|Home | About | Journals | Submit | Contact Us | Français|
Colorectal cancer (CRC) is a leading cause of cancer-related deaths worldwide. Chronic inflammation is recognized as a predisposing factor for the development of colon cancer, but the molecular mechanisms linking inflammation and tumorigenesis remained elusive. Recent studies revealed a crucial role for the NOD-like receptor (NLR) protein Nlrp3 in regulating inflammation through the assembly of pro-inflammatory protein complexes termed inflammasomes. However, its role in colorectal tumor formation remains unclear. Here, we showed that mice deficient for Nlrp3 or the inflammasome effector caspase-1 were highly susceptible to azoxymethane (AOM)-dextran sodium sulfate (DSS)-induced inflammation and suffered from dramatically increased tumor burdens in the colon. This was a consequence of markedly reduced interleukin (IL)-18 levels in mice lacking components of the Nlrp3 inflammasome, which led to impaired production and activation of the tumor suppressors interferon-γ (IFN-γ) and STAT1, respectively. Thus, IL-18 production downstream of the Nlrp3 inflammasome is critically involved in protection against colorectal tumorigenesis.
Colorectal cancer (CRC) is one of the leading causes of cancer-related deaths. Patients with inflammatory bowel diseases (IBD), most commonly Crohn’s disease and ulcerative colitis, are at increased risk of developing colorectal cancer (1–3). Indeed, IBD is considered the third most important risk factor for the development of colorectal cancer (4). Although the precise molecular mechanism of IBD-related colorectal tumor formation is not clearly understood, existing studies suggest that chronic inflammation primes the mucosal tissue in the gut for increased cell proliferation, angiogenesis and tumor invasiveness (5). In this regard, pro-inflammatory cytokines such as interleukin (IL)-1β, IL-6, IL-18, TNF-α and interferons (IFNs) have been demonstrated to exert key roles in inducing gut inflammation and colorectal tumor formation (6–8). The synthesis and secretion of these cytokines is controlled by transcription factors of the signal transducers and activator of transcription (STAT), NF-κB and AP-1 families (6). Notably, recent evidence suggests that inhibition of NF-κB reduces tumorigenesis (9, 10).
NF-κB activation and induction of additional inflammatory signalling pathways is initiated by engagement of pathogen recognition receptors (PRRs) of the Toll-like receptor (TLR) and NOD-like receptor (NLR) families (11, 12). TLRs are membrane-bound receptors that detect pathogen-associated molecular patterns (PAMPs) in the extracellular milieu (13). The role of TLRs in the recruitment of immune cells at mucosal surfaces and in protection against tumorigenesis in the gut is well-established. For example, TLR5 activation in a mouse xenograft model of human colon cancer elicited powerful antitumor activity (14). In addition to TLRs, several members of the cytosolic NLR family have been identified as key regulators of cytokine production (11). The NLR proteins NOD1 and NOD2 mediate activation of NF-κB and MAP kinases in response to the cytosolic presence of peptidoglycan fragments. In contrast, the NLR protein Nlrp3 (also referred to as Nalp3/CIAS1/Cryopyrin) is involved in activation of the cysteine protease caspase-1 (15). Homotypic interactions between the pyrin domain in the N-terminus of Nlrp3 and the bipartite adaptor protein ASC bridge the association of caspase-1 to Nlrp3 in a large protein complex known as the ‘inflammasome’ (16). Activated caspase-1 processes the cytosolic precursors of the related cytokines IL-1β and IL-18, thus allowing secretion of the biologically active cytokines. Hence, mice lacking caspase-1 are defective in the maturation and secretion of IL-1β and IL-18 (17, 18). IL-1β participates in the generation of systemic and local responses to infection, injury and immunological challenges by generating fever, activating lymphocytes, and by promoting leukocyte infiltration at sites of injury or infection. Binding of IL-18 to the IL-18 receptor complex triggers many of the signaling pathways that are engaged by the IL-1 receptor, including activation of NF-κB, STAT1 and MAP kinases (19, 20). IL-18 (previously known as IFN-γ inducing factor) also promotes the production of IFN-γ in activated T cells and NK cells, thereby contributing to TH1 cell polarization (8, 21, 22). Finally, IL-18 was shown to induce Fas ligand production and the generation of multiple secondary pro-inflammatory cytokines, chemokines, cell adhesion molecules and nitric oxide species (23, 24).
The profound role of NLR-mediated inflammatory responses in shaping the micro-environment during colitis-associated colorectal tumorigenesis is starting to emerge. For instance, the NLR family member NOD1 is important for protection against colitis-associated colorectal tumor formation (25). In addition, defective activation of the Nlrp3 inflammasome was linked to increased susceptibility to Crohn’s disease in patients (26). Moreover, the Nlrp3 inflammasome was recently shown to confer protection against experimental colitis in mice (27–29). In this regard, mice lacking the inflammasome components Nlrp3, ASC or caspase-1 all presented with more severe clinical manifestations of colitis and suffered from increased epithelial injury, bacterial invasion and death rates (27–29). The increased susceptibility to colitis was correlated with defective IL-18 production in inflammasome-deficient mice (27, 29). Despite the role of the Nlrp3 inflammasome in controlling colitis-associated inflammation, its role in controlling colitis-associated tumorigenesis and the relevant inflammasome effector pathways in this process have remained unclear. To resolve these issues, we determined the rate of colorectal tumor formation in Nlrp3−/− and caspase-1−/− mice in the commonly used azoxymethane (AOM)-dextran sodium sulfate (DSS) model. IL-18 production downstream of the Nlrp3 inflammasome was found to exert a protective role against colorectal tumor formation. IL-18-mediated activation and induction of the respective tumor suppressors STAT1 and IFN-γ may represent a potentially critical mechanism for Nlrp3-mediated resistance against colitis-associated tumorigenesis.
Nlrp3−/−, ASC−/− and Casp1−/− mice backcrossed to C57BL/6 background for at least 10 generations have been described before (30). IL-18−/− were kindly donated by Dr. Paul G. Thomas, St. Jude children’s Research Hospital. All mice were 8–10 weeks old males and maintained in a pathogen-free facility and the animal studies were conducted under protocols approved by St. Jude Children’s Research Hospital Committee on Use and Care of Animals.
Mice were injected intraperitoneally with 10 mg/kg AOM (Sigma). After 5 days, 3% DSS (Molecular mass 36–40 kDa; MP Biologicals) was given in drinking water over 5 days followed by regular drinking water for 2 weeks. This cycle was repeated twice and mice were sacrificed 4 weeks after the last DSS cycle.
Formalin-preserved sections of cecum and colon (proximal, middle, and distal) were processed and embedded in paraffin by standard techniques. Longitudinal sections of 5 μm thick were stained with hematoxylin and eosin (H&E) and examined by a pathologist blinded to the experimental groups. Colitis scores of each segment were assigned based on the extent and severity of inflammation, ulceration, and hyperplasia of the mucosa. Severity scores for inflammation were as follows: 0 = normal (within normal limits); 1 = mild (small, focal, or widely separated, limited to lamina propria); 2 = moderate (multifocal or locally extensive, extending to submucosa); 3 = severe (transmural inflammation with ulcers covering >20 crypts). Scores for ulceration were as follows: 0 = normal (no ulcers); 1 = mild (1–2 ulcers involving up to a total of 20 crypts); 2 = moderate (1–4 ulcers involving a total of 20–40 crypts); 3 = severe (more than 4 ulcers or over 40 crypts). Mucosal hyperplasia scores were assigned as follows: 0 = normal (within normal limits); 1 = mild (Crypts 2–3 times normal thickness, normal epithelium); 2 = moderate (Crypts 2–3 times normal thickness, hyperchromatic epithelium, reduced goblet cells, scattered arborization); 3 = severe (Crypts >4 times normal thickness, marked hyperchromasia, few to no goblet cells, high mitotic index, frequent arborization). Scoring for extent of lesions: 0 = normal (0% involvement); 1 = mild (up to 30% involvement); 2 = moderate (30% to 70% involvement); 3 = severe (over 70% involvement). The individual scores from the four segments were summed, such that the maximum colitis score for a given animal is 48, and the minimum score is 0. For immunohistochemistry, formalin-fixed paraffin-embedded tissues were cut into 4 μm sections and slides were stained with antibodies against the macrophage marker F4/80 and phospho-STAT1 (Cell Signaling), respectively.
To measure the cytokine levels in colon tissue, a part of the colon was homogenized mechanically in PBS containing 1% NP-40 and a Complete protease inhibitor cocktail tablet (Roche). Mouse cytokines and chemokines in serum and colon homogenates were determined with Luminex (Bio-Rad) and ELISA (R&D Systems) assays.
Total RNA from colon tissue was isolated with Trizol (Invitrogen). First-strand cDNA was synthesized from 250 ng RNA using superscipt III (Invitrogen). Real-time PCR for COX-2 and IFN-γ was performed using SYBR Green master mix (Invitrogen) on an ABI Prism7500 Real-time PCR system (Applied Biosystems). mRNA levels were determined by means of the standard curve method. A standard sample was serially diluted and used for constructing a standard curve. Simultaneous quantification of GAPDH mRNA was used as an internal control.
The number of proliferating cells in intestinal epithelium was determined using the immunoperoxidase staining protocol with the thymidine analogue 5′-bromo-2′deoxyuridine (BrdU) as described earlier (29). In brief, 1 mg/ml BrdU in PBS was injected intraperitoneally. 3 hours later, colon tissue was collected, fixed in 10% neutral buffered formalin and embedded in paraffin. Immunohistochemistry was performed using an in situ BrdU staining kit (BD Bioscience). Tissues were counterstained with hematoxylin.
Tissue homogenates were lysed in lysis buffer (10 mM Tris-HCl, 150 mM Nacl, 5 mM EDTA, 0.1% Nonidet P-40, 0.25% sodium deoxycholate, supplemented with protease and phosphatase inhibitor cocktails (Roche)) and membranes were removed by centrifugation at 11,000 g. Before separation by SDS-PAGE, protein samples were denatured with SDS plus 100 mM DTT and boiled for 5 min. Separated proteins were transferred to PVDF membranes and immunoblotted with primary antibodies against phospho-STAT-1 (Cell Signaling), STAT1 (Cell Signaling), rabbit phospho-IκB (Cell Signaling), IκB (Cell Signaling) and β-actin (Sigma)
Recombinant IL-18 (MBL international) was injected intraperitoneally at a concentration of 0.5 μg/mouse on days 0, 2 and 4; and 0.1 μg/mouse on days 6 and 8 after DSS administration. Alternatively, Casp1−/− mice were injected intraperitoneally with recombinant mouse IFN-γ (R&D Systems) at a concentration of 200 IU/mouse on days 0, 2, 4 and 6 after DSS administration.
Data are represented as mean ± S.E. Statistical significance was determined by Student’s t-test or chi-square test. P < 0.05 was considered statistically significant.
Nlrp3-deficient mice were recently shown to be highly susceptible to the induction of inflammation and tissue damage in the acute DSS-induced colitis model (29). Similarly, Nlrp3-deficient mice developed more severe symptoms of chronic colitis when mice were administered multiple cycles of DSS (28, 29). To determine whether the increased and prolonged gut inflammation in inflammasome-deficient mice led to increased tumorigenesis, Nlrp3−/−, ASC−/− and Casp1−/− mice were administered a single dose of the DNA methylating agent AOM (10 mg/kg), followed by repeated cycles of a 3% DSS solution (31). 12 weeks after AOM injection (Fig 1A), the development of adenomatous polyps and well-formed tumors in the colons of Nlrp3 inflammasome-deficient mice was examined and compared to colons of treated wild-type mice. Mice in the wild-type, Nlrp3−/−, ASC−/− and Casp1−/− groups developed tumors after AOM and DSS administration, but tumor burdens were significantly increased in Nlrp3−/−, ASC−/− and Casp1−/− mice over wild-type mice (Fig. 1B, C). Nevertheless, statistically significant differences in tumor size could not be observed (data not shown). Most tumors were located in the distal area of the colon in all genotypes (Fig. 1D), although a fraction of the tumors were found in the mid-colon section (Fig. 1E). Notably, the number of mice presenting with tumors in the mid-colon region increased from about 40% of wild-type mice to over 60% in the Nlrp3−/− and Casp1−/− cohorts (Fig. 1F). This may be due to the more severe inflammation and extended tissue damage that DSS induced in Nlrp3−/− and Casp1−/− mice (27–29).
Colons and ceca of representative tumor-bearing wild-type, Nlrp3−/− and Casp1−/− mice were sectioned and stained with hematoxylin and eosin (H&E) to study mucosal dysplasia in more detail. Significantly more dysplastic cells and hyperplastic areas, adenomatous polyps and well-formed tumors were visible in the distal colons of Nlrp3−/− and Casp1−/− mice relative to those of wild-type mice (Fig. 2A). Moreover, the number of mice presenting with dysplastic events was significantly higher in the Nlrp3−/− and Casp1−/− cohorts (Fig. 2B). In all three genotypes, tumors were mainly derived from dysplastic epithelial cells at the site of inflammation (Fig. 2C). The tumors appeared as sessile tubuovillous adenomas, and evidence for adenocarcinoma development was not observed. Histopathological scoring for severity of inflammation, inflamed area, ulceration and hyperplasia in the colon and cecum was in agreement with significantly increased disease progression in Nlrp3−/− and Casp1−/− mice (Fig. 2D). As indicated before (Fig. 1D), most tumors were located in the distal area of the colon, but occasionally tumors were found in the mid-colon area. Separate histopathological scorings for the distal and mid-colon regions were performed to determine whether spatial differences in pathology could be observed. In both the distal (Fig. 2E) and mid-colon (Fig. 2F) regions, read-outs for inflammation severity, inflamed area and hyperplasia were significantly increased in Nlrp3−/− and Casp1−/− mice over wild-type controls. Collectively, these results demonstrate that a functional Nlrp3 inflammasome is critical for protection against colitis-associated dysplasia and tumorigenesis in the gut. These observations are in agreement with a recent report showing increased AOM/DSS-induced colon tumorigenesis in mice lacking Nlrp3 or the inflammasome effectors ASC and caspase-1 (28).
The Nlrp3 inflammasome is required for maturation and secretion of the inflammatory cytokines IL-1β and IL-18 (32). To determine whether the absence of a functional Nlrp3 inflammasome affects local IL-1β and IL-18 production in the gut during early stages of tumorigenesis, the levels of these cytokines were measured in colon homogenates of Nlrp3−/− and Casp1−/− mice 5 days after completion of the 1st DSS cycle (Day 10 post-AOM treatment). IL-1β amounts in the colon homogenates of AOM/DSS-treated wild–type, Nlrp3−/− and Casp1−/− mice remained low and barely rose above those of untreated animals (Fig. 3A and data not shown). In contrast, significant levels of IL-18 were measured in colon homogenates of AOM/DSS-treated wild-type mice (Fig. 3A). However, IL-18 levels in colon homogenates of Nlrp3−/− and Casp1−/− mice were nearly 50% lower than those of treated wild-type controls (Fig. 3A). Unlike IL-18, the levels of the cytokines IL-6, IL-12 and TNF-α did not differ significantly from those found in wild-type controls (data not shown), demonstrating the specificity of these results. Moreover, we measured significantly higher levels of the chemokines MIP-1α and eotaxin in colon homogenates of Nlrp3−/− and Casp1−/− mice (Fig. 3B, C), suggesting that deregulated IL-18 production triggered an increased recruitment of inflammatory cells in colons of Nlrp3−/− and Casp1−/− mice. In agreement, colon sections of the latter genotypes contained significantly more F4/80-positive cells than wild-type colons (Fig. 3D), indicating a dramatically increased infiltration of macrophages in colons of Nlrp3−/− and Casp1−/− mice. Macrophages exert a regulating role in the colorectal tumor micro-environment through the production of a variety of tumorigenic factors including COX-2, which promotes tumor development through the synthesis of its enzymatic product prostaglandin E(2) (33). Consistent with the increased macrophage infiltration in colons of Nlrp3−/− and Casp1−/− mice, real-time PCR analysis demonstrated significantly higher COX-2 mRNA levels in colon tissue of Nlrp3−/− and Casp1−/− mice (Fig. 3E).
We next sought to determine the effect of deregulated IL-18 production, increased macrophage infiltration and COX-2 production on epithelial cell proliferation in the colons of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice. To this end, epithelial cell proliferation was examined by BrdU staining at the early and late time points of 10 days and 12 weeks post-AOM injection, respectively. Interestingly, the number of proliferating cells located in dysplastic regions of Nlrp3−/− and Casp1−/− colons was significantly higher than in tumor tissue of AOM/DSS-treated wild-type mice at both analyzed time points (Fig. 4). In contrast, no significant differences in proliferation were noted between the three genotypes in regions of the normal mucosa (Fig. 4). AOM-induced mutagenic events are likely to be critical for inducing neoplasia in the context of the colonic micro-environment of Nlrp3−/− and Casp1−/− mice because DSS-administration alone failed to induce increased cell proliferation in colonic crypts of Nlrp3−/− and Casp1−/− (29). Together, these results suggest a critical role for Nlrp3 inflammasome-mediated production of IL-18 in protection against colitis-associated immune cell invasion and neoplasia.
To further examine the role of IL-18 in protection against colitis-associated dysplasia and tumor development, we characterized colon inflammation and tumor development in il-18−/− mice that were subjected to the AOM/DSS regimen described in figure 1A. In agreement with an important role for IL-18 in protection against colitis-associated tumorigenesis, colons of il-18−/− mice contained significantly more tumors than those of treated wild-type mice (Fig 5A, B). To determine whether increased tumor formation in il-18−/− could be linked to increased epithelial cell damage and colon inflammation during the early stages of disease, we examined phenotypic and histological signs of colitis and hyperplasia during acute colitis. To this end, wild-type and il-18−/− mice were administered AOM followed by a 3% DSS solution during 5 days before parameters of colitis development and tumorigenesis were analyzed. Il-18−/− mice presented with aggravated colitis as evidenced by higher body weight loss (Fig. 5C), severe inflammation, hyperplasia and an increased number of dysplastic cells (Fig. 5D). Semi-quantitative scoring of histological colon sections for inflammation, ulceration, affected area and hyperplasia was consistent with markedly increased disease development in il-18−/− mice relative to the group of wild-type mice (Fig 5E).
As a complementary approach to the use of il-18−/− mice, we studied whether recombinant IL-18 could reverse disease progression in AOM/DSS-treated Casp1−/− mice. Importantly, the group of Casp1−/− mice that received recombinant IL-18 lost significantly less weight compared to Casp1−/− mice that were refused the recombinant cytokine (Fig. 5F). Moreover, IL-18 administration provided protection against histological signs of inflammation and dysplasia (Fig. 5G). Consistently, semi-quantitative scoring of inflammation, ulceration, affected area and hyperplasia on histological colon sections was indicative of milder disease in IL-18-treated Casp1−/− mice (Fig. 5H). These results demonstrate that IL-18 signaling downstream of the Nlrp3 inflammasome confers protection against colitis-associated colorectal tumorigenesis.
Our results showed that impaired production of IL-18 downstream of the Nlrp3 inflammasome contributes to aggravated colitis-associated tumorigenesis (Fig. 3 and and5).5). IL-18 was initially described as the cytokine responsible for induction of IFN-γ production (22) and IFN-γ was attributed potent anti-tumor activity in a variety of experimental tumorigenesis models (34–38). In agreement with the defective IL-18 production in colons of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice (Fig. 3A), we found that IFN-γ mRNA levels were dramatically lower in colon homogenates of the latter genotypes relative to those of AOM/DSS-treated wild-type mice (Fig. 6A). Diminished IFN-γ production was confirmed at the protein level by IFN-γ-specific ELISA (Fig. 6B). IFN-γ-mediated anti-tumor signaling involves activation of the transcription factor STAT1 (36). Notably, stat1−/− mice are highly susceptible to tumorigenesis, classifying STAT1 as a tumor suppressor (39, 40). To determine whether decreased production of IL-18 and IFN-γ in colons of AOM/DSS-treated Casp1−/− mice affected STAT1 activation levels, phospho-specific STAT1 antibodies were used to examine STAT1 activation by Western blotting. STAT1 activation was a consequence of AOM/DSS treatment, because basal STAT1 activation levels in wild-type colons were significantly upregulated following AOM/DSS treatment (Fig. 6C). However, AOM/DSS-induced STAT1 activation was dramatically reduced in colons of Casp1−/− mice during early stages of tumorigenesis (day 15 post-AOM) (Fig. 6C). In contrast, phosphorylation of the NF-κB inhibitor IκB was not affected (Fig. 6D), demonstrating the specificity of these results. Immunohistochemical analysis of wild-type colons indicated increased phospho-STAT1 activation in epithelial cells and infiltrating immune cells upon AOM/DSS treatment (Fig. 6E). Colons of AOM/DSS-treated Casp1−/− mice contained significantly less cells staining positive for phospho-STAT1 (Fig. 6E), suggesting that STAT1 signaling in epithelial and immune cells may both contribute to protection against tumorigenesis.
In agreement with an important role for IFN-γ in inducing STAT1 activation, STAT1 phosphorylation was restored by treating Casp1−/− mice with 200 IU recombinant IFN-γ at days 5, 7, 9 11 post-AOM (Fig. 6F). Similarly, administration of recombinant IL-18 restored phospho-STAT1 levels in AOM/DSS-treated Casp1−/− mice to those observed in colons of AOM/DSS-treated wild-type mice (Fig. 6G). Thus, IL-18- and IFN-γmediated activation of the tumor suppressor STAT1 may play a critical role in protection against colitis-associated tumorigenesis upon activation of the Nlrp3 inflammasome.
Chronic inflammation is increasingly recognized as a critical risk factor for the development of colorectal cancer (4). Members of the NLR protein family are expressed on epithelial and professional antigen presenting cells residing in the colonic mucosa and lamina propria and play key roles in regulating the immune response against commensal microorganisms in the gut. Notably, defective activation of the NLR member NOD1 has been reported to enhance inflammatory cytokine production against commensal bacteria in the gut and prime the colorectal mucosa for increased cell proliferation and tumor formation during colitis in mice (25). Moreover, mutations in the NLR protein NOD2 are linked with the development of Crohn’s disease in humans (41, 42). It has been established that Crohn’s disease patients are at increased risk of developing sporadic colorectal cancer (43). In agreement, polymorphisms in the gene encoding NOD2 have been associated with increased susceptibility to gastrointestinal tumorigenesis (44). More recently, mutations in the gene encoding Nlrp3 were linked with increased susceptibility to Crohn’s disease in humans (26). Recent reports from our and other groups demonstrated that DSS-induced colitis in Nlrp3 deficient mice is associated with an increased destruction of the epithelial barrier in the gut, inducing systemic dispersion of colonic microflora and an exaggerated inflammatory response (27, 29). Nlrp3 plays a central role in activation of caspase-1 and secretion of the pro-inflammatory cytokines IL-1β and IL-18 (30). Caspase-1 deficient mice also were hypersensitive to DSS- and TNBS-induced colitis (27, 29), indicating that Nlrp3 protects against colitis through the production of caspase-1-dependent cytokines. Indeed, the phenotype of Casp1−/− mice was rescued by administration of recombinant IL-18 (27, 29). Moreover, mice lacking the inflammasome inhibitor caspase-12 were resistant to acute colitis, although (paradoxically) they were more susceptible to AOM/DSS-induced colorectal tumorigenesis (27).
Here, we showed that increased inflammatory responses and destruction of the epithelial barrier led to enhanced dysplasia and tumorigenesis in colons of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice. Our observations are in agreement with a recent report showing that mice lacking Nlrp3, ASC or caspase-1 were hyper-susceptible to AOM/DSS-induced colorectal tumor formation (28). However, the mechanism by which the Nlrp3 inflammasome confers protection against colitis-associated tumorigenesis remained obscure (45). We demonstrated that IL-18 production was significantly reduced in colons of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice, and more importantly, that colons of AOM/DSS-treated il-18−/− mice recapitulated the increased tumor burdens seen in mice lacking Nlrp3 or caspase-1. These results suggested a critical role for IL-18 production downstream of the Nlrp3 inflammasome in protection against colitis-associated neoplasia. In agreement, administration of recombinant IL-18 markedly reduced disease progression in AOM/DSS-treated Casp1−/− mice.
IL-18 was previously assigned an anti-tumor function in a variety of experimental tumor models (46–49). It was reported to inhibit tumor growth and angiogenesis (50–52) and associated with repair and restitution of ulcerated epithelium (53). We and others showed that IL-18 is involved in repair of the epithelial layer of the gut by maintaining proper levels of epithelial cell proliferation during the acute stage of DSS-induced colitis (27, 29). DSS-induced damage and erosion of the epithelial layer is repaired by rapid proliferation of stem cells residing at the base of crypts (54). Intriguingly, while IL-18 promotes enterocyte proliferation to repair chemically-induced injury of colonic epithelium, we showed here that it also inhibits hyperplasia during chronic stages of colitis. This apparent discrepancy may be explained by differential roles of IL-18 during the acute and chronic stages of colitis (53). Moreover, we only observed higher proliferation rates in dysplastic regions of the colon epithelium of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice, but not in non-tumor regions. These observations suggest that IL-18 exerts its protective effect in two stages. Firstly, during acute DSS-induced colitis, it contributes to restoring epithelial barrier integrity by induced controlled proliferation of stem cells at the crypt base and turnover of damaged epithelial cells. This prevents systemic dispersion of commensal microflora and the induction of exaggerated inflammatory responses. However, during remission and chronic stages of colitis, IL-18 inhibits epithelial cell proliferation in neoplastic regions of the colon epithelium. This may be achieved, at least in part, through the induction of IFN-γ production. Indeed, IL-18 was originally identified as the ‘IFN-γ-inducing factor’ (22), and IFN-γ has been described as a pleiotropic cytokine with potent anti-tumor activity (34, 37). In this regard, we demonstrated a markedly diminished production of IFN-γ in colons of AOM/DSS-treated Nlrp3−/− and Casp1−/− mice. Notably, IFN-γ signaling was previously shown to confer protection against experimental colitis (55). In agreement with a biphasic role for Nlrp3-mediated IL-18 production in colitis-associated tumorigenesis, a recent report described a biphasic role for IFN-γ during DSS-induced colitis with promotion of intestinal epithelial cell proliferation at early stages and induction of anti-proliferative responses at later stages (56) IFN-γ mediates its effect through the IFN-γReceptor (IFN-γR), which is expressed on both normal and malignant cells (57). Its biological effects are mediated by a number of intracellular signaling pathways, the best characterized of which is the JAK-STAT pathway. Once IFN-γR is activated, it phosphorylates JAK1 and JAK2, which further induces phosphorylation and nuclear translocation of STAT1. We showed that phosphorylated STAT1 levels were markedly reduced in colons of AOM/DSS-treated Casp1−/− mice, but restored upon stimulation with either IFN-γ or IL-18. These results indicate that STAT1 signaling is affected in the absence of a functional Nlrp3 inflammasome. Once in the nucleus, STAT1 binds with gamma activated sequences (GAS) in IFN-γ-responsive genes to induce transcription of genes involved in cell proliferation, differentiation and cell death (58, 59). Thus, IFN-γ-mediated STAT1 activation downstream of IL-18 may play an important role in maintaining gut homeostasis and inhibiting tumor development during colitis.
In conclusion, we characterized the role and the mechanism by which activation of the Nlrp3 inflammasome confers protection against the development of inflammation-associated colorectal tumorigenesis. We showed that Nlrp3 inflammasome-dependent IL-18 production prevents neoplastic events, possibly through the induction of IFN-γ production and STAT1 signaling. These results suggest that strategies aimed at producing or delivering mature IL-18 in the colon may prove beneficial in preventing colorectal tumor development in the context of chronic inflammation.
We thank Anthony Coyle, Ethan Grant, John Bertin (Millennium Pharmaceuticals), Gabriel Nuñez (University of Michigan) and Richard Flavell (Yale) for generous supply of mutant mice. We also thank Dorothy Bush of the Veterinary Pathology Core at St. Jude Children’s Research Hospital for technical assistance.
1This work was supported by National Institute of Health Grants AR056296 and AI088177, a Cancer Center Support Grant (CCSG 2 P30 CA 21765) and the American Lebanese Syrian Associated Charities (ALSAC) to T-D.K. ML is supported by the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen.
2 The authors declare that they have no competing financial interests.