Immunologic Research

, Volume 52, Issue 1, pp 89–99

IL-33/ST2 axis in inflammation and immunopathology


  • Marija Milovanovic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
  • Vladislav Volarevic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
  • Gordana Radosavljevic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
  • Ivan Jovanovic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
  • Nada Pejnovic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
  • Nebojsa Arsenijevic
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
    • Faculty of Medicine, Center for Molecular Medicine and Stem Cell ResearchUniversity of Kragujevac
Immunology in Serbia

DOI: 10.1007/s12026-012-8283-9

Cite this article as:
Milovanovic, M., Volarevic, V., Radosavljevic, G. et al. Immunol Res (2012) 52: 89. doi:10.1007/s12026-012-8283-9


Interleukin-33 (IL-33), a member of the IL-1 family of cytokines, binds to its plasma membrane receptor, heterodimeric complex consisted of membrane-bound ST2L and IL-1R accessory protein, inducing NFkB and MAPK activation. IL-33 exists as a nuclear precursor and may act as an alarmin, when it is released after cell damage or as negative regulator of NFκB gene transcription, when acts in an intracrine manner. ST2L is expressed on several immune cells: Th2 lymphocytes, NK, NKT and mast cells and on cells of myeloid lineage: monocytes, dendritic cells and granulocytes. IL-33/ST2 axis can promote both Th1 and Th2 immune responses depending on the type of activated cell and microenvironment and cytokine network in damaged tissue. We previously described and discuss here the important role of IL-33/ST2 axis in experimental models of type 1 diabetes, experimental autoimmune encephalomyelitis, fulminant hepatitis and breast cancer. We found that ST2 deletion enhance the development of T cell-mediated autoimmune disorders, EAE and diabetes mellitus type I. Disease development was accompanied by dominantly Th1/Th17 immune response but also higher IL-33 production, which suggest that IL-33 in receptor independent manner could promote the development of inflammatory autoreactive T cells. IL-33/ST2 axis has protective role in Con A hepatitis. ST2-deficient mice had more severe hepatitis with higher influx of inflammatory cells in liver and dominant Th1/Th17 systemic response. Pretreatment of mice with IL-33 prevented Con A-induced liver damage through prevention of apoptosis of hepatocytes and Th2 amplification. Deletion of IL-33/ST2 axis enhances cytotoxicity of NK cells, production of IFN-γ in these cells and systemic production of IFN-γ, IL-17 and TNF-α, which leads to attenuated tumor growth. IL-33 treatment of tumor-bearing mice suppresses activity of NK cells, dendritic cell maturation and enhances alternative activation of macrophages. In conclusion, we observed that IL-33 has attenuated anti-inflammatory effects in T cell-mediated responses and that both IL-33 and ST2 could be further explored as potential therapeutic targets in treatment of immune-mediated diseases.


IL-33/ST2 axisCon A hepatitisEAEMLD-STZ diabetesMouse breast cancer

IL-33/ST2 axis

Interleukin-33 (IL-33) is a member of the IL-1 cytokine family, originally described as a nuclear protein in cerebral arteries [1] and later as NF-HEV, a nuclear factor expressed in human high endothelial venules in secondary lymphoid organs [2]. Recently, IL-33 was identified as the ligand for the orphan receptor, ST2 (IL-1RL1). ST2 molecule is a member of the IL-1 receptor family [3] that exists in two forms: a transmembrane full-length form (ST2L) and a soluble, secreted form (sST2) due to differential splicing of ST2 mRNA [4]. Soluble ST2 acts as a decoy receptor for IL-33 [5]. In normal conditions, the serum concentration of soluble ST2 is below the detectable level, but elevated level of ST2 has been reported in patients with autoimmune diseases [6], asthma [7], idiopathic pulmonary fibrosis [8], myocardial infarction and heart failure [9]. ST2L is expressed by many hematopoietic cells, NK and NKT cells, mast cells, monocytes, dendritic cells and granulocytes and selectively expressed by murine and human Th2 cells but not by Th1 lymphocytes [10]. ST2 associates with IL-1R accessory protein (IL-1RAcP) to form an IL-33 receptor (IL-33R1) [11], and IL-33 signals via this heterodimer. Soluble form of IL-1RAcP interacts with the sST2–IL-33 complex to increase blocking of IL-33 signaling [12]. There is another IL-1R family member, SIGIRR, which, accompanied by ST2L (IL-33R2), negatively regulates IL-33 effects [13].

The binding of IL-33 to IL-33 receptor results in the recruitment of myeloid differentiation primary response protein 88 (MyD88), IL-1R-associated kinase 1 (IRAK1) and IRAK4 to the receptor complex in cytoplasmic region of ST2, which induces activation of various signaling proteins, including nuclear factor-κB (NF-κB), inhibitor of NF-κB-α (IκBα) and extracellular signal-regulated kinase 1 (eRK1), eRK2, p38 and c-Jun N-terminal kinase (JNK) leading to the induction of inflammatory mediators IL-1β, IL-3, IL-6, TNF, IL-5 and IL-13 [1416].

At the mRNA level, IL-33 is expressed in many organs [11] in humans and mice. However, at the protein level, IL-33 is mainly and constitutively expressed in epithelial and endothelial cells [17]. Immune cells, macrophages and dendritic cells also produce IL-33 after adequate stimulation [18]. Pathogen-associated molecular pattern (PAMP) molecules and also cytokines TNF-α, IL-1 and IFN-γ stimulate the production of IL-33 in macrophages [19, 20]. On the other hand, some proinflammatory cytokines such as TNF-α and IL-6 are also potent inducers of soluble ST2 (sST2) [21], which block the effects of IL-33. Analogous to other IL-1 family cytokines, it was proposed that IL-33 becomes activated by caspase activity, but pro-IL-33 does not have a typical cleavage site seen in pro-IL-1β and IL-18. Later findings indicated that caspase-1, caspase-3 and caspase-7 [2224] released after apoptotic cell death inactivated IL-33 and that IL-33 is active in pro-form. IL-33 has no leader sequence and it is not clear, at present, how it is released from cells [25]. It was shown that biologically active pro-IL-33 can be released by necrotic cells as “alarmin” [26]. Since IL-33 is mainly expressed in lining, epithelial and endothelial cells [17] and is released after cell damage, it is proposed to have an important role in sensing damage in various infectious and inflammatory diseases. In the absence of proinflammatory stimuli, IL-33 is localized to the nucleus [2, 27]. Additionally, IL-33, as a full-length protein, can act in an intracrine manner, translocating to the nucleus, where it binds to the chromatin and stimulates expression of IkBa, TNF-a, and C-REL [27, 28]. Since it has been shown that IL-33 is released from necrotic cells of structural but not hematopoietic origin [29], intracrine action of IL-33 could be the main way of action in immune cells.

However, it is well known that IL-33 affects the function of cells that express ST2 molecule. IL-33 polarizes naive T cells to produce Th2-associated cytokines IL-4, IL-5 and IL-13 [11] and functions as a chemoattractant for Th2 cells in vitro and in vivo [30], but also induces secretion of proinflammatory cytokines and chemokines by mast cells [16], basophils [31] and Th1 type cytokines from NK and NKT cells [31, 32]. Also, IL-33 amplifies polarization of alternatively activated M2 macrophages [33], induces maturation of dendritic cells [34] and may promote Th1-type response depending on the local cytokine milieu [35].

IL-33/ST2 axis as the first line of defense in infection

IL-33, as a full-length molecule, is mainly expressed in epithelial and endothelial cells [17, 36] especially in high endothelial venules [27], and it is proposed that IL-33 serves as the first line of defense against microbes that shape adaptive immune response. Humphreys et al. [37] were the first to show that IL-33 protects from parasitic infection. They demonstrated increased expression of IL-33 mRNA in the colon of resistant, but not susceptible mice after infection with Trichiuris muris. Exogenous IL-33 administered to susceptible mice made them resistant to T. muris infection. IL-33 acts as the initial innate signal of parasite invasion to the host that polarizes adaptive immune response toward Th2-biased response, which is host protective in this disease. It also appears that IL-33/ST2 axis affects significantly innate as well as adaptive immune response, as IL-33 was unable to induce parasite expulsion in SCID mice. However, IL-33 can exacerbate the pathology associated with chronic T. muris infection through T cell-independent mechanisms, probably increasing the production of IFN-γ in NK cells [2]. Regulatory role of IL-33 in infection is also shown during Toxoplasma gondii infection [38]. ST2 mRNA expression was upregulated, but IL-33 mRNA expression was not altered, in brain lesions of mice infected with T. gondii compared with naive mice [38]. However, ST2 knockout mice had infection with T. gondii with more severe pathology and increased mRNA levels of iNOS, IFN-γ and TNF-α in the central nervous system [37].

IL-33/ST2 signaling affects immune response to viruses. The serum levels of s ST2 protein were increased in patients infected with dengue virus [39]. Infection of mice with Theiler’s murine encephalomyelitis virus is followed by increased expression of IL-33 in the glia, which is followed by enhanced innate effector functions of glial cells, which suggested that IL-33 has important role in host defense in the CNS [38]. The role of IL-33 in antiviral defense was also shown in respiratory syncytial virus infection in mice. Treatment with monoclonal ST2-specific antibody reduced lung inflammation and disease severity in mice with Th2 type of immune response and clinical signs of bronchiolitis [40].

IL-33/ST2 axis in fulminant hepatitis

Recently, it has been demonstrated that both IL-33 and soluble ST2 were elevated in sera of patients with liver failure [41]. Therefore, we decided to dissect out the role of IL-33/ST2 axis in liver pathology by using experimental model of acute liver injury, concanavalin A (Con A)-induced hepatitis. Con A-induced liver injury is well-established murine model of T cell-mediated hepatitis [4245]. Intravenous injection of Con A induces massive necrosis of hepatocytes and immune cell activation that resembles the pathology of immune-mediated fulminant hepatitis in humans [46].

We demonstrated that IL-33/ST2 axis has protective role in Con A hepatitis [47] showing that ST2 knockout mice had more sever Con A-induced hepatitis than wild-type (WT) animals. Liver damage in ST2 knockout mice was accompanied by intrahepatic accumulation of macrophages, CD4+ and CD8+ T lymphocytes, NK and NKT cells, while the number of CD4+Foxp3+ regulatory T cells, that have protective role in Con A-induced hepatitis [48, 49], was reduced in ST2-deficient mice. Further, ST2−/− mice had altered systemic immune response toward Th1/Th17 type: we found elevated levels of systemic proinflammatory cytokines (TNF alpha, IFN gamma and IL-17) and attenuated serum level of IL-4 in ST2 knockout mice. Elevated serum level of TNF-α correlated with higher number of TNF-α producing CD4+ T cells in livers of ST2−/− mice, probably as a consequence of increased capacity of CD4+ T cells to secrete TNF in the absence of T1/ST2 as was reported in mice treated with anti-T1/ST2 monoclonal antibodies [50]. It is well known that CD8+ effector T cells are potent producers of IFN-γ [51], and that IFN-γ-producing NK and NKT cells are major effector cells involved in Con A-induced liver injury [52, 53]. Consistent with these findings, we showed massive infiltration of IFN-γ-producing CD8+T, NK and NKT cells in livers of Con A-treated ST2−/− mice. Thus, we concluded that IFN-γ, absolutely necessary in the pathogenesis of Con A hepatitis [5456], was mainly produced by CD8+T lymphocytes, NK and NKT cells. In Con A hepatitis, IL-17 has been reported to be both proinflammatory and without a direct inflammation modulating role [57, 58]. In our study, elevated serum levels of IL-17 correlated with severe liver damage and massive infiltration of IL-17-producing NKT cells in the liver. Recently published data [59] confirm that IL-17 plays a pathological role in acute liver damage. Xu et al. [59] demonstrated that a pathological role of exogenous IL-23 in Con A hepatitis depends on the production of IL-17, suggesting that in fulminate hepatitis IL-17 plays a pathological role.

In line with these observations, we showed that pretreatment of WT mice with IL-33 attenuates Con A-induced liver injury. The injection of single dose of IL-33 significantly reduced Con A-induced liver damage in wild-type BALB/c mice, prevented the recruitment of mononuclear cells into the liver and increased the influx of T regulatory cells. Further, levels of Th1 and Th17 type cytokines in the serum of IL-33 pretreated mice were decreased after Con A injection. Therefore, we assume that the main mechanism by which IL-33/ST2 pathway has a protective role in Con A-mediated liver injury is prevention of Th1/Th17 cell-mediated hepatic immune response.

It is well known that after Con A injection, polyclonally activated lymphocytes, particularly CD4+ T lymphocytes and NKT cells, stick to liver sinusoidal endothelium and destroy SECs and underlying hepatocytes [46]. IL-33 is proposed to be released as an alarmin from necrotic cells while caspase-1, caspase-3 and caspase-7 released after apoptotic cell death inactivate it [2225]. Therefore, we decided to investigate the influence of IL-33 on the expression of pro- and anti-apoptotic genes during Con A-induced liver injury. Pretreatment of WT BALB/c mice with IL-33 suppressed the activation of pro-apoptotic caspase-3 and mitochondrial BAX and enhanced the expression of anti-apoptotic Bcl-2 and p-ERK leading to the attenuation of hepatitis. We and Erhardt and Tiegs [47, 60] suggest that during acute hepatitis IL-33 is released from damaged LSECs and hepatocytes, activates cells that express ST2 molecule, particularly CD4+ Th2 lymphocytes, NKT cells and activated macrophages, shifts immune response toward Th2 type, suppresses caspase-3 activity and enhances expression of anti-apoptotic Bcl-2 and p-ERK that limits liver damage and promotes healing of the liver tissue (illustrated in Fig. 1].
Fig. 1

Proposed role of IL-33 in tissue regeneration in Con A-induced hepatitis. Upon Con A injection, resident liver (M1) phagocyte Con A produces proinflammatory cytokines TNF-α, IL-12, IL-6 and IL-1β that attract CD4+ and CD8+ T lymphocytes, NK and NKT cells. These immune cells either directly or through different soluble mediators induce apoptosis (A) or necrosis (N) of hepatocytes (H), (gray arrow). IL-33 released from necrotic hepatocytes binds to the ST2 receptor expressed on immune cells (Th2 and M2) and converts immune response toward Th2 type, and stimulates secretion of matrix metalloproteinase and arginase I from alternatively activated macrophages that promote liver regeneration (Hr) (based on [47, 60])

IL-33/ST2 axis in allergy and asthma

Asthma is a chronic inflammatory disease classically characterized by increased Th2 cytokine production. IL-33 is a strong inducer of Th2 immune responses and its role in immunopathology of asthma had been recently reviewed elsewhere [61, 62]. Higher expression of IL-33 and sST2 was found in sera and endobronchial biopsies from asthmatic patients [63] as well as in mouse models of asthma induced by ovalbumin [64, 65] compared to healthy controls. In the lung tissue, IL-33 and ST2 were mainly expressed in bronchial epithelial cells [66]. Nevertheless, the precise role of IL-33 and ST2 in mouse models of asthma is unknown. There are the opposite results found in the studies using ST2-deficient mice, mice treated with anti-ST2 proteins or IL-33-deficient mice. Massive cell influx in the lung and airway hyper-responsiveness in a murine model of ovalbumin-induced airway inflammation was found after intranasal administration of IL-33 [67, 68] and in IL-33 transgenic mice [69]. IL-33 potentiates maturation of dendritic cells upregulating the expression of CD80, CD40 and OX40L, accompanied by the release of proinflammatory cytokines, IL-6, IL-1β, TNF-α, and IL-33 also induces allergen-specific proliferation of naive T cells. IL-33 also affects migration of dendritic cells to the lymph nodes, where they can contribute to the priming of Th2 cells and the induction of allergic airway inflammation [70]. In line with these findings, IL-33 knockout mice sensitized with ovalbumin emulsified in alum showed attenuated recruitment of inflammatory cells to the lung and attenuated airway hyper-responsiveness [71]. Furthermore, application of blocking anti-ST2 antibodies or ST2-Ig fusion protein inhibited eosinophilic pulmonary inflammation and airways hyper-responsiveness [72]. In contrast to these reports, ST2-deficient mice were not protected in the ovalbumin-induced airway inflammation model [73] but have attenuated inflammation in different model of asthma. Further, an exacerbated disease was found in wild-type or Rag-1−/− mice that had undergone adoptive transfer of ST2−/− ovalbumin-specific Th2 cells [74].

The reason for these differences is not clear. It could be due to different expression of IL1RAcP in ST2-deficient mice. IL1RAcP forms receptor complex not only with ST2 but also with IL-1 receptor (IL1R) and amplifies the signal [75]. Thus, if there are no ST2 molecules on cell membranes, IL-1 signaling could be overamplificated, inducing inflammation. Or, if there is no soluble ST2 that blocks IL-33, IL-33 can bind to some other receptor and induce inflammation.

Anaphylaxis is characterized by elevated immunoglobulin-E (IgE) antibodies that signal via the high affinity Fcε receptor (FcεRI) to release inflammatory mediators [14]. IL-33 is markedly elevated in the serum of patients during an anaphylactic shock and in inflamed skin tissue of patients with atopic dermatitis. In the presence of IgE, IL-33 activates mast cells and directly induces degranulation following IgE sensitization. In animals that are systemically sensitized with IgE, IL-33 administration exacerbates antigen-induced anaphylaxis and induces the degranulation of IgE-sensitized mast cells in the skin even in the absence of antigen [14].

IL-33/ST2 axis in T cell-mediated autoimmunity

We explored the effects of IL-33/ST2 signaling in several autoimmune disorders mediated by T lymphocytes. Our findings indicate that ST2 deletion and exclusion of IL-33/ST2 axis is accompanied by enhanced susceptibility to dominantly T cell-mediated organ-specific autoimmune diseases.

BALB/c mice are relatively resistant to the induction of T cell-mediated diabetes by multiple low doses of streptozotocin (MLD-STZ). In BALB/c mice deletion of ST2 molecule leads to enhanced susceptibility to MLD-STZ-induced disease as evaluated by level of glycemia and glycosuria, number of infiltrating islet cells and β cell loss [76]. Thus, ST2-deficient mice develop insulitis and β cell loss by apoptosis after MLD-STZ induction of disease while there was minimal apoptosis of beta cells and no infiltrates in the islets of wild-type, BALB/c mice. Based on these findings, we considered that the deletion of ST2 molecule and exclusion of IL-33/ST2 axis alters naturally prevailing Th2 response in BALB/c mice and thus allows development of autoreactive Th1/Th17 cell-mediated disease such as diabetes. We found higher expression of proinflammatory cytokines TNF-α and IFN-γ in pancreatic lymph nodes of diabetic ST2-deficient mice in the early phase of disease, and in later phase constantly higher expression of TNF-α. Additionally, IL-17 was detectable in the later phase of the disease in ST2-deficient mice, but not in the draining lymph nodes of the WT mice. Based on these findings in MLD-STZ diabetes, we suggest that ST2 alters Th1/Th2 balance and leads to enhanced Th1/Th17 immune response responsible for the destruction of beta cells and diabetes.

In another model of T cell-mediated autoimmunity, experimental autoimmune encephalomyelitis, EAE, we have found that ST2/IL-33 axis has important role in regulating the encephalitogenic potential of T cells. Although BALB/c mice are resistant to EAE, ST2 deletion in BALB/c mice is accompanied by clinically and pathologically similar expression of EAE as in susceptible C57Bl/6 mice (unpublished results). Deletions of ST2 gene in BALB/c mice induced development of highly pathogenic helper cells in the induction phase of the disease and thereafter increase influx of these cells in CNS. In spinal cords as well as in brains of ST2−/− mice at the peak of the disease, we found higher numbers of CD4+ lymphocytes containing IFN-γ, IL-17, TNF-α and GM-CSF, but also myeloid cells containing IL-33, while in CNS of BALB/c mice at any time after disease induction, there was negligible number of inflammatory cells. The importance of ST2 molecule for EAE development was also shown by adoptive transfer of immune ST2−/− cells that induced the disease in BALB/c wild-type mice as in ST2−/− mice. It appears that deletion of ST2 gene facilitates the development of highly encephalitogenic T helper cells, which can be able to transfer the disease to WT BALB/c mice. Conversely T cells from the draining lymph nodes of MOG35-55 immunized WT BALB/c mice did not induce clinical disease to ST2−/− and WT recipient. These findings also suggest that ST2 expression on immune cells in CNS is not a crucial factor that controls inflammation in CNS tissue in EAE. Among MOG35–55 stimulated mononuclear cells that passively transferred EAE (ST2−/− cells), we found higher frequency of cells that contain inflammatory cytokines (IFN-γ, IL-17, TNF-α and GM-CSF) in comparison with cells isolated from WT mice that were unable to passively induce EAE. Development of T helper phenotype depends on the function of APC [77], and our next goal was to determine the effects of ST2 molecule on APC during EAE. We found that ST2 did not affect the expression of CD80, CD86 and MHCII markers but affected relative frequencies of different subpopulations of DC. Inflammatory dendritic cells, CD11c+CD11b+, that migrate to lymph nodes after initial Th1 polarizing stimulus produce abundant IL-12 and stimulate IFN-γ production in T lymphocytes [78], while CD11c+CD8+ cells act as regulatory cells that suppress CD4+ T lymphocytes [79]. Draining lymph nodes of ST2−/− BALB/c mice contain higher percentage of inflammatory type of dendritic cells and lower percentage of CD11c+CD8+ cells. Higher percentage of myeloid cells, CD11b+, isolated from ST2−/− BALB/c mice contains proinflammatory cytokines such as IL-1, IL-12, IL-6 and also IL-33. Similar results were obtained after in vitro culture of dendritic cells stimulated with TLR agonist. There were higher amounts of IL-6 and IL-23 and lower amounts of IL-10 in cell culture supernatants of ST2−/− DC compared to WT derived DC. Based on these findings, we assume that ST2 deletion alters polarization of APC that secrete predominantly inflammatory cytokines that leads to the development of highly encephalitogenic T cells.

Based on our findings, it could be suggested that ST2 gene deletion leads to the production of inflammatory innate cytokines that induce the development of dominantly inflammatory Th1/Th17 response. Additionally, the fact that strong proinflammatory type of ST2−/− antigen-presenting cells is accompanied by higher production of IL-33 suggests that IL-33 exhibits intracrine inflammatory role in APC and therefore may also promote inflammation (illustrated in Fig 2).
Fig. 2

Possible mechanism of IL-33/ST2 axis impact on autoimmune disorder development. Autoantigen challenge of ST2−/− mice in the presence of adjuvant (or STZ-induced release of autoantigens, not shown) induce proinflammatory polarization of antigen-presenting cells in the draining lymph node. ST2−/− APC after stimulation with adjuvant produce high amount of proinflammatory cytokines IL-12, IL-1, IL-6 (see text). These APC also contain high level of IL-33 that possibly could, by intracrine action, potentiate production of inflammatory cytokines. These APC present antigen to naive T cells and induce differentiation toward Th1 type cells that express T-bet and produce IL-17, IFN-γ, GM-CSF and TNF-α. These inflammatory T cells pass blood–tissue barrier, enter the target tissue, secrete inflammatory cytokines and attract other immune cells that secrete soluble factors that damage tissue. If ST2 molecule is present in APC, autoantigen induces limited production of inflammatory cytokines and in these cells production of IL-10 prevails. Thus, these APC polarize naive T cells toward Th2 type that express GATA-3. Th2 cells produce minor amount of inflammatory cytokines, do not pass blood–tissue barrier and therefore tissue damage is not seen

IL-33/ST2 axis in anti-tumor immunity

Although there are numerous findings about the role of IL-33 in inflammation, allergy and autoimmunity, there were no data about the role of ST2/IL-33 axis in anti-tumor immunity and tumor growth. Recently, we showed importance of ST2/IL-33 axis in experimental metastatic 4T1 breast cancer model in mice [80]. ST2−/− mice had delayed appearance of palpable primary tumor as well as slower tumor growth and reduced number and size of metastatic colonies in lungs and livers. ST2 deletion was accompanied by increased number of CD4+ and CD8+ T lymphocytes, enhanced cytotoxic activity in vitro of splenocytes, NK and CD8+ T lymphocytes, and also increased the numbers of IFN-γ expressing NK cells. In contrast, number of IL-10-producing NK cells was higher in wild-type mice after tumor inoculation, while undetectable in ST2-deficient mice. ST2−/− mice also have constitutively higher percentages of activated CD27highCD11bhigh NK cells and CD69+KLRG NK cells. In vivo depletion of CD8+ or NK cells revealed the key role for NK cells in enhanced anti-tumor immunity in ST2−/− mice.

The role of IL-33/ST2 axis on NK cells function is not fully understood. There are reports of IL-33 to directly stimulate [32] or indirectly amplify [31] responses of iNKT and NK cells. However, the IL-33-dependent enhancement of IFN-γ production by these cells always required the presence of IL-12. Our results appear to be at variance with these reports [31, 32]. Possible explanation of this discrepancy may be related to in vivo maturation of dendritic cells in ST2−/− mice and their effect on NK cells [81]. Dendritic cells with mature phenotype appear to be required for the functional maturation of NK cells [81]. Mayuzumi et al. [82] have recently demonstrated that conventional myeloid dendritic cells from IL-33 supplemented cultures are immature and resistant to phenotypic and functional maturation. Thus, it could be assumed that in vivo lack of ST2 signaling may facilitate maturation of dendritic cells. In fact, we have obtained data indicating that percentage and number of CD11c+CD80highCD86high dendritic cells were significantly higher in the local lymph nodes of ST2−/− tumor-bearing mice compared to WT mice [80]. Thus, it appears that IL-33/ST2 signaling facilitates primary tumor progression and metastatic dissemination probably affecting cytotoxic activity and cellular makeup of local lymph node and spleen, indicating an important regulatory role of IL-33/ST2 pathway in NK physiology and anti-tumor immunity.

It also appears that ST2 deletion affects macrophage differentiation in tumor-bearing host. Macrophages can be categorized into two main subsets in parallel with Th1/Th2 dichotomy. M1 macrophages (classically activated) are induced with IFN-γ and characterized as IL-12- and IL-23-producing cells that exert enhanced cytotoxic activity against neoplastic cells [8387]. M1 macrophages kill tumor cells, secrete high amounts of proinflammatory cytokines and activate anti-tumor immune response [86, 88], while M2 macrophages play a crucial role in type 2 immune response: promoting angiogenesis, remodeling and repairing of damaged tissues and also controlling inflammatory response by downregulation of M1-mediated functions [86, 8992]. Since ST2 is constitutively expressed on alternatively activated macrophages [33], we wondered whether IL-33/ST2 axis is involved in control of tumor development through activity of macrophages. We showed that target disruption of ST2 is associated with constitutive frequency of alternatively activated (CD206+) macrophages in the spleen [80].

Thus, IL-33 could be one of the cytokines that influence modulation of macrophages toward M2 cells, which produce IL-10 and suppress innate and adaptive anti-tumor immune responses.

Our recent data (unpublished results) show that exogenously administrated IL-33 enhances primary 4T1 tumor growth and inhibits innate anti-tumor immunity. It appears that IL-33 acts as an important amplifier of the development of alternatively activated macrophages and also markedly reduced NK cell cytotoxic activity.

Based on our findings in mouse mammary adenocarcinoma 4T1 cancer model [80, 93], it could be suggested that the absence of ST2 molecule decreases the immunosuppressive Th2-type immune response mediated by IL-33 released from epithelial, endothelial or maybe tumor cells. Then, IL-12 produced by classically activated M1 macrophages promote maturation of DC and consequently strong Th1/Th17 response that activates tumoricidal NK, NKT cells and CD8+ T cytotoxic lymphocytes (illustrated in Fig 3). In the wild-type mice, if IL-33 is overexpressed either endogenously or exogenously, it binds to ST2-positive Th2-polarized cells and also promotes generation of relatively immature dendritic cells that could induce T-regs and therefore facilitate tumor progression and metastasis.
Fig. 3

Scheme of the observed effects of IL-33/ST2 axis on tumor growth. The effects of endogenous and also exogenous IL-33 in mammary adenocarcinoma (4T1)-bearing hosts. IL-33 activates T cells toward Th2 phenotype and generates relatively immature dendritic cells that do not produce IL-12p70. Immature dendritic cells induce T-regs that contribute to an immunosuppressive environment and facilitate metastasis. Subsequently, IL-33/ST2 signaling could upregulate OX40L on dendritic cells leading to induction of IL-4, and more importantly immunosuppressive IL-10- and IL-13-producing Th2-cells that promote cancer escape. IL-33/ST2 signaling could possibly enhance the production of tumor cells factors with immunosuppressive activity. Also, IL-33 induces IL-10 production in NK cells and decreases their cytotoxicity. In the absence of ST2, IL-33 produced by epithelial cells and possibly tumor cells does not lead to the activation of Th2-associated immunosuppressive response. Concomitantly, IL-12 produced by classically activated M1 macrophages leads to the maturation of DC and induction of strong Th1-polarized immune response followed by IFN-γ production, which activates tumoricidal CTLs and NK cells. These cells with enhanced cytotoxic activity delay 4T1 tumor growth and the development of metastases (modified from [80, 93])

Therapeutic potential of IL-33 and soluble ST2

IL-33 is a dual-role cytokine. It can not only promote but also reduce inflammation depending on the tissue environment [35]. In addition, although firstly described as Th2-type promoting cytokine, now it is known that IL-33 has pleiotropic effects, it can contribute to development of Th1-type of immune response as well as enhanced IL-1 and IL-18 secretion. It implies that IL-33/ST2 axis could be considered as therapeutic target in different diseases. IL-33 can reduce inflammation depending on the tissue context, for example, in blood vessel inflammation associated with atherosclerosis [36]. Unexpectedly, application of IL-33 in established Th1/Th17 mediated inflammatory conditions such as collagen-induced arthritis exacerbated the disease [94]. Besides, in chronic inflammation, addition of IL-33 can promote fibrosis through enhancing the production of cytokines such as IL-13 and factors secreted by alternatively activated macrophages.

Studies with IL33−/−, ST2−/− mice and application of different anti-ST2 antibodies in the same model of disease indicated that absence of IL-33 do not have the same impact on disease development/attenuation as well as absence of ST2 molecule [95]. The fundamental mechanisms of the synthesis, processing, releasing and active form of IL-33 are still poorly defined [25]. In addition, it is unknown whether is there any other receptor that could bind IL-33 or weather IL-33 shares ST2 as receptor with some other cytokine. Better understanding of these processes is essential for the future studies of IL-33 as therapeutic agent.

Serum level of sST2 molecule is increased in many inflammatory conditions with anti-inflammatory effects. Treatment of mice with mixed Th1/Th2 type of inflammatory bowel disease with standard anti-TNF therapy led to attenuation of the disease that was accompanied by higher sST2/IL-33 ratio in serum [96]. Also, adenovirus-mediated overexpression of soluble ST2 protects from LPS-mediated lung injury [97]. The finding that serum sST2 increases in response to serial injection of IL-33 indicates that ST2 is induced as a negative regulator of IL-33 [95]. Thus, it seems that modulation of serum level of sST2 and its relative ratio with IL-33 should be further explored as potential therapeutic target in many inflammatory conditions. To this, we also added the evidence that ST2 deletion/blocking may enhance anti-tumor immunity and stimulate NK cell activity. This may have therapeutic implication in immune response to viruses and experimental immunotherapy of malignances.


This study is supported by grants ON175069, ON175071 and ON175103 from Ministry of Education and Science, Republic of Serbia. We thank Dr. Andrew McKenzie for providing us ST2 knockout mice and also thank Milan Milojevic for excellent technical assistance.

Copyright information

© Springer Science+Business Media, LLC 2012