Polyps Wrap Mast Cells and Treg within Tumorigenic Tentacles

Gounaris and colleagues describe a previously unrecognized cross-talk between mast cells and Treg in colon adenomatous polyposis (Gounaris et al., Cancer Res 2009;69:5490–7). Adoptively transferred Treg suppress the focal mastocytosis that fosters tumor initiation and progression. In contrast, endogenous Treg, which abundantly infiltrate polyps, show proinflammatory activity under unknown microenvironmental cues that promote mast cell differentiation and expansion. Compartmentalized Treg plasticity seems to be a key factor in establishing the optimal milieu for cancer development in the intestines. Treg partnership with mast cells recapitulates the complexity of innate-adaptive networks characterizing gut inflammation and represents a novel target for cancer immunotherapy. [Cancer Res 2009;69(14):5619–22]

Gut, as other mucosal tissues exposed to the external environment, is a site where inflammation and cancer are tightly linked. Bacterial infections, exposure to toxic molecules damaging the epithelial barrier, genetic predisposition, and enhanced immune reactivity, may promote chronic colitis that induces epithelial cell proliferation, stroma remodeling, neoangiogenesis, and suppression of the antitumor adaptive response (1).

Regulatory T cells (Treg) play nonredundant roles in colitis (2, 3). The intestinal Treg pool includes naturally arising CD4⁺Foxp3⁺ Treg, thymus-derived and endowed with a self-directed antigenic repertoire, and also adaptive Treg, developing in the gut upon encounter with foreign antigens under transforming growth factor (TGF)-β and retinoic acid exposure (4). Although systemic Treg suppress immune responses through an array of different mechanisms, the inhibitory function of intestinal Treg is critically dependent on interleukin (IL)-10, especially under microbial triggers (5). Treg activity under homeostatic conditions is as crucial as Treg inactivation when response to pathogens must initiate. TLR ligands and inflammatory cytokines can paralyze Treg functions directly or indirectly, by turning tolerogenic antigen-presenting cells into immunogenic cells that no longer sustain Treg-mediated suppression (6). Under precise microenvironmental stimuli, Treg may be not only “deactivated” but also “reprogrammed” toward alternative differentiation pathways, for instance skewed into IL-17–producing T cells (Th17; ref. 7). Since the discovery of Foxp3 as their master transcription factor, Treg have been considered as a committed lineage, but many data support the idea that Treg can adapt to external and internal challenges with high plasticity and heterogeneity. Distinct Treg subsets, expressing T helper–associated transcription factors, selectively suppress different effector responses (8, 9). It could be predicted that specialized Treg and Teff subsets develop in parallel under different microenvironmental stimuli to maximize immunity and minimize collateral damage (10). A forced conversion of “hybrid” Treg into the corresponding class of effector cells may lie behind pathologic conditions (10).

In tumor immunology, Treg are viewed as one of the major obstacles to the endogenous antitumor immunity and to successful immunotherapy. High Treg frequency, observed in a variety of human and murine cancers, has been associated with poor prognosis (11). Immunotherapeutic strategies aimed at Treg depletion and/or inactivation, as a single approach or in combination with conventional treatments, have shown to significantly improve antitumor response (12). The case of Treg involvement in colon carcinogenesis is unique. Studies from the Erdman's group have shown that transfer of Treg prevents cancer development associated to colitis, induced by microbial infection or by genetic susceptibility to polyposis in Apc^Min mice, an effect dependent on IL-10 acting on cells of the innate response (13–15). From a clinical point of view, high-density Treg infiltration in colorectal cancer tissue has been associated to better prognosis, in sharp contrast with other types of solid cancers (16). It could be argued that, during cancer progression, Treg may exert two opposing roles, the suppression of tumor immunity and the inhibition of inflammation; the former occurs in the majority of solid cancer, whereas the latter may be relevant during the carcinogenetic processes fostered by inflammatory conditions. This latter event is not restricted to colorectal but extended to prostatic cancer spontaneously arising in Apc^Min mice, a context in which the adoptive transfer of Treg decreases inflammation and protects from tumor development (17).

Among the innate myeloid cells partnering Treg activities, mast cells are gaining new interest. Mast cells produce a wide array of molecules and express panoply of surface receptors that allow interactions with all immune cells (18). In the forefront of immune reactions, they release very rapidly the prepackaged granules containing early-response mediators, while synthesizing ex novo second-phase molecules. In recent years, mast cells have been recognized to play immunoregulatory, beside immunostimulatory, functions, mainly through the secretion of IL-10 (19). In spite of their relative paucity in tissues, mast cells play nonredundant and complex roles not only in allergy, but also in autoimmune diseases, allograft tolerance, and cancer (20). Mastocytosis is quite commonly associated to neoplasia. Lisa Coussens and coworkers (21, 22) first showed that mast cells, recruited early into skin cancer, support tumor growth by favoring neoangiogenesis and invasiveness through matrix metalloproteinase 9 secretion. Tumor-associated mast cells favor tumor growth not only participating to stroma remodeling, but also exerting immunoregulatory functions on adaptive antitumor response, thus behaving as tolerogenic cells (23).

In this issue, Gounaris and colleagues (24) describe the molecular and cellular loops coiling mast cells and Treg in intestinal polyposis. Based on the premises that adoptively transferred Treg suppress polyposis in the Apc^Min model (15) and that mastocytosis is a necessary event to polyp outgrowth in Apc^Δ468 mice (25), the authors show that the two events are causally related. Indeed, transferred Treg were able to suppress focal mastocytosis in vivo and to inhibit mast cell progenitor differentiation and expansion in vitro. Strikingly, polyps were populated by high numbers of endogenous Treg, which were anergic and suppressive ex vivo, while unable to counteract mastocytosis and polyposis “from the inside.” IL-10 has been identified as the discriminating molecule that, produced by exogenous and not by endogenous Treg, suppressed in vitro mast cell differentiation. However, the definitive prove of the essential role of Treg-derived IL-10 in suppressing polyposis, mainly through the inhibition of focal mastocytosis, in vivo, is still lacking. Endogenous Treg from polyp-ridden mice not only failed to inhibit, but also enhanced, mast cell differentiation in vitro, suggesting that IL-10 deficit alone could not account for the overall effect. Of note, endogenous Treg displayed not only defective IL-10 secretion but also high IL-17 production. IL-17, but also other molecules produced by polyp-infiltrating Treg, may actively recall mast cell in vivo. Thus, locally diverted Treg may help tumor growth through the arrangement of a mast cell–supported proinflammatory microenvironment.

This picture raises the intriguing and topical issue of which factors are redirecting Treg conventional suppressive ability into tissue-restricted proinflammatory activity. Tumor cells, but also infiltrating immune cells, may produce cytokines capable to induce Treg diversion. It could be even suggested that mast cells themselves may be responsible for IL-10 decline and/or IL-17 promotion in polyp-infiltrating Treg. Mast cell–derived tumor necrosis factor (TNF)-α was shown necessary for IL-17 induction in T cells in a mouse model of neutrophilic airway inflammation (26) and may play a similar role in polyposis. Accordingly, the same authors have shown huge TNF-α secretion by polyp-infiltrating mast cells (25). We also have data showing that mast cells can break Treg suppression and anergy in vitro and redirect Treg toward Th17 cells,¹

supporting the notion of mast cells sustaining Th17-mediated immune responses. In a mouse model of hepatocarcinoma, mast cells, massively recruited by tumor-derived stem cell factor (SCF), create an inflammatory microenvironment in which both Treg and Th17 cells can expand thus establishing immunosuppression (27). Therefore, a positive feedback between mast cells and Treg may exist in polyps, with mast cells “reprogramming” Treg toward inflammation and Treg, in turn, recalling additional mast cells. Further experiments will elucidate the kinetics of these events and the distinct signals that, acting on Treg or mast cells, initiate the proinflammatory, tumorigenic loop.

The Treg–mast cell interaction has been described at least in other two reports. We take part in describing that Treg inhibit histamine degranulation (but not TNF-α release) in mast cells, through the OX40/OX40L axis, in anaphylaxis (28). The OX40/OX40L pathway, already associated to inflammatory bowel disease (29), deserves a close look in intestinal polyposis. However, Treg preserve TNF-α production in mast cells, an autocrine mechanism known to sustain mast cell precursor proliferation and expansion in the polyps (25). Therefore, Treg may not only recruit mast cells, but also dictate the kind of response that mast cells should arrange locally to maintain inflammation and sustain IL-17 production. Another report from Noelle's group has shown that Treg-infiltrating skin allografts promote, through the production of IL-9, the recruitment of tolerogenic mast cells. Such Treg were described to be fully competent in terms of suppressive function, as capable to release granzyme B against allo-reactive effector T cells in vivo (30). It could be argued that, in both grafts and polyps microenvironments, Treg recall mast cells as valuable allies in arranging the tissue-restricted immune response, but their partnership may result in tolerance or inflammation depending on the molecular milieu (Fig. 1).

It remains to be determined whether intestinal Treg directly recruit mast cells in the polyps (possibly through IL-9) or rather promote microenvironmental changes indirectly leading to mastocytosis. Supporting the latter possibility is the observation that adoptive transfer of Treg provokes a general down-regulation of inflammatory signals and, consequently, of mastocytosis. Therefore, still unknown intermediaries may link Treg to mast cells. IL-9, one of the most important growth factor and chemoattractant for mast cells (31), has recently come in the limelight as the master cytokine characterizing a novel T helper subset, developing from naïve T cells in the concurrent presence of IL-4 and TGF-β or from committed Th2 cells upon exposure to TGF-β. The inability to differentiate this subset in vivo results in the failure to recruit mucosal mast cells and to eradicate helminth infections in the intestines (32). Future investigation will uncover the role of “Th9” cells in intestinal polyposis, a Th2-biased pathologic condition (33) in which Treg may locally provide the TGF-β necessary to skew Th2 cells into Th9 effector cells. Of note, although IL-10 production has been detected in vitro in Th9 cells, in vivo, in a mouse model of inflammatory bowel disease, they behave as proinflammatory rather than regulatory (34), suggesting that IL-10 secretion may be inhibited in the colon microenvironment by local inflammatory cues, as occurs for polyp-infiltrating Treg.

Adoptive transfer of Treg is currently considered a suitable preventive and therapeutic option for a variety of autoimmune and inflammatory diseases in humans (35). This idea is supported by many studies in mice showing the protective role of exogenous Treg, colitis included (4). Less clear is the reason why endogenous Treg, abundantly present in inflamed colonic mucosa, fail to prevent pathology. A possible explanation is that Treg, routinely activated during ongoing responses to resolve inflammation eventually, are outcompeted by persistent proinflammatory signals or even subverted into fully armed effector cells despite persistent Foxp3 expression. Very recent is the finding that adoptively transferred Treg lose Foxp3 expression and gain, in the gut, phenotype and function of follicular B-helper T cells, supporting germinal center formation and arranging the main networks of intestinal immunity (36). It would be of interest to study if exogenous Treg may modulate the germinal center response also in intestinal polyposis, a condition correlated to IgA-mediated humoral immune reactivity (33). In light of the novel findings about Treg transcriptional and functional heterogeneity, it could be suggested that the same local cues skewing the development of a specific class of effector T cells may concurrently promote an exaggerated specialization of the corresponding Treg subset, impairing the correct regulatory program. Mast cells, thought to contribute to the differentiation of different T helper subsets (18), may act also on specific Treg subsets exacerbating their plasticity.

The novel immune pathways described by Gounaris and colleagues (24) in colon polyposis offer new therapeutic targets and provide additional explanations for the efficacy of old treatments. Cyclooxygenase (COX)-2 inhibitors are one of the most promising drugs for the treatment of gastrointestinal malignancies. It should be noted that their efficacy may rely also on the inhibition of Th17 differentiation in the colon, as prostaglandin-E2 has been recently shown to promote Th17 skewing (37). Administration of tyrosine kinase inhibitors, such as imatinib mesylate, may lead to successful cancer treatment also by directly targeting SCF-mediated mast cell migration and activation in vivo, as recently shown in murine and human neurofibromatosis (38). Also, dietary restriction may help inhibiting IL-17 induction in T cells; for instance, retinoic acid assumption negatively impacts Th17 differentiation while promoting Treg induction (39). Last but not least, exposure to pollutants can alter immune homeostasis; indeed, stimulation of the aryl hydrocarbon receptor, a molecule involved in the toxicity of dioxin, increases Th17 differentiation and effector functions in autoimmunity (40).

In a mouse model of colitis-associated carcinogenesis, TNF-α blockade decreased the recruitment of inflammatory cells, the expression of COX-2 and the incidence of cancer (41). Results of phase I and II clinical trials indicate that TNF-α antagonist etanercept or the antibody infliximab have some therapeutic activity against metastatic breast cancer (42), ovarian cancer (43), renal cell carcinoma (44), melanoma, and also colorectal cancer (45), although the involvement of Treg and Th17 in such settings remains to be investigated. Of note, blockade of TNF-α induces fully suppressive Treg in patients of rheumatoid arthritis (46), while rapidly reducing the Th17 response in psoriatic plaques (47). Interestingly, murine tumor-infiltrating Treg were found to express huge levels of TNF receptor II (48), suggesting that TNF-α may exert its protumorigenic role also through Treg expansion and activation.

The mast cell–Treg interaction offers novel therapeutic targets in inflammatory disease and cancer. On one side, the inhibition of Treg-promoted mastocytosis could be obtained blocking the relevant cytokines, such as IL-17 and IL-9, likely involved in mast cell recruitment and expansion. On the other side, mast cell–induced Treg deprogramming could be prevented by the selective inhibition of proinflammatory molecules, such as TNF-α, IL-23, and IL-6, or by the generalized blockade of mast cell degranulation with a mast cell stabilizer such as cromolyn. Finally, molecules involved in direct Treg–mast cell crosstalk, such as the OX40/OX40L axis, could be targeted to functionally tune the relevant pathways at the crossroad of the innate and the adaptive immune response.

No potential conflicts of interest were disclosed.

We thank Associazione Italiana per la Ricerca sul Cancro, Italian Ministry of Health and Istituto Superiore di Sanità within the Alliance Against Cancer Program, for financial support. S. Piconese is supported by a fellowship from Fondazione Italiana per la Ricerca sul Cancro.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.