinoma of the oesophagus. J. Clin. Pathol. 54, 940–944 (2001).
Article CAS Google Scholar
Ribatti, D. et al. Mast cells and angiogenesis in gastric carcinoma. Int. J. Exp. Pathol. 91, 350–356 (2010).
Article CAS Google Scholar
Yano, H. et al. Mast cell infiltration around gastric cancer cells correlates with tumor angiogenesis and metastasis. Gastric Cancer 2, 26–32 (1999).
Article Google Scholar
Theoharides, T. C. Mast cells and pancreatic cancer. N. Engl. J. Med. 358, 1860–1861 (2008).
Article CAS Google Scholar
Nielsen, H. J. et al. Independent prognostic value of eosinophil and mast cell infiltration in colorectal cancer tissue. J. Pathol. 189, 487–495 (1999).
Article CAS Google Scholar
Rajput, A. B. et al. Stromal mast cells in invasive breast cancer are a marker of favourable prognosis: a study of 4,444 cases. Breast Cancer Res. Treat. 107, 249–257 (2008).
Article Google Scholar
Soucek, L. et al. Mast cells are required for angiogenesis and macroscopic expansion of Myc-induced pancreatic islet tumors. Nat. Med. 13, 1211–1218 (2007).
Article CAS Google Scholar
Schonhuber, N. et al. A next-generation dual-recombinase system for time- and host-specific targeting of pancreatic cancer. Nat. Med. 20, 1340–1347 (2014).
Article Google Scholar
Lüthi, A. U. et al. Suppression of interleukin-33 bioactivity through proteolysis by apoptotic caspases. Immunity 31, 84–98 (2009).
Article Google Scholar
Rickard, J. A. et al. RIPK1 regulates RIPK3-MLKL-driven systemic inflammation and emergency hematopoiesis. Cell 157, 1175–1188 (2014).
Article CAS Google Scholar
Martin, N. T. & Martin, M. U. Interleukin 33 is a guardian of barriers and a local alarmin. Nat. Immunol. 17, 122–131 (2016).
Article CAS Google Scholar
Nile, C. J., Barksby, E., Jitprasertwong, P., Preshaw, P. M. & Taylor, J. J. Expression and regulation of interleukin-33 in human monocytes. Immunology 130, 172–180 (2010).
Article Google Scholar
Kouzaki, H., Iijima, K., Kobayashi, T., O’Grady, S. M. & Kita, H. The danger signal, extracellular ATP, is a sensor for an airborne allergen and triggers IL-33 release and innate Th2-type responses. J. Immunol. 186, 4375–4387 (2011).
Article CAS Google Scholar
Ohno, T. et al. Caspase-1, caspase-8, and calpain are dispensable for IL-33 release by macrophages. J. Immunol. 183, 7890–7897 (2009).
Article CAS Google Scholar
Wang, J.-X. et al. IL-33/ST2 axis promotes mast cell survival via BCLXL. Proc. Natl Acad. Sci. USA 111, 10281–10286 (2014).
Article CAS ADS Google Scholar
Licona-Limon, P., Kim, L. K., Palm, N. W. & Flavell, R. A. TH2, allergy and group 2 innate lymphoid cells. Nat. Immunol. 14, 536–542 (2013).
Article CAS Google Scholar
Schiering, C. et al. The alarmin IL-33 promotes regulatory T-cell function in the intestine. Nature 513, 564–568 (2014).
Article CAS ADS Google Scholar
Vasanthakumar, A. et al. The transcriptional regulators IRF4, BATF and IL-33 orchestrate development and maintenance of adipose tissue-resident regulatory T cells. Nat. Immunol. 16, 276–285 (2015).
Article CAS Google Scholar
Schmitz, J. et al. IL-33, an Interleukin-1-like cytokine that signals via the IL-1 receptor-related protein ST2 and induces T helper type 2-associated cytokines. Immunity 23, 479–490 (2005).
Article CAS Google Scholar
Molofsky, A. B., Savage, A. & Locksley, R. M. Interleukin-33 in tissue homeostasis, injury and inflammation. Immunity 42, 1005–1019 (2015).
Article CAS Google Scholar
Gramatzki, D. et al. Interleukin-33 in human gliomas: expression and prognostic significance. Oncol. Lett. 12, 445–452 (2016).
Article CAS Google Scholar
Tong, X. et al. Interleukin-33 predicts poor prognosis and promotes ovarian cancer cell growth and metastasis through regulating ERK and JNK signaling pathways.Mol. Oncol. 10, 113–125 (2016).
Article CAS Google Scholar
Chen, S.-F. et al. The paracrine effect of cancer-associated fibroblast-induced interleukin-33 regulates the invasiveness of head and neck squamous cell carcinoma. J. Pathol. 231, 180–189 (2013).
Article CAS Google Scholar
Maywald, R. L. et al. IL-33 activates tumor stroma to promote intestinal polyposis. Proc. Natl Acad. Sci. USA 112, E2487–E2496 (2015).
Article CAS ADS Google Scholar
O’Donnell, C. et al. An antitumorigenic role for the IL-33 receptor, ST2L, in colon cancer. Br. J. Cancer 114, 37–43 (2016).
Article Google Scholar
Malik, A. et al. IL-33 regulates the IgA-microbiota axis to restrain IL-1α–dependent colitis and tumorigenesis. J. Clin. Investig. 126, 4469–4481 (2016).
Article Google Scholar
Eissmann, M. F. et al. Interleukin 33 signaling restrains sporadic colon cancer in an interferon-gamma dependent manner.Cancer Immunol. Res. 6, 409–421 (2018).
Article CAS Google Scholar
Thiem, S. et al. mTORC1 inhibition restricts inflammation-associated gastrointestinal tumorigenesis in mice. J. Clin. Investig. 123, 767–781 (2013).
CAS PubMed Google Scholar
Putoczki, TracyL. et al. Interleukin-11 Is the dominant IL-6 family cytokine during gastrointestinal tumorigenesis and can be targeted therapeutically. Cancer Cell 24, 257–271 (2013).
Article CAS Google Scholar
Ernst, M. et al. STAT3 and STAT1 mediate IL-11–dependent and inflammation-associated gastric tumorigenesis in gp130 receptor mutant mice. J. Clin. Investig. 118, 1727–1738 (2008).
CAS PubMed Google Scholar
Jenkins, B. J. et al. Hyperactivation of Stat3 in gp130 mutant mice promotes gastric hyperproliferation and desensitizes TGF-beta signaling. Nat. Med. 11, 845–852 (2005).
Article CAS Google Scholar
Grimbaldeston, M. A. et al. Mast cell-deficient W-sash c-kit Mutant KitW-sh/W-sh mice as a model for investigating mast cell biology in vivo. Am. J. Pathol. 167, 835–848 (2005).
Article CAS Google Scholar
Wolters, P. J. et al. Tissue-selective mast cell reconstitution and differential lung gene expression in mast cell-deficient KitW-sh/KitW-sh sash mice. Clin. Exp. Allergy 35, 82–88 (2005).
Article CAS Google Scholar
Duttlinger, R. et al. W-sash affects positive and negative elements controlling c-kit expression: ectopic c-kit expression at sites of kit-ligand expression affects melanogenesis. Development 118, 705 (1993).
CAS PubMed Google Scholar
Lilla, J. N. et al. Reduced mast cell and basophil numbers and function in Cpa3-Cre; Mcl-1fl/fl mice. Blood 118, 6930–6938 (2011).
Article CAS Google Scholar
Traum, D. et al. IL-10-induced gp130 expression in mouse mast cells permits IL-6 trans-signaling. J. Leukoc. Biol. 91, 427–435 (2012).
Article CAS Google Scholar
Gyotoku, E. et al. The Il-6 family cytokines, interleukin-6, interleukin-11, oncostatin M, and leukemia inhibitory factor, enhance mast cell growth through fibroblast-dependent pathway in mice. Arch. Dermatol. Res 203, 508–514 (2001).
Article Google Scholar
Li, X. et al. Targeting of tumour-infiltrating macrophages via CCL2/CCR2 signalling as a therapeutic strategy against hepatocellular carcinoma. Gut 66, 157–167 (2017).
Article CAS Google Scholar
Ries, C. H. et al. Targeting tumor-associated macrophages with anti-CSF-1R antibody reveals a strategy for cancer therapy. Cancer Cell 25, 846–859 (2014).
Article CAS Google Scholar
Komai-Koma, M. et al. Interleukin-33 amplifies IgE synthesis and triggers mast cell degranulation via interleukin-4 in naïve mice. Allergy 67, 1118–1126 (2012).
Article CAS Google Scholar
Belo, A. V. et al. Murine chemokine CXCL2/KC is a surrogate marker for angiogenic activity in the inflammatory granulation tissue. Microcirculation 12, 597–606 (2005).
Article CAS Google Scholar
Michalec, L. et al. CCL7 and CXCL10 orchestrate oxidative stress-induced neutrophilic lung inflammation. J. Immunol. 168, 846 (2002).
Article CAS Google Scholar
Amann, B., Perabo, F., Wirger, A., Hugenschmidt, H. & Schultze-Seemann, W. Urinary levels of monocyte chemo-attractant protein-1 correlate with tumour stage and grade in patients with bladder cancer. Br. J. Urol. 82, 118–121 (1998).
Article CAS Google Scholar
Szász, A. M. et al. Cross-validation of survival associated biomarkers in gastric cancer using transcriptomic data of 1,065 patients. Oncotarget 7, 49322–49333 (2016).
Article Google Scholar
Zhao, Y. et al. Increased numbers of gastric-infiltrating mast cells and regulatory T cells are associated with tumor stage in gastric adenocarcinoma patients. Oncol. Lett. 4, 755–758 (2012).
Article ADS Google Scholar
Engblom, C., Pfirschke, C. & Pittet, M. J. The role of myeloid cells in cancer therapies. Nat. Rev. Cancer 16, 447–462 (2016).
Article CAS Google Scholar
Kanbe, N. et al. Human mast cells produce matrix metalloproteinase 9. Eur. J. Immunol. 29, 2645–2649 (1999).
Article CAS Google Scholar
Ranieri, G. et al. Tryptase-positive mast cells correlate with angiogenesis in early breast cancer patients. Int. J. Oncol. 35, 115–120 (2009).
Article Google Scholar
Kondo, K. et al. Expression of chymase-positive cells in gastric cancer and its correlation with the angiogenesis. J. Surg. Oncol. 93, 36–42 (2006).
Article CAS Google Scholar
Bot, I. et al. Perivascular mast cells promote atherogenesis and induce plaque destabilization in apolipoprotein E–deficient mice. Circulation 115, 2516 (2007).
Article CAS Google Scholar
Rigoni, A. et al. Mast cells infiltrating inflamed or transformed gut alternatively sustain mucosal healing or tumor growth. Cancer Res. 75, 3760 (2015).
Article CAS Google Scholar
von Moltke, J., Ji, M., Liang, H. E. & Locksley, R. M. Tuft-cell-derived IL-25 regulates an intestinal ILC2-epithelial response circuit. Nature 529, 221–225 (2016).
Article ADS Google Scholar
Petersen, C. P. et al. A signalling cascade of IL-33 to IL-13 regulates metaplasia in the mouse stomach. Gut 67, 805–817 (2017).
Article Google Scholar
Liew, F. Y., Girard, J.-P. & Turnquist, H. R. Interleukin-33 in health and disease. Nat. Rev. Immunol. 16, 676–689 (2016).
Article CAS Google Scholar
Kobori, A. et al. Interleukin-33 expression is specifically enhanced in inflamed mucosa of ulcerative colitis. J. Gastroenterol. 45, 999–1007 (2010).
Article CAS Google Scholar
Nishida, A. et al. Expression of interleukin 1-like cytokine interleukin 33 and its receptor complex (ST2L and IL1RAcP) in human pancreatic myofibroblasts. Gut 59, 531 (2010).
Article CAS Google Scholar
Yanagawa, Y., Suzuki, M., Matsumoto, M. & Togashi, H. Prostaglandin E2 enhances IL-33 production by dendritic cells. Immunol. Lett. 141, 55–60 (2011).
Article CAS Google Scholar
Suen, Y., Chang, M., Lee, S. M., Buzby, J. S. & Cairo, M. S. Regulation of interleukin-11 protein and mRNA expression in neonatal and adult fibroblasts and endothelial cells. Blood 84, 4125–4134 (1994).
CAS PubMed Google Scholar
Liu, X. et al. Intratumor IL-17-positive mast cells are the major source of the IL-17 that is predictive of survival in gastric cancer patients. PLoS One 9, e106834 (2014).
Article ADS Google Scholar
Dai, X.-M. et al. Targeted disruption of the mouse colony-stimulating factor 1 receptor gene results in osteopetrosis, mononuclear phagocyte deficiency, increased primitive progenitor cell frequencies, and reproductive defects. Blood 99, 111 (2002).
Article CAS Google Scholar
Hoshino, K. et al. The absence of interleukin 1 receptor–related T1/St2 does not affect T helper cell Type 2 development and its effector function. J. Exp. Med. 190, 1541 (1999).
Article CAS Google Scholar
Thiem, S. et al. Stomach-specific activation of oncogenic KRAS and STAT3-dependent inflammation cooperatively promote gastric tumorigenesis in a preclinical model. Cancer Res. 76, 2277–2287 (2016).
Article CAS Google Scholar
Boussioutas, A. et al. Distinctive patterns of gene expression in premalignant gastric mucosa and gastric cancer. Cancer Res. 63, 2569–2577 (2003).
CAS PubMed Google Scholar
Thompson, P. J., Hanson, J. M. & Morley, J. Asthma, mast cells, and sodium cromoglycate. Lancet 322, 848–849 (1983).
Article Google Scholar
Bancroft, J. D. & Alan, S. Theory and practice of histological techniques. 3rd edn, (Churchill Liningstone Elsevier, New York, 1990).

Download references

Acknowledgements

We thank the animal facility staff of La Trobe University and the Ludwig Institute for Cancer Research as well as the Austin and WEHI histology department for technical assistance. We also acknowledge the excellent support for cell sorting by David Baloyan (ONJCRI) and image generation and analysis by Cameron Nowell (Monash University) and Lachlan Whitehead (WEHI). This work was supported through the Victorian State Government Operational Infrastructure Support, the Nation Health and Medical Research Council (NHMRC) of Australia grants 1092788, 1067244, 1069024 and Cancer Council Victoria’s Grant-in-Aid 1160708. M.E. is a NHMRC Principal Research Fellow and also received funding from Ludwig Cancer Research.

Author information

Toby Phesse
Present address: Cell Signaling and Cancer Laboratory, European Cancer Stem Cell Research Institute and Cardiff University, Cardiff, CF24 4HQ, UK
Jamina Brunnberg
Present address: Institute of Biochemistry, Goethe University Frankfurt, Frankfurt am Main, 60438, Frankfurt, Germany
Nima Etemadi
Present address: Cell Signalling and Cell Death Division, Walter and Eliza Hall Institute of Medical Research, and Department of Medical Biology, University of Melbourne, Melbourne, VIC, 3052, Australia
Michele A. Grimbaldeston
Present address: OMNI-Biomarker Development, Genentech Inc., South San Francisco, CA, 94080, USA
Robert J. J. O’Donoghue
Present address: Department of Pharmacology and Therapeutics, University of Melbourne, Melbourne, VIC, 3010, Australia
Frederick Masson
Present address: Team 5, Centre of Physiopathology Toulouse-Purpan, INSERM UMR 1043/CNRS UMR 5282, University Toulouse III, CHU Purpan, 31024, Toulouse, France

Authors and Affiliations

Cancer and Inflammation Laboratory, Olivia Newton-John Cancer Research Institute and School of Cancer Medicine, La Trobe University, Heidelberg, VIC, 3084, Australia
Moritz F. Eissmann, Christine Dijkstra, Toby Phesse, Jamina Brunnberg, Ashleigh R. Poh, Nima Etemadi, Stefan Thiem, Michael Buchert, Robert J. J. O’Donoghue, Frederick Masson & Matthias Ernst
Department of Immunology and Pathology, Monash University, Melbourne, VIC, 3004, Australia
Evelyn Tsantikos & Margaret L. Hibbs
Molecular Immunology Division, The Walter and Eliza Hall Institute of Medical Research, and Department of Medical Biology, University of Melbourne, Melbourne, VIC, 3052, Australia
Nicholas D. Huntington
Department of Medicine, University of Melbourne, Melbourne, VIC, 3050, Australia
Alex Boussioutas
Centre for Cancer Biology, University of South Australia and SA Pathology, Adelaide, SA, 5000, Australia
Michele A. Grimbaldeston
Department of Pharmacology and Therapeutics, University of Melbourne, Melbourne, VIC, 3010, Australia
Andrew Jarnicki

Authors

Moritz F. Eissmann
View author publications
You can also search for this author in PubMed Google Scholar
Christine Dijkstra
View author publications
You can also search for this author in PubMed Google Scholar
Andrew Jarnicki
View author publications
You can also search for this author in PubMed Google Scholar
Toby Phesse
View author publications
You can also search for this author in PubMed Google Scholar
Jamina Brunnberg
View author publications
You can also search for this author in PubMed Google Scholar
Ashleigh R. Poh
View author publications
You can also search for this author in PubMed Google Scholar
Nima Etemadi
View author publications
You can also search for this author in PubMed Google Scholar
Evelyn Tsantikos
View author publications
You can also search for this author in PubMed Google Scholar
Stefan Thiem
View author publications
You can also search for this author in PubMed Google Scholar
Nicholas D. Huntington
View author publications
You can also search for this author in PubMed Google Scholar
Margaret L. Hibbs
View author publications
You can also search for this author in PubMed Google Scholar
Alex Boussioutas
View author publications
You can also search for this author in PubMed Google Scholar
Michele A. Grimbaldeston
View author publications
You can also search for this author in PubMed Google Scholar
Michael Buchert
View author publications
You can also search for this author in PubMed Google Scholar
Robert J. J. O’Donoghue
View author publications
You can also search for this author in PubMed Google Scholar
Frederick Masson
View author publications
You can also search for this author in PubMed Google Scholar
Matthias Ernst
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Experimental work: M.F.E., A.J., M.B., R.J.J.O’D., T.P., C.D., J.B., S.T., E.T., N.D.H., N.E., A.R.P., and F.M. Data interpretation: M.F.E, A.J., M.B., R.J.J.O’D., A.B., M.L.H., M.A.G., F.M., and M.E. Writing and editing of manuscript: M.F.E., M.L.H., F.M., and M.E. Study concept: M.F.E., A.J., F.M., and M.E. Funding: M.E.

Corresponding author

Correspondence to Matthias Ernst.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information: Nature Communications thanks Jean Marshall and other anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Eissmann, M.F., Dijkstra, C., Jarnicki, A. et al. IL-33-mediated mast cell activation promotes gastric cancer through macrophage mobilization. Nat Commun 10, 2735 (2019). https://doi.org/10.1038/s41467-019-10676-1

Download citation

Received: 23 January 2018
Accepted: 22 May 2019
Published: 21 June 2019
DOI: https://doi.org/10.1038/s41467-019-10676-1

This article is cited by

Prognostic role of macrophages and mast cells in the microenvironment of hepatocellular carcinoma after resection
- Esraa Ali
- Lenka Červenková
- Andriy Trailin
BMC Cancer (2024)
Synovial microenvironment-influenced mast cells promote the progression of rheumatoid arthritis
- Yunxuan Lei
- Xin Guo
- Guangjie Chen
Nature Communications (2024)
IL-33 aggravates extranodal NK/T cell lymphoma aggressiveness and angiogenesis by activating the Wnt/β-catenin signaling pathway
- Mingli Ni
- Yuhui Wang
- Yunlong Wang
Molecular and Cellular Biochemistry (2024)
Gastric intestinal metaplasia: progress and remaining challenges
- Qi-Yue Tong
- Min-Jiao Pang
- Zhi-Feng Miao
Journal of Gastroenterology (2024)
Single-cell analysis unveils activation of mast cells in colorectal cancer microenvironment
- Zhenyu Xie
- Liaoran Niu
- Jianyong Zheng
Cell & Bioscience (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.