The NOD-like receptor NLRP12 attenuates colon inflammation and tumorigenesis

SUMMARY

NLRP12 is a member of the intracellular Nod-like receptor (NLR) family that has been suggested to down-regulate the production of inflammatory cytokines, but its physiological role in regulating inflammation has not been characterized. We generated mice deficient in Nlrp12 and studied its role in inflammatory diseases such as colitis and colorectal tumorigenesis. We show that Nlrp12-deficient mice are highly susceptible to colon inflammation and tumorigenesis, which is associated with increased production of inflammatory cytokines, chemokines and tumorigenic factors. Enhanced colon inflammation and colorectal tumor development in Nlrp12-deficient mice are due to a failure to dampen NF-κB and ERK activation in macrophages. These results reveal a critical role for NLRP12 in maintaining intestinal homeostasis and providing protection against colorectal tumorigenesis.

INTRODUCTION

Inflammation is generally considered to be a host protective response against infection and injury (Medzhitov, 2008). However, uncontrolled inflammation is a major risk factor for the development of cancer (Grivennikov et al., 2010). Colorectal cancer is the third most common form of cancer and the second leading cause of cancer related death in developed countries (Eaden et al., 2001; Ekbom et al., 1990; Itzkowitz and Yio, 2004). Notably, patients with inflammatory bowel diseases (IBD) such as Crohn’s disease and ulcerative colitis are at increased risk for the development of colorectal cancer (Fiocchi, 1998). Although the precise molecular mechanism of IBD-related colorectal tumor formation is incompletely understood, it is widely viewed that cytokines, chemokines, matrix degrading enzymes and growth factors produced during chronic inflammation in IBD patients contribute to mutagenic transformation of colonic epithelial cells into neoplastic cells (Grivennikov et al., 2010).

Chronic inflammatory diseases of the gut are initiated by the aberrant interaction of the host immune system with commensal microflora (Goyette et al., 2007; Zaki et al., 2011). Innate immune receptors such as Toll-like receptors (TLR) at the surface of epithelial cells and immune cells initiate this inflammatory process by activating the downstream transcription factor NF-κB, which is a central mediator of pro-inflammatory cytokine and chemokine production. However, a tight regulation of NF-κB signaling is essential to maintaining a beneficial level of homeostatic interactions with the gut microflora. Therefore, deregulated NF-κB signaling may represent a key mechanism contributing to gut inflammation, colitis and colorectal tumorigenesis (Leu et al., 2003; van Vliet et al., 2005; Yu et al., 2009). Recent studies demonstrated a key role for molecules that negatively regulate NF-κB activation in maintenance of gut homeostasis. For instance, enterocyte-specific deletion of the NF-κB regulator A20 rendered mice hypersusceptible to colitis and colorectal tumorigenesis as a consequence of uncontrolled production of NF-κB-dependent pro-inflammatory cytokines (Lee et al., 2000; Vereecke et al., 2010). Similarly, mice deficient in TIR8/SIGGIR, a molecule that negatively regulates Toll-like receptor (TLR)- and interleukin (IL)-1 receptor-mediated NF-κB signaling, suffered from increased susceptibility to colitis and colorectal tumorigenesis (Garlanda et al., 2004; Xiao et al., 2010).

In addition to TLRs, the immune system makes use of a limited set of germ-line encoded pattern recognition receptors (PRRs) to induce the production of inflammatory cytokines in response to microbial components (Kawai and Akira, 2007). This includes C-type lectin receptors (CLRs), RIG-I-like receptors (RLRs), HIN-200 proteins and nucleotide binding and oligomerization domain-like receptors (NLRs) (Inohara et al., 2005; Kanneganti et al., 2007; Kanneganti et al., 2006). NLR proteins represent platform proteins that are characterized by the presence of a conserved nucleotide binding and oligomerization domain (referred to as NBD; NOD or NACHT domain) and are located in intracellular compartments (Kanneganti et al., 2007). NLRs are implicated in a multitude of innate immune signaling pathways ranging from the regulation of MAP kinase and NF-κB signaling pathways by NOD1 and NOD2, over modulation of MHC class II genes by CIITA, to the assembly of caspase-1-activating protein complexes named ‘inflammasomes’ by the NLR proteins NLRP1, NLRP3 and NLRC4 (Kanneganti et al., 2007). Unlike the above-mentioned NLRs, the in vivo role of the NLR protein NLRP12 is not clear. Notably, polymorphisms in the gene encoding NLRP12 have been linked with increased susceptibility to periodic fever syndromes and atopic dermatitis (Arthur et al., 2010; Borghini et al., 2010; Jeru et al., 2008; Macaluso et al., 2007). Moreover, NLRP12 was recently suggested to negatively regulate canonical and non-canonical NF-κB signaling in vitro by targeting the kinases IRAK1 and NIK for proteasomal degradation (Arthur et al., 2007; Lich et al., 2007; Williams et al., 2005). However, NLRP12 missense mutations in periodic fever syndrome patients were recently linked to increased caspase-1 activation rather than to inhibition of NF-κB signaling (Borghini et al., 2010; Jeru et al., 2010). Therefore, the physiological relevance of NLRP12-mediated regulation of NF-κB pathways remains to be defined. In this study we focused on determining the physiological role of NLRP12 in regulating inflammation in a mouse model of colitis and colorectal tumorigenesis.

RESULTS

Generation of Nlrp12-deficient mice

Mouse NLRP12 shares with other NLRs a structural composition that exists of an amino-terminal Pyrin motif, followed by a central nucleotide-binding domain (NBD) and a C-terminal leucine rich repeat domain (Figure S1A). The product is encoded on mouse chromosome 7 and contains 10 exons spanning a region of 28.3 kb. We initially investigated the expression pattern of murine Nlrp12 transcripts in a variety of primary immune cell populations. Cells with the highest expression levels of Nlrp12 mRNA were neutrophils and T cells, followed by dendritic cells and macrophages, respectively (Figure 1A). Recently, it was reported that NLRP12 is expressed in the colon tissue (Lech et al., 2010). We further verified the expression of Nlrp12 in different parts of the gastrointestinal tract and lymphoid organs. Nlrp12 was found to be expressed in the small intestine, caecum, colon, spleen, liver and mesenteric lymph nodes (MLN) (Figure 1B). To define the role of NLRP12 in regulating inflammatory responses in the gut, we generated Nlrp12-deficient mice by homologous recombination. To this aim, exon II encoding the Pyrin domain of Nlrp12 – which is required for the recruitment of downstream effectors and functions of the protein – was replaced with a neomycin selection cassette in the targeting construct (Figure S1B, C). Positive embryonic stem (ES) cells were used to generate chimeric mice and Nlrp12^−/− mice were backcrossed to the C57BL/6 genetic background for 10 generations. Nlrp12^−/− mice were fertile and appeared healthy when housed in a specific pathogen-free environment.

Figure 1. Nlrp12-deficient mice are hypersusceptible to DSS-induced colitis.

(A) Real time PCR analysis for mRNA of Nlrp12 in macrophages, dendritic cells (DC), PMN and T cells collected from wild-type mouse. (B) Real time PCR analysis for mRNA of Nlrp12 in liver, spleen, small intestine (SI), cecum, colon and MLN. (C, H) Wild-type and Nlrp12^−/− mice were fed a 3% DSS-solution in drinking water for 6 days followed by regular drinking water until the end of the study. (C) Body weight, (D) clinical score (stool consistency plus rectal bleeding score) were scored daily. Data represent means ± SEM; n=20/group; _*, p<0.05; _{* *}, p<0.01. (E, F) Mice were sacrificed on day 5 and 9 to measure colon length. Histopathological changes in colon tissue collected at day 5, 9, and 20 were examined by H&E staining. (G) Semi-quantitative scoring of histopathology. (H) Representative images of H&E staining of colon section at day 0, 5, 9 and 20 after DSS administration. Scale bar; 50 μM. Data represent means ± SEM. _*, p<0.05; _{* *}, p<0.01. See also Figure S1.

Nlrp12-deficient mice are hypersusceptible to DSS-induced colitis

To define the role of NLRP12 in colitis-induced inflammation, Nlrp12^−/− mice were fed 3% DSS in drinking water for 5 days and susceptibility was monitored by measuring body weight, assessing stool consistency and rectal bleeding, and measuring colon length during both the acute (day 5) and recovery (days 9–20) stages of disease. Notably, disease progression and clinical scores of wild-type and Nlrp12-deficient mice were not statistically different during the acute phase of disease. However, wild-type mice started to recover once DSS was omitted from the drinking water, whereas Nlrp12-deficient mice suffered from continued body weight loss (Figure 1C), diarrhea and rectal bleeding (Figure 1D). This inflammatory phenotype was further evidenced by the gross appearance of the colon. During the acute stage of colitis (at day 5), colons of wild-type and Nlrp12-deficient mice appeared similar (data not shown). At day 9, however, colons of Nlrp12-deficient mice were significantly shorter than those of wild-type mice (Figure 1E, F). To determine whether recovery in Nlrp12^−/− mice was simply delayed, or whether NLRP12-deficiency prevented healing responses at later time points as well, colon length and weight of Nlrp12^−/− mice were analyzed at day 20 after DSS-induced colitis. Notably, colons of Nlrp12^−/− mice were significantly shorter and weighted more than those of wild-type mice (Figure S1D, E), suggesting that Nlrp12^−/− mice continued to suffer from colon inflammation 2 weeks after DSS administration was stopped. Consistent with an inflammatory phenotype, MLN and spleens of DSS-fed Nlrp12^−/− mice were found to be significantly heavier and enlarged at day 20 compared to those of wild-type mice (Figure S1F, G).

To obtain further evidence of sustained inflammation in Nlrp12-deficient mice, colon tissue was analyzed histologically at days 5, 9 and 20 following DSS administration. Consistent with the clinical parameters discussed above, colons of Nlrp12-deficient mice contained markedly more infiltrating inflammatory cells, and displayed significantly more ulceration and hyperplasia during the recovery phase of the disease (at days 9 and 20), but not at early stages (Figure 1G, H). In agreement, colon tissue of Nlrp12-deficient mice contained significantly higher levels of pro-inflammatory cytokines than wild-type mice at day 9 (Figure S1H), but not at day 5 post-DSS administration (data not shown). Together, these results indicate that NLRP12 plays a critical role in resolving the inflammatory response following DSS-induced injury of the colonic epithelium.

NLRP12 suppresses colitis-associated tumorigenesis

The observation that Nlrp12-deficient mice suffered from sustained gut inflammation upon DSS-treatment, prompted us to investigate the role of Nlrp12 in colitis-associated tumorigenesis. To this aim, we induced colon tumorigenesis by injecting a single dose of the DNA-methylating agent azoxymethane (AOM), which was followed by 3 cycles of 3% DSS-administration (Figure 2A). Changes in body weight were monitored daily throughout the study duration and colonic tumor burden was determined 12 weeks after AOM/DSS treatment. Nlrp12-deficient mice lost significantly more body weight relative to wild-type mice (Figure 2B) and showed increased rectal bleeding (Figure S2A). Consistently, Nlrp12-deficient mice had significantly higher tumor burdens in the colon, although tumor size was not significantly different (Figure 2C–E). Tumors in wild-type mice were mostly contained within the colorectal and distal areas of the colon, whereas tumors in Nlrp12-deficient mice were commonly found throughout the entire colonic tract (Figure S2B–C). Increased tumor burdens in Nlrp12-deficient mice were associated with more inflammation and hyperplasia (Figure 2F, G), and a higher incidence of dysplasia (Figure 2H). Histological examination of tumors and adenomatous polyps showed that all Nlrp12-deficient mice included in the study developed high-grade dysplasia, of which around 30% were classified as adenocarcinoma (Figure 2H, I). By contrast, only 20% of the wild-type cohort displayed high-grade dysplasia in the colon, and adenocarcinoma development was not evident in this group (Figure 2I). Collectively, these results indicate a critical role for NLRP12 in protection against colitis-associated tumorigenesis.

Figure 2. Increased colitis-associated colorectal tumorigenesis in Nlrp12-deficient mice.

(A) WT (n=16), Nlrp12^−/− (n=14), were injected with AOM on day 0 (d0), and were then given a 3% DSS solution during three 5-day cycles as described in Methods. (B) Body weight changes were monitored throughout the study. (C) 12 weeks after AOM injection, mice were sacrificed to determine tumor development in the colon. Number of tumors in the whole colon was counted (D), and diameter of the tumors (E) was measured at the end of the study. At the same time colons were processed for histopathological analysis. (F–G) H&E stained sections were scored for inflammation, ulceration, hyperplasia and inflamed area. (H) Overall grading of dysplasia in each genotype. (I) Representative high magnification images showing dysplasia in colon tissue. Scale bar; 100 μM. Data represent means ± SEM; ^*, p<0.05, ^** p<0.01. See also Figure S2.

NLRP12 dampens inflammatory responses after colitis induction

The splenomegaly of DSS-fed Nlrp12-deficient mice (Figure S1G) was also apparent in AOM/DSS-administered animals (Figure S3A, B). We therefore hypothesized that NLRP12 may protect from colitis-associated tumorigenesis by dampening immune cell activation and inflammatory responses in response to DSS-treatment. To characterize this possibility in additional detail, we carefully examined the histopathological changes that occur during early stages of tumor induction (at day 15 after AOM injection). In line with our hypothesis, histological analysis of colon sections revealed markedly more tissue damage, mucosal edema, inflammation and hyperplasia in Nlrp12-deficient mice than in wild-type mice (Figure 3A, B). Semi-quantitative scores for colon inflammation, ulceration and hyperplasia were all significantly higher in Nlrp12-deficient mice (Figure S3C). Moreover, Nlrp12-deficient colons showed increased infiltration of macrophages, PMNs and T cells (Figure 3C). Hyper-infiltration of immune cells in Nlrp12^−/− mice was not confined to inflamed sections of the colon, but extended to the entire colon as evidenced by an increased F4/80-staining in relatively non-inflamed parts of the Nlrp12^−/− colon (Figure 3D).

Figure 3. Nlrp12-deficiency leads to increased and prolonged inflammatory responses in colon tissue.

Wild-type and Nlrp12^−/− mice were injected with AOM and 5 days later mice were fed with 3% DSS in drinking water for 5 days. Distal colons were collected at day 15 after AOM injection and examined after H&E staining. (A) Representative images of H&E stained sections of distal colon. Scale bar; 200μM. (B) Total histological scores of whole colon at the same time. Data represent means ± SEM; n=5/group; ^*, p<0.05. (C) Colon tissue was immunostained for the macrophage marker F4/80, PMN marker Gr-1 and T cell marker CD3. Scale bar; 100μM. (D) F4/80 staining in non-inflamed part of the colon at day 10 after DSS. Scale bar; 20μM. (E–G) In a separate experiment, mice were fed with 3% DSS for 5 days followed by normal drinking water until day 20. Colon, MLN and spleen were collected at day 5, 9, and 20 after DSS. Colonic lamina propria cells, MLN cells, and splenocytes were analyzed flow cytometrically after staining for myeloid cell markers CD11b, F4/80, CD11c and Gr-1. Total number of CD11b⁺, F4/80⁺, CD11c⁺, and Gr-1⁺ cells at day 5 (E), day 9 (F), and day 20 (G). Data represent means ± SEM; n=5/group; ^*, p<0.05; ^**, p<0.01. See also Figure S3.

To further characterize the immune cell types associated with the induction of hyper-inflammatory responses in the colon of Nlrp12^−/− mice, myeloid cells present in the colonic lamina propria were isolated at different stages of colitis and analyzed by flow cytometry. During early stages of colitis (at day 5), neutrophils were the most prevalent cell type found in the lamina propria of both wild-type and Nlrp12^−/− mice, but their number was nearly doubled in Nlrp12-deficient colons (Figure 3E). Notably, at later stages of colitis (days 9 and 20), cell counts of all analyzed myeloid cell types (CD11b⁺, F4/80⁺, CD11c⁺, Gr-1⁺) in Nlrp12^−/− mouse colons were significantly higher than in wild-type colons (Figure 3F, G). At day 20, a similarly dramatic increase in infiltration of myeloid cell populations was evident in the mesenteric lymph node (MLN) and spleen of Nlrp12^−/− mice, although myeloid cell counts in these tissues were comparable to those of wild-type mice at earlier time points (Figure 3G and Figure S3D). Notably, number and frequency of myeloid cell populations in untreated control mice of wild-type and Nlrp12^−/− background were not different (data not shown). In addition to being more prevalent, a larger number of CD11b⁺ myeloid cells that were collected from the spleen and MLN of Nlrp12^−/− mice at day 20 produced IL-6 and TNF-α in response to LPS and following PMA plus ionomycin stimulation (Figure 4A). Moreover, the mean fluorescent intensity (MFI) for intracellular expression for IL-6 and TNF-α was significantly higher for Nlrp12-deficient CD11b⁺ cells than for wild-type cells, indicating that Nlrp12^−/− myeloid cells produced higher levels of these pro-inflammatory cytokines (Figure 4B). Thus, together these results indicate that NLRP12 plays a critical role in dampening the inflammatory response in myeloid cells and during DSS-induced colitis.

Figure 4. Nlrp12-deficiency leads to enhanced cytokine production.

WT and Nlrp12^−/− mice were fed with 3% DSS for 5 days followed by normal drinking water until day 20. MLN and spleen were collected at day 20 after DSS administration and cells were stimulated with either LPS (1 μg/ml) or PMA (25ng/ml) plus ionomycin (500ng/ml) for 5h. After stimulation cells were stained for myeloid cell marker CD11b, and intracellular cytokines IL-6 and TNFα. (A) Frequency of CD11b⁺TNFα⁺ and CD11b⁺IL-6⁺ cells. (B) Mean fluorescent intensity (MFI) of IL-6⁺CD11b⁺ and TNFα ⁺ CD11b⁺ cells in LPS stimulated MLN and spleen. Data represent means ± SEM; n=5/group; ^*, p<0.05, ^** p<0.01. In a separate experiment, wild-type and Nlrp12^−/− mice were injected with AOM and 5 days later mice were fed with 3% DSS in drinking water for 5 days. (C) Whole colons were collected at day 14 and day 18 after AOM injection and homogenates were used to determine cytokines by multiplex ELISA. Data represent means ± SEM; n=7–8/group; ^*, p<0.05, ^** p<0.01.

Enhanced cytokine and chemokine production in Nlrp12-deficient mice drives hyperplasia and tumorigenesis

Consistent with the enhanced infiltration and hyper-activation of myeloid cells in the absence of NLRP12, the production of pro-inflammatory cytokines such as IL-1β, IL-6, TNF-α, IL-17, IL-15 were all found to be elevated in the colon of Nlrp12^−/− mice relative to the levels found in wild-type mice (Figure 4C and data not shown). Similarly, colons of Nlrp12-deficient mice contained higher levels of the chemokines, G-CSF, eotaxin, KC, IP-10, MIP-1α, MIP-1β, MIP2 (Figure 4C and data not shown). Given the association of higher tumor burdens with enhanced production of pro-inflammatory cytokines and chemokines in Nlrp12^−/− mice, we hypothesized that unlike most NLRs, NLRP12 may operate as a negative regulator of inflammatory signaling pathways. Such mechanism may also explain its antitumor function because increased cytokine and chemokine production, along with tumorigenic growth factors, may create a microenvironment that promotes unwarranted cell proliferation and adenomatous polyp development in Nlrp12^−/− mice.

To understand the nature of the tumorigenic signals that are deregulated in Nlrp12^−/− mice, we first studied apoptosis induction in colon tissue of AOM/DSS-treated mice. However, mRNA and protein expression of the anti-apoptotic protein Bcl-X_L as well as the number of TUNEL-positive cells in colon tissue of wild- type and Nlrp12^−/− mice were comparable (Figure S4). We next analyzed the expression of multiple cytokines and tumorigenic factors such as cycloxygenase 2 (COX2) and inducible nitric oxide synthase (iNOS /NOS2) in colon tissue because these factors often drive tumorigenesis. The transcript level of pro-inflammatory cytokines such as IL-6 and TNF-α and the chemokine MIP2 were markedly more induced in Nlrp12^−/− mice relative to the levels in wild-type colon (Figure 5A). Unlike IL-6 and TNF-α, the levels of the tumor-suppressing cytokine IFN-γ and its effector IFN-γ-dependent NOS2 transcripts were not significantly changed (Figure 5A). In contrast, we measured 3-fold higher mRNA levels of the prostaglandin-synthesizing enzyme COX2, which has previously been linked to colon carcinogenesis (Buchanan and DuBois, 2006; Shiff and Rigas, 1999). Moreover, elevated COX2 and MIP2 expression was also evident in colonic tumors of Nlrp12-deficient mice (Figure 5B).

Figure 5. Enhanced pro-inflammatory cytokine levels in colon tissue of Nlrp12-deficient mice augments production of tumor-inducing factors.

Wild-type and Nlrp12^−/− mice were injected with AOM and 5 days later mice were fed with 3% DSS in drinking water for 5 days. (A) Distal colons collected at day 15 after AOM injection were used to isolate RNA for expression analysis of IL-6, TNF-α, IFN-γ, MIP2, COX2 and NOS2 by real time PCR. Data represent means ± SEM; n=5–7/group; ^*, p<0.05, ^** p<0.01. (B) Real time PCR analysis for COX2 and MIP2 in distal colon tumors collected at 12 weeks after AOM/DSS treatment. Data represent means ± SEM; n=5–7/group;^** p<0.01. (C) Whole colon tissue homogenates collected at 15 days after AOM injection (10 days after DSS) were examined for activation of NF-κB, p38, ERK and STAT3 by Western blot analysis. Each lane corresponds individual mouse. (D) Densitometric analysis of pIκBα, p-ERK and p-STAT3 relative to their total protein. Data represent means ± SD; n=5/group; ^*, p<0.05. In a separate experiment, mice were injected i.p. with BrdU either at day 15 or 12 weeks after AOM injection. Colon sections were stained for BrdU-positive cells and macrophages. (E) Representative images of BrdU staining at 15 days and 12 weeks after tumor induction. Scale bar; 50μM (F) Number of BrDU-positive cells at day 10 after DSS treatment (day 15 after AOM). Data represent means ± SEM; n=4 mice/group; 3 different section per mouse; 50 field per section; ^*, p<0.05. (G) Serial sections of Nlrp12^−/− mouse colon at day 10 (day 15 after AOM) were stained for macrophage marker F4/80 (lower panel) and BrdU-positive cells (upper panel). Scale bar; 20μM. See also Figure S4.

The expression of tumorigenic and pro-inflammatory genes is modulated by signal transduction pathways mediated by NF-κB, MAPK, STAT and AKT proteins. To understand whether these pathways and molecules were deregulated in the absence of Nlrp12, we examined activation of NF-κB, MAPK and STAT signaling pathways by Western blotting. Indeed, significantly higher activation levels of NF-κB, ERK and STAT3 were observed in Nlrp12-deficient colon tissue at day 15 after AOM injection (day 10 after DSS) relative to the levels found in wild-type mice (Figure 5C, D). Hyper-activation of these inflammatory pathways is associated with an increased proliferation of epithelial cells in hyperplastic colon regions of AOM/DSS treated Nlrp12-deficient mice (Figure 5E, F). By contrast, untransformed colon tissue and unaffected colon regions of both Nlrp12^−/− and wild-type mice displayed similar proliferation levels as measured by BrdU staining (data not shown). Notably, hyperplastic colonic tissue in Nlrp12-deficient mice was surrounded by a massive infiltration of macrophages (Figure 5G), suggesting that myeloid cells may provide signals that promote tumorigenesis in the absence of NLRP12 signaling.

NLRP12 signaling in the hematopoietic compartment is critical for protection against colitis and colitis-associated tumorigenesis

To determine the cell populations that contribute to NLRP12-mediated protection against tumorigenesis, we generated 4 groups of Nlrp12 bone marrow chimeras (Figure 6A). Eight weeks after bone marrow reconstitution, mice were subjected to tumor induction using AOM plus DSS. NLRP12 deficiency in either compartment led to increased body weight loss compared to wild-type mice (Figure 6A), suggesting that NLRP12 may contribute to protection against colitis in both immune and non-immune cells. However, the body weight of Nlrp12-deficient mice with wild-type hematopoietic cells later recovered to level similar to those of wild-type mice, whereas mice lacking NLRP12 in the hematopoietic compartment failed to do so (Figure 6A). This suggests that NLRP12 signaling in immune cells is critical to controlling colitis, while its role in epithelial cells may be redundant. Consistently, the groups lacking Nlrp12 in the hematopoietic compartment had significantly higher tumor burdens and shortened colons (Figure 6B, C). On the other hand, no significant differences in tumor burdens and colon length were observed between wild-type and Nlrp12-deficient mice having wild-type bone marrow cells (Figure 6B, C). In agreement, colon tissue of chimera groups with Nlrp12-deficient immune cells that was collected at day 15 after AOM injection (day 10 after DSS) showed increased NF-κB and ERK activation (Figure 6D, E). Together, these results suggest that Nlrp12-deficient myeloid cells, particularly macrophages, fail to silence NF-κB and ERK signaling pathways, which ultimately results in elevated cytokine levels, inflammatory responses and colon tumorigenesis during colitis.

Figure 6. NLRP12 signaling in hematopoetic cells is critical for protection against colon tumorigenesis.

Bone marrow chimera mice for Nlrp12 were generated as described in Methods. 8 weeks after, mice (8–10/group) were induced tumorigenesis using AOM/DSS regimen as described in Methods. (A) Body weight was monitored every alternate day. (B–C) Tumor burden in whole colon and length of colon were measured at 12 weeks after tumor induction. Data represent means ± SEM; ^*, p<0.05. (D) In a separate study, colon tissue collected at 15 days after AOM injection were processed for Western blot analysis for IκB, p-IκB, ERK and p-ERK. (E) Densitometric analysis of p-IκBα and p-ERK relative to their total protein. Data represent means ± SD; n=3/group; ^*, p<0.05. ^**, p<0.01.

NLRP12 negatively regulates NF-κB and ERK signaling in macrophages

To gather further evidence for deregulated NF-κB and ERK signaling in Nlrp12-deficient macrophages, bone marrow-derived macrophages from wild-type and Nlrp12-deficient mice were stimulated with LPS, and activation of NF-κB, MAPK and STAT3 was analyzed by Western blotting at different time points. In agreement with the notion that NF-κB activation is enhanced in Nlrp12-deficient cells, both IκBα degradation and phosphorylation were found to be increased over a range of time points covering 10 min and up to 2 h after LPS exposure (Figure 7A, B). Furthermore, enhanced phosphorylation of ERK was apparent in LPS-stimulated Nlrp12-deficient macrophages (Figure 7A). In contrast, activation levels of STAT3 and the MAP kinases p38 and JNK were comparable to those of wild-type macrophages (Figure 7A and data not shown), confirming the specificity of these results. Enhanced NF-κB and ERK activation in Nlrp12-deficient macrophages was not due to deregulated inflammasome activity, because caspase-1 activation in response to the inflammasome activators LPS plus ATP, and upon infection with Salmonella or Listeria was not affected (Figure S5A).

Figure 7. NLRP12 negatively regulate NF-κB and ERK activation in macrophages.

Bone marrow derived macrophages were cultured as described in Methods and were stimulated with (A) LPS (1 μg/mL), (C) Pam3-CSK4 (5 μg/mL) (E) poly(I:C) (1 μg/mL). Cell lysates collected at indicated time points were analyzed for p-IκB, I-κB, p-P38, pERK, ERK, p-JNK and JNK by Western blotting. Densitometric analysis of band intensity of p-IκBα and p-ERK relative to IκBα and ERK of LPS (B), Pam3CSK4 (D) and Poly(I:C) (F) treated cells. (G) LPS-stimulated wild-type and Nlrp12-deficient macrophages were collected at different time points and mRNA was isolated for real-time quantitative PCR analysis of IL-6, KC, TNF-α, MIP2, COX2 and NOS2 synthesis. Data represent means ± SD of triplicate wells; ^*, p<0.05; ^**, p<0.01. See also Figure S5.

We next addressed the question whether the increased activation of NF-κB and ERK in Nlrp12-deficient macrophages is confined to the TLR4 ligand LPS, or a general response seen with other TLR ligands as well. Wild-type and Nlrp12^−/− macrophages stimulated with the TLR2 ligand Pam3-CSK4 and the TLR3 ligand poly(I:C) showed that ligation of both receptors induced enhanced NF-κB and ERK activation in Nlrp12-deficient macrophages (Figure 7C–F), indicating a role for NLRP12 in negatively regulating NF-κB and ERK signaling downstream of multiple TLRs. In agreement with the in vivo data presented earlier, expression of IL-6, KC, TNF-α, COX2 and MIP2 were all significantly increased in LPS-stimulated Nlrp12-deficient macrophages at the mRNA (Figure 7G) and protein levels (Figure S5B).

Previous studies suggested NLRP12 to suppress canonical NF-κB activation by preventing IRAK1 phosphorylation, and to downregulate non-canonical NF-κB signaling by inducing proteosomal degradation of NIK (Lich et al., 2007; Williams et al., 2005). To determine whether the canonical and/or non-canonical NF-κB signaling was altered in Nlrp12-deficient macrophages upon LPS stimulation, we analyzed the phosphorylation status of the non-canonical effector P100, and that of P105 for canonical NF-κB signaling. Interestingly, the levels of phosphorylated P105 were markedly induced in Nlrp12-deficient macrophages after LPS-stimulation (Figure 8A). In contrast, the levels of phosphorylated P100 appeared comparable in LPS-stimulated wild-type and Nlrp12^−/− macrophages (Figure 8A). To further confirm the suppressive effect of NLRP12 on the canonical NF-κB pathway, P105 and P100 phosphorylation, and translocation of RelA (p65) and RelB were analyzed in the cytosolic and nuclear fractions of LPS-stimulated macrophages, respectively. As shown in Figure 8B, phosphorylated P105 levels were markedly higher in Nlrp12^−/− macrophages than in wild-type cells. Moreover, increased amounts of RelA were detected in the nucleus of Nlrp12-deficient macrophages that had been stimulated with LPS (Figure 8C). Finally, we measured the DNA-binding activity of p65 in nuclear isolates of LPS-stimulated macrophages. A significantly higher p65 DNA-binding activity was measured in nuclear lysates of LPS-stimulated Nlrp12-deficient macrophages (Figure 8D). Although our results do not exclude the possibility of NLRP12 regulating non-canonical NF-κB signaling in response to TNF-α receptor family ligands such as CD40L (Lich et al., 2007), it confirms that NLRP12 potently downregulates TLR-induced activation of canonical NF-κB signaling, as previously suggested (Jeru et al., 2010). In conclusion, NLRP12 dampens inflammation and colon tumorigenesis by attenuating activation of NF-κB and ERK signaling in myeloid cells. In the absence of NLRP12, upregulated production of inflammatory cytokines and tumorigenic factors drives the transformation of epithelial cells and supports colitis-associated tumorigenesis.

Fig 8. NLRP12 is a negative regulator of canonical NF-κB signaling pathway in response to LPS.

Bone marrow derived macrophages from wild-type and Nlrp12^−/− mice were stimulated with LPS (1 μg/mL). Total cell lysate (A) and cytoplsmic fraction (B) were analyzed for phosphorylation of P100 and P105 by Western blotting. (C) Western blot analysis of NF-κB complex component RelA and RelB in nuclear fraction. (D) 5 μg nuclear protein was assayed for p65 DNA binding activity using commercial ELISA kit as described in Methods. Absorbance at 450 nm represents corresponding DNA binding activity of P65 (RelA) in the nuclear extract. Data represent means ± SD of triplicate wells in ELISA assay of a single sample; ^**, p<0.01.

DISCUSSION

Our studies demonstrate that NLRP12 plays an essential role in the suppression of proinflammatory cytokines and chemokines by controlling the activation of NF-κB and ERK pathways in response to microbial components and in colitis and colorectal tumorigenesis. Colorectal tumorigenesis is a leading cause of cancer related death. IBD is a predisposing factor of colorectal cancer (Itzkowitz and Yio, 2004). Chronic colitis develops due to hyperactivation of immune cells upon permeabilization of the colonic epithelial barrier in genetically susceptible hosts (Hill and Artis, 2010). Inflammatory processes initiated upon detection of commensal flora by NLRs and other PRRs primarily aim to control the infection and to restore the damage to the epithelial layer (Medzhitov, 2007). However, a tight regulation of inflammatory signaling pathways is critical to maintain immune responses at homeostatic levels. Excessive inflammation is destructive by itself and activates cells of the adaptive immune system, which may ultimately result in the development of autoimmunity (Liew et al., 2005). Therefore, the mechanisms controlling inhibition of NF-κB and other inflammatory signaling pathways are equally important as those driving inflammation. Indeed, deletion of negative regulators of NF-κB such as A20, TIR8 and DUBA was previously shown to lead to increased susceptibility to DSS-induced colitis (Garlanda et al., 2004; Gonzalez-Navajas et al., 2010; Lee et al., 2000; Vereecke et al., 2010; Xiao et al., 2010). Similarly, we showed here that Nlrp12 deficiency leads to increased susceptibility to colitis and colitis-associated tumorigenesis in mice by a failure to dampen inflammatory signaling pathways.

In this study we propose that NLRP12 activity in macrophages plays a major role in attenuating colon inflammation and tumorigenesis in mice. Our interpretation is based on several evidences. At first, we demonstrated that NLRP12 activity in the mouse myeloid compartment is critical in protection against colitis-associated colon tumorigenesis. Secondly, Nlrp12-deficient macrophages are hyper-responsive to TLR ligands, showing increased activation of NF-κB and ERK. Increased macrophage density in tumor tissue is strongly linked with poor prognosis of human cancer (Chen et al., 2005). A growing body of evidence suggests that activated macrophages in human colorectal tumors produce tumor promoting cytokines such as IL-6 and TNF-α, chemokines KC and MIP2, enzymes matrix metaloprotinses, COX2 and NOS2, and growth factors (Qian and Pollard, 2010). Notably, NLRP12 is expressed in both human and mouse monocytic cells (Lich et al., 2007; Williams et al., 2005). It was shown that human monocytic cells having mutation in NLRP12 are hyperinflammatory in nature and linked to inflammatory disease periodic fever syndrome (Jeru et al., 2008). Mutations in other NLR genes such as NOD2 and NLRP3 were previously shown to be associated with autoinflammatory diseases and IBD (Hugot et al., 2001; Maeda et al., 2005; Miceli-Richard et al., 2001; Ogura et al., 2001; Schoultz et al., 2009; Villani et al., 2009; Zaki et al., 2010). Therefore, hyperinflammatory nature of Nlrp12-deficient macrophages as seen in this study of mouse model of colorectal tumorigenesis suggests a critical role of NLRP12 in the protection of human colon inflammation and colorectal cancer.

Several tumorigenic factors produced by tumor-associated macrophages are regulated by signaling pathways NF-κB, STAT3 and ERK. The critical role of NF-κB in both mouse model and human colorectal tumorigenesis is well-known (Greten et al., 2004; Luo et al., 2004; Terzic et al., 2010). Recent studies in a mouse model of colorectal tumorigenesis demonstrated that both ERK and STAT3 activation is an integral part of tumor development (Bollrath et al., 2009; Kujawski et al., 2008; Lee et al., 2010). ERK regulates several tumorigenic factors such as COX2, by activating oncogenic transcription factor cMyc (Wilkins and Sansom, 2008). STAT3 regulates proinflammatory cytokines IL-17 and IL-23, anti-apoptotic protein Bcl-xL and several growth factors (Yu et al., 2009). Therefore, higher activation of NF-κB, ERK and STAT3 in Nlrp12-deficient colon tissue during the recovery stage of the disease strongly supports our phenotypic observation as well as explains the mechanism of increased tumor incidence in Nlrp12-deficient mice.

In summary, our study provides evidence for an anti-inflammatory and anti-tumorigenic role for NLRP12 in vivo by negatively regulating NF-κB and ERK signaling. Considering the importance of anti-inflammatory signals in maintaining colonic homeostasis, our study on the anti-inflammatory function of NLRP12 in colitis and colon tumorigenesis bears enormous importance. This study demonstrates a regulatory mechanism of intestinal inflammation and tumorigenesis by PRRs and paves the way to further understanding the role of NLR proteins in gastrointestinal disorders. This may help identify new therapeutic approaches to control this increasingly important health problem.

EXPERIMENTAL PROCEDURES

Mice

Nlrp12 knock-out mice were generated by homologous recombination in ES cells by replacement of exon II of the Nlrp12 gene encoding the N-terminal Pyrin domain with an IRES-β-gal-neomycin-resistance cassette via a targeting vector (Figure S1B). A positive ES clone was used to generate chimeric mice. 129/C57BL/6 chimeric mice were crossed with C57BL/6 females to generate heterozygous mice. Nlrp12^−/− mice were further backcrossed to C57BL/6 background for 10 generations. 8–10 week old male mice maintained in a pathogen-free facility were used in this study. Animal studies were conducted under protocols approved by St. Jude Children’s Research Hospital Committee on Use and Care of Animals.

Induction of DSS-induced Colitis

Acute colitis was induced with 3% (w/v) DSS (Molecular mass 36–40 kDa; MP Biologicals) dissolved in sterile, distilled water ad libitum for the experimental days 1–5 followed by normal drinking water until the end of the experiment. The DSS solutions were made fresh on day 3.

Clinical scoring of colitis

Scoring for stool consistency and occult blood was done as previously described (Zaki et al., 2010). Briefly, stool scores were determined as follows: 0 = well-formed pellets, 1 = semiformed stools that did not adhere to the anus, 2 = semiformed stools that adhered to the anus, 3 = liquid stools that adhered to the anus. Bleeding scores were determined as follows: 0 = no blood by using hemoccult (Beckman Coulter), 1 = positive hemoccult, 2 = blood traces in stool visible, 3 = gross rectal bleeding. Stool consistency scores and bleeding scores were added and presented as clinical score.

Induction of Colorectal Cancer

Mice were injected intraperitoneally with 10 mg/kg AOM (Sigma). After 5 days, 3% DSS 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.

Histopathological Analysis

Formalin-preserved colon sections 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 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). 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, neutrophil marker Gr-1, and T cell marker CD3.

In situ Intestinal Proliferation assay

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 (Zaki et al.). 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.

In vitro Signaling Assays

Bone marrow cells were cultured in L-cell-conditioned IMDM medium supplemented with 10% FBS, 1% non-essential amino acid and 1% penicillin-streptomycin for 5 days to differentiate into macrophages. Bone marrow derived macrophages were seeded in 12-well cell culture plates, cultured overnight and stimulated with ultrapure E. coli-derived LPS (Invivogen), or Pam3CSK4 (Invivogen) or Poly (I:C) (Invivogen). For analysis of caspase-1 activation, macrophages were cultured with LPS for 3h and then with 5 mM ATP (Sigma) for 30 min. or infected with Salmonella typhimurium or Listeria monocytogenes for 4 h.

Bone Marrow Chimeras

Bone marrow transfer was used to create Nlrp12^−/− chimera mice wherein the genetic deficiency of Nlrp12 was confined to either circulating cells (Nlrp12^−/− > WT chimera) or non-hematopoietic tissue (WT > Nlrp12^−/−). Briefly bone marrows were collected from femur and tibia of congenic WT (expressing CD45.1 leukocyte antigen) or Nlrp12^−/− (expressing CD45.2 leukocyte antigen) donor mice by flushing with HBSS. After several washing steps, cells were resuspended in PBS at a concentration of 1×10⁸/ml. 100 μl of this cell suspension was injected retro-orbitally in irradiated donor mice. 4 chimera groups were generated WT > WT (WT cells expressing CD45.1 into WT expressing CD45.2); WT > Nlrp12^−/− (WT cells expressing CD45.1 into Nlrp12^−/− expressing CD45.2); Nlrp12^−/− > Nlrp12^−/− (Nlrp12^−/− cells expressing CD45.2 cells into Nlrp12^−/− expressing CD45.2) and Nlrp12^−/− > WT (Nlrp12^−/− cells expressing CD45.2 into WT expressing CD45.1). The use of CD45.1-expressing congenic mice facilitated verification of proper reconstitution in the chimera mice. Bone marrow reconstitution efficiency was verified after 6 weeks by staining for CD45.1 and CD45.2 in blood cells using FITC-conjugated anti-CD45.1 and PE-conjugated anti-CD45.2. All chimera mice used in this study had around 95% reconstitution.

Isolation of lamina propria cells

Colons were dissected, washed with ice-cold PBS supplemented with antibiotics (penicillin plus streptomycin) and cut into small pieces. Colons pieces were then incubated with RPMI medium supplemented with 3% FBS, 0.5 mM DTT, 5mM EDTA and antibiotics at 37°C for 30m with gentle shaking. After removing epithelial layer, the remaining colon segments were incubated at 37°C with RPMI medium containing 0.5% CollagenaseD (Roche) and 0.05% DNAse (Roche) for 30 min with gentle shaking. The supernatant was passed through 70 μM cell strainer to isolate lamina propria cells.

NF-κB activity assay

Nuclear p65 DNA binding activity was determined by ELISA based NF-kB activity assay (Cayman, Cat # 10007889) according to manufacturer instruction. Briefly, 10 μl (5 μg) nuclear extract was incubated in 96-well ELISA plate pre-coated with a specific dsDNA sequence containing NF-κB response element. A primary antibody for NF-κB (p65) was then added which was detected by a HRP-conjugated secondary antibody. Finally P65 activity was measured as a colorimetric readout at 450 nm.

Statistical Analysis

Data are represented as mean ± SD or SEM. Statistical significance was determined by Student’s t-test. p < 0.05 was considered statistically significant.

Highlights.

• NLRP12 dampens inflammation and tumorigenesis in the colon

• NLRP12 regulates cytokine and chemokine production, and epithelial proliferation

• NLRP12 negatively regulates NF-κB and ERK activation in the macrophages

• NLRP12 activity in myeloid compartment is essential for colonic homeostasis

SIGNIFICANCE.

Colorectal cancer is the third most common form of cancer and the second leading cause of cancer-related death in developed countries. Chronic inflammation shapes the tumorigenic micro-environment in the gut by inducing cytokines, chemokines and other factors through NF-κB, ERK and STAT3 signaling. In this study, we showed that the NOD-like receptor family member NLRP12 plays a critical role in down-regulating these tumor-inducing signaling pathways. Given the importance of anti-inflammatory signals in maintaining colonic homeostasis, these results reveal a regulatory mechanism controlling inflammation and tumorigenesis in the gut, and may help identify new therapeutic approaches to control inflammatory bowel diseases.

Abbreviations

NLR

NOD-like receptor

DSS

dextran sodium sulfate

BrdU

5′-bromo-2′-deoxy-uridine

AOM

Azoxymethane

TLR

Toll-like receptor

WT

wild-type