If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The intestine-specific transcription factor Cdx2 is essential for intestinal homeostasis and has been implicated in the pathogenesis of disorders including inflammatory bowel disease. However, the mechanism by which Cdx2 influences intestinal disease is not clear. Here, we present evidence supporting a novel Cdx2–TRIM31–NLRP3 (NLR family, pyrin domain containing 3) signaling pathway, which may represent a mechanistic means by which Cdx2 impacts intestinal inflammation. We found that conditional loss of Cdx function resulted in an increase in proinflammatory cytokines, including tumor necrosis factor alpha, interleukin (IL)-1β, and IL-6, in the mouse colon. We further show that TRIM31, which encodes a suppressor of NLRP3 (a central component of the NLRP3 inflammasome complex) is a novel Cdx2 target gene and is attenuated in the colon of Cdx conditional mutants. Consistent with this, we found that attenuation of TRIM31 in Cdx mutant intestine occurs concomitant with elevated levels of NLRP3 and an increase in inflammasome products. We demonstrate that specific inhibition of NLRP3 activity significantly reduced IL-1β and IL-6 levels and extended the life span of Cdx conditional mutants, reflecting the therapeutic potential of targeting NLRP3. Tumor necrosis factor-alpha levels were also induced independent of NLRP3, potentially via elevated activity of the proinflammatory NF-κB signaling pathway in Cdx mutants. Finally, in silico analysis of ulcerative colitis patients revealed attenuation of CDX2 and TRIM31 expression coincident with enhanced expression of proinflammatory cytokines. We conclude that the novel Cdx2–TRIM31–NLRP3 signaling pathway promotes proinflammatory cytokine expression, and its inhibition may have therapeutic potential in human intestinal diseases.
). Given the long-term morbidity, rising incidence, and increased susceptibility to colorectal cancer in those with IBD, it is vital to better understand the molecular mechanisms underlying these disorders. In the absence of such advances, management of IBD is expected to remain suboptimal despite the availability of potential therapeutic options (
Although the etiology of IBD is not fully understood, it is generally believed to result from a dysbiosis, where an inappropriate immune response is evoked in reaction to environmental cues, such as commensal microbes (
), and genome-wide association studies and candidate gene analyses have provided compelling evidence that IBD represents a class of polygenic disorders, with over 240 susceptibility loci reported to date (
). Cdx members elicit their effects by direct binding to target loci at Cdx response elements (CDREs) to increase or decrease target gene transcription, at least in part, by alterations in chromatin architecture (
). These findings underscore the prospective involvement of CDX2 in the pathogenesis of IBD, although the mechanisms remain poorly understood. In this regard, although Cdx2 null mice are early lethal, Cdx2 heterozygous offspring are sensitized to dextran sodium sulfate–induced colitis (
), consistent with a protective role for Cdx in IBD. The clinical relevance of this relationship is further evidenced by the attenuation of expression of CDX2 in UC patients and its re-expression following leukocytapheresis in responsive individuals (
). Moreover, increased levels of tumor necrosis factor alpha (TNF-α), a cytokine central to the proinflammatory events in IBD, correlate with attenuated CDX2 expression in UC patients, and TNF-α can likewise suppress CDX2 expression in model systems (
). To circumvent this functional overlap and early lethality, we created a Cdx2 “floxed” conditional allele and crossed this into a Cdx1 null mutant background. We then assessed the impact of the loss of all Cdx functions in the intestine, using a tamoxifen-regulated Villin-CreERT transgene to delete Cdx2. Using this approach, we previously demonstrated that Cdx2, and to a lesser extent Cdx1, is required for intestinal homeostasis and that Cdx members function as tumor suppressors in an adenomatous polyposis coli model of colorectal cancer (
In the present study, we found that loss of Cdx function led to induction of a proinflammatory cascade in colon epithelium, including elevated NLRP3 protein and accumulation of products of the NLRP3 inflammasome. We further found that TRIM31, which encodes a negative regulator of the NLRP3 inflammasome (
), is directly regulated by Cdx2, offering a mechanistic basis for these observations. The physiological relevance of this relationship is underscored by the finding that chemical inhibition of NLRP3 activity decreased levels of IL-1β and IL-6 induced by Cdx loss and lengthened the life span of Cdx conditional mutants. We also observed increased expression of NF-κb2 and Rela, suggesting a basis for the elevated levels of TNF-α in Cdx mutants. Finally, in silico analysis revealed a strong inverse correlation between CDX2, TRIM31, and the levels of proinflammatory markers, including products of the NLRP3 inflammasome, in UC patients, underscoring the potential clinical relevance of these findings.
Cdx deletion leads to intestinal inflammation
We generated Cdx1−/−–Cdx2fl/fl-Villin-Cre ERT conditional mutant mice and treated them with tamoxifen to induce Cdx2 deletion, yielding mice lacking Cdx1 and Cdx2 throughout the intestinal epithelium (termed Cdx-DKO hereafter; DKO [double KO]) as previously described (Fig. S1A) (
). Three days post-treatment, mice developed diarrhea, lethargy, rapid weight loss, and other indices of intestinal insufficiency and necessitating endpointing at 6-day post-treatment. Histological analysis revealed pathological alterations in intestinal epithelium as previously reported (
Colon epithelial cells were isolated from WT and Cdx-DKO mice 6 days following treatment with tamoxifen and used for transcriptome profiling by RNA-Seq. This analysis revealed 6442 genes that were differentially expressed at a padj/false discovery rate of 0.05, with approximately equal numbers of genes upregulated (3231) and downregulated (3211) in the Cdx-DKO colon compared with controls (Fig. 1A). Kyoto Encyclopedia of Genes and Genomes functional enrichment analysis revealed a marked upregulation of genes involved in cytokine signaling, particularly the TNF-α pathway (Fig. 1B).
Additional bioinformatics analysis was performed for the top 500 differentially expressed genes. For each replicate, normalized counts were plotted, and data bars were added to the side to show the mean level of expression and log2FoldChange for each gene. A list of all genes downloaded from Ensembl was annotated with the Gene Ontology (GO) term “Inflammatory response” (GO: 0006954) comprising a total of 759 genes. Any gene in the heatmap that is annotated with this term was flagged with a red bar. In colon epithelium from Cdx-DKO mice, 19 proinflammatory genes were upregulated, whereas 13 genes were downregulated (Fig. 1C).
Cdx deficiency promotes the expression of proinflammatory cytokines
Genes encoding a number of cytokines and chemokines were induced in the Cdx DKO intestine, including TNF-α, IL-1β, CXCL1, and CCL2, as determined by RNA-Seq analyses and confirmed by RT–quantitative PCR (qPCR) (Figs. 2, A–E and S2, A–D). We were also able to detect upregulation of IL-6, which is a downstream target of IL-1β (
IL-1β is activated by the NLRP3 inflammasome, prompting us to examine the consequence of Cdx deletion on the expression of the components of this pathway. We found that the protein levels of NLRP3, pro-caspase-1, caspase-1, IL-1β, and IL-18 were all significantly elevated in Cdx mutants compared with WT (Figs. 3, A–G and S6).
Inflammatory cytokines secreted by the intestinal epithelium have a functional role in recruiting cells of the innate immune system. We therefore used flow cytometry to compare the levels of infiltrating monocytes (Ly6C-hi:Ly6G-low) and neutrophils (Ly6C-hi:Ly6G-hi) in the colons of Cdx-DKO and WT mice and found a significant increase in both types of immune cells in the mutant samples (Fig. 3, H–J).
Identification of Cdx target genes
To better understand the means by which Cdx function impacts intestinal inflammation, we conducted chromatin immunoprecipitation followed by deep sequencing (ChIP-Seq) using chromatin prepared from colon epithelial cells and a Cdx2 antibody as described previously (
). Approximately 5100 sites were identified using MACS2 peak calling annotated to genes that have Cdx2 bound in the region from −20 kb to +5kb relative to the transcriptional start site (TSS). Genomic distribution analysis revealed that these peaks were primarily localized in intergenic (35.1%) and intronic (37.7%) regions, followed by proximal promoter (≤1 kb from the TSS; 13%) and upstream promoter (1–3 kb; 8.8%) regions (Fig. S3A). Analysis of sequences under the ChIP-Seq peaks revealed significant enrichment for the motif TTTATGGCT, consistent with previous observations for the consensus CDRE (Fig. S3B) (
), ATAC-Seq was also conducted to examine the impact of Cdx status on chromatin accessibility. ATAC-Seq and ChIP-Seq profiles for all protein-coding genes from Ensembl were created using the mouse reference genome UCSC mm10. The signal was normalized to counts per million, and the Cdx2 binding profile shown on a distinct color scale (range, 0–0.2 for Cdx2, 0–0.5 for ATAC-Seq and H3K4me3). Genes were examined for significant differential expression in Cdx mutants (Up_in_KO, Down_in_KO) or not differentially expressed (NC_in_KO). Heatmap analyses showed chromatin accessibility (ATAC-Seq) in control cells aligning with Cdx2 peaks and H3K4me3. Markedly lower signals were observed for ATAC-Seq in a subset of genes that were downregulated in DKO mutants as determined by RNA-Seq, suggesting Cdx impacts the expression of these genes via altered chromatin accessibility (Fig. 4A).
TRIM31 is a direct Cdx target gene
To further identify potential direct Cdx target genes implicated in intestinal inflammation, we intersected Cdx2 occupancy as well as altered genome accessibility and differential expression (>1.4-fold altered expression). This yielded a list of 227 genes that were bound by Cdx2, exhibited altered chromatin accessibility, and were differentially expressed in Cdx mutant colon (Fig. 4B). TRIM31 was recovered from among these genes.
Examination of the TRIM31 locus revealed a major Cdx2 binding peak within the promoter region (Fig. 4C) with occupancy confirmed by ChIP–PCR (Fig. S5A). Alignment of the ATAC-Seq tracks between WT and Cdx mutant samples revealed Cdx-dependent loss of accessibility in mutant cells (Fig. 4C), which correlated with attenuation of TRIM31 expression (Fig. 4D). Cdx-dependent regulation of TRIM31 was also validated by RT–qPCR and immunoblot analysis from both mouse colon epithelium (Fig. 4, E–G) and Cdx conditional mutant colon organoid cultures (Fig. S4). In addition, Cdx2 occupancy, chromatin accessibility, and expression correlated with loss of the active chromatin mark H3K4me3 at the promoter region of TRIM31 following Cdx deletion (Fig. S5B). Finally, we identified a putative CDRE under the ChIP-Seq peak, comprised of the sequence TTTATG (Fig. 4H).
To further assess Cdx-dependent regulation of TRIM31, we employed heterologous reporter assays using TRIM31 promoter sequences containing either a WT or a mutated CDRE driving expression of a luciferase reporter gene. This analysis revealed that expression of the WT reporter was Cdx dependent in C2bbe1 intestinal cells, and that this response required the CDRE (Fig. 4I). Taken together, these findings strongly suggest that TRIM31 is a direct Cdx target gene in the intestinal epithelium.
NLRP3-dependent intestinal inflammation in Cdx mutant mice
TRIM31 encodes a ubiquitin E3 ligase and has previously been identified as a suppressor of the NLRP3 inflammasome (
). This raises the possibility that Cdx-dependent intestinal inflammation could be due, at least in part, to attenuated expression of TRIM31 and subsequent induction of the NLRP3 inflammasome, as observed in Cdx-DKO mice (Fig. 3).
To further examine this relationship, we assessed the effect of inhibiting the NLRP3 inflammasome with a specific antagonist, CY-09 (
). To this end, WT or Cdx conditional mutant mice were treated with CY-09 or saline for 2 days, followed by tamoxifen treatment and continuation of CY-09 (or saline) treatment (Fig. 5A). Mice were euthanized on day 5 post-tamoxifen or when moribund, and the expression of the proinflammatory cytokines TNF-α, IL-1β, and IL-6 measured by ELISA from colon extracts. In addition, NLRP3 was examined by immunofluorescence (IF), and IL-18 expression and processing assessed by immunoblot from colon epithelial cells after depletion of CD-45-positive immune cells.
When compared with saline-treated conditional mutant mice, CY-09-treated Cdx-DKO mice survived longer (Fig. 5B), suggesting that the NLRP3-mediated inflammatory response contributes to their demise. Notably, CY-09 had no inhibitory effect on the baseline levels of inflammatory markers in WT mice, whereas it reduced IL-1β and IL-6 in Cdx mutants, with TNF-α remaining unchanged (Fig. 5, C–E). CY-09 did not affect the levels of IL-1β or IL-6 transcripts (Fig. S7, A and B), consistent with the inhibitor acting post-transcriptionally. The epithelial nature of this effect is consistent with IF analysis, which revealed that an induction of NLRP3 was observed in the epithelial cells of the mutant colon, and that this expression was reduced by CY-09 treatment (Fig. 5F). In addition, IL-18 protein expression and activation was observed in immune-purified mutant colon epithelial cells (Fig. S6).
NF-κB expression in Cdx mutants
The activation and nuclear translocation of NF-κB transcription factors is a crucial step for the transcription of a number of proinflammatory genes, including TNF-α. TNF-α receptor activation can further enhance the cytokine signaling cascade by activating both mitogen-activated protein kinase and NF-κB pathways (
). Aberrant intestinal differentiation may contribute to the initiation of these cascades in Cdx mutants, with inflammation further amplified by upregulation of NLRP3 inflammasome activity. In this regard, we noted that stress-activated mitogen-activated protein kinase pathways were upregulated in Cdx mutant intestine (data not shown). Moreover, we found a significant increase in expression of Nf-κb2, Rela, and the activated form Pser536-NF-κB (Fig. 6, A–C), whereas Nf-κb and Relb remained unchanged (Fig. S8, A and B). Although the basis for these observations is presently unclear, it is tempting to speculate that loss of Cdx function contributes to induction of NF-κB signaling, thereby further amplifying TNF-α-mediated inflammatory cascades. Related to this, we also observed downregulation of MEP1A in Cdx mutant mice (Fig. 6D) consistent with previous studies suggesting that MEP1A is a UC susceptibility gene and is suppressed by TNF-α (
To determine if CDX status may impact IBD in human populations, we examined publicly accessible RNA-Seq data from UC patients comprised of 24 samples from noninflamed and patient-matched inflamed mucosal tissue. This analysis revealed that CDX2 was one of the top differentially expressed genes, being reduced in almost all inflamed UC patient samples (Fig. 7A). TRIM31 levels were also generally reduced in the same samples concomitant with elevation of inflammatory markers, including the NLRP3 product IL-1β (Fig. 7, B–G). While these findings are consistent with a conserved CDX–TRIM31–NLRP3 pathway that contributes to IBD pathology, there also appears to be a relative reduction in epithelial cells in samples from inflamed UC patients, as assessed by EPCAM expression (Fig. S9), and further analysis will be needed to better establish this relationship.
This study provides evidence for a novel role for Cdx in intestinal inflammation, which has implications for the etiology and therapeutic management of IBD. We demonstrated that TRIM31 is a novel direct Cdx target gene. TRIM31 expression was attenuated in Cdx mutant mice concomitant with an increase in the levels of its target, NRLP3, and of proinflammatory cytokines, including IL-1β, IL-6, IL-18, and TNF-α. Extending these findings, we showed that CDX2 and TRIM31 levels were also suppressed in UC patients, concomitant with an increase in IL-1β and TNF-α, suggesting that this pathway is conserved. We also found that the expression of the NF-κb members Nf-kb2 and Rela were upregulated in Cdx mutant epithelium, suggesting that Cdx loss impacts the expression of these proinflammatory transcription factors through unknown basis.
A number of IBD susceptibility loci have been identified, indicative of complex genetics contributing to the disease (
). Cdx2 has been shown to impact the expression of certain of these genes, prompting us to identify additional Cdx target genes implicated in inflammation. Analysis of genomic distribution of Cdx2 occupancy through ChIP-Seq revealed binding sites that were both gene proximal and intergenic, consistent with Cdx functioning through both distal enhancer motifs as well as proximal regulatory elements, as previously described (
). We also observed that loss of Cdx function impacted chromatin accessibility at a number of putative target genes, including TRIM31. This is in agreement with previous findings in both embryonic tissue and the small intestine in which Cdx2 deletion resulted in a loss of chromatin accessibility for a subset of its targets (
). Consistent with a role for Cdx in this process, we found that loss of Cdx function was associated with a rapid and pronounced inflammatory response, including induction of IL-1β and TNF-α, among other cytokines and chemokines. We further found that Cdx is a direct regulator of TRIM31, which encodes a negative regulator of the NLRP3 inflammasome (
). Consistent with this, Cdx mutants exhibited decreased levels of TRIM31 concomitant with elevated NLPP3 protein levels and increased inflammasome activity as evidenced by activation of caspase 1 and maturation of its substrate IL-1β. In addition, increased levels of IL-6 were observed in Cdx mutant intestine, in agreement with the known role for IL-1β upstream of the production of this proinflammatory cytokine (
Certain other genes encoding proinflammatory effectors, such as Nfkb2 and Rela, also exhibited a gain in chromatin accessibility and concomitant increase in expression in Cdx mutants. While the basis for this is presently unknown, it is possible that the increased TNF-α levels observed in the mutant mice promotes activation of the NF-κB pathway (
). Such activation would provide a mechanistic basis for the induction of expression of NF-κB target genes including those encoding proinflammatory players such as IL-1β, IL-6, IL-17, Tnfsf13b as well as chemokines such as Ccl2 and Cxcl1, as also observed in the mutant intestine. In this regard, increased IL-6 protein levels in Cdx mutants could be attributed to both NLRP3-dependent IL-1β production, or to induction of NF-κB signaling, another known regulator of IL-6 expression (
). In the case of Cdx-dependent intestinal inflammation, a significant amount of IL-6 production would appear to arise from the former, as inhibition of the NLRP3 inflammasome by CY-09 attenuated generation of both IL-1β and IL-6. Irrespective of the source, elevated levels of IL-6 could also contribute to the recruitment of Ly6c-positive monocytes as observed in the Cdx mutants, as IL-6 is required for neutrophil trafficking during the inflammatory response (
). Other chemokines and proinflammatory cytokines, such as Ccl2 and Cxcl1, were also increased in Cdx mutants and may also contribute to this chemotaxis.
Treatment with the NLRP3 inhibitor CY-09 reduced the levels of IL-1β and IL-6 and increased the average survival time of Cdx mutants. This finding underscores the pathophysiological consequences of dysregulation of Cdx function specifically upstream of NLRP3 and suggests a potential therapeutic benefit of pharmacological inhibition of NLRP3 in intestinal inflammation. Notably, there was no difference in TNF-α levels between saline-treated and CY-09-treated Cdx mutant mice, consistent with TNF-α production independent of NLRP3, the nature of which is presently unknown. TRIM31 is also known to regulate NLRP3 activity via ubiquitin-mediated proteasomal degradation, thereby impacting inflammasome assembly (
). In this regard, it is interesting to note that GO analysis identified “proteasome protein catabolic process” as one of the pathways altered in Cdx-DKO mice, suggesting that Cdx may have additional functions related to protein degradation.
A relationship between TNF-α, CDX2, and IBD has been previously suggested. For example, TNFα has been shown to suppress CDX2 expression in model systems, whereas, in vivo, anti-TNF-α therapy can restore CDX2 expression levels. In addition, TNF-α suppresses the expression of the CDX2 target gene MEP1A, which has also been identified as an IBD susceptibility locus. Taken together, these findings suggest that TNF-α elicits its inflammatory effects, in part, through attenuation of CDX2 and subsequent impact on downstream targets, at least some of which are implicated in IBD (
). Our present results evoke an additional, novel, mechanism that involves the TRIM–NLRP3–inflammatory pathway wherein loss of Cdx function leads to attenuation of the target gene TRIM31, resulting in increased NLRP3 levels and subsequent generation of proinflammatory cytokines. In addition, in silico analysis suggests a similar relationship between CDX2 and TRIM31, and inflammation is conserved in UC patients. Taken together, these findings lend support to an important pathogenic role for the CDX2–TRIM31–NLRP3 pathway in IBD and suggest new therapeutic options for the management of chronic intestinal inflammation, which may be particularly germane for patients exhibiting reduction of CDX2 expression.
Mouse models and treatment
Cdx1−/−-Cdx2fl/fl-Villin-Cre ERT mice have been previously described (
). Mice were treated at 12 to 14 weeks of age with a single 4.0 mg dose of tamoxifen to delete Cdx2, generating DKO mice. Animals were maintained according to the guidelines of the Canadian Council on Animal Care as approved by the Animal Care and Veterinary Services of the University of Ottawa.
Cell isolation, organoid culture, and treatment
Colon epithelial cells were enriched by incubating fresh colon tissue in 5 mM EDTA in PBS (pH 8) for 10 min at 4 °C with three to four repeated washes. Each aliquot was collected, and cells run through a 70-μm strainer and either fixed for ChIP-Seq or used to prepare samples for Western blot, RNA-Seq, or ATAC-Seq. In some experiments, filtered cells were incubated with Magnisort beads (Invitrogen; catalog no.: MSNB-6002-74) conjugated to an anti-CD45 antibody (Invitrogen; catalog no.: 13-0451-82) as per the manufacturer’s directions. Unbound epithelial cells were recovered and processed for Western blot. Organoid culture from mouse colon was performed as described (
). Briefly, colon epithelial cells were isolated as aforementioned, crosslinked with 1% formaldehyde (Sigma; catalog no.: F8775), resuspended in radioimmunoprecipitation buffer, and sonicated to obtain 200- to 500-bp chromatin fragments using a Branson sonicator. Lysates were incubated overnight at 4 °C with primary antibodies or with isotype-specific immunoglobulin G control and DNA recovered and purified. RNA-Seq library construction and sequencing was conducted by Genome Quebec using an Illumina HiSeq 4000 system. Approximately 25 to 34 M reads were generated for the ChIP-Seq analyses.
RNA extraction, RNA-Seq, and RT–qPCR
Total RNA was isolated using Trizol or a Qiagen RNeasy mini kit according to the manufacturer's protocol, and RNA quality was assessed by Bioanalyzer. Library construction for and sequencing was performed by Genome Quebec with 19 to 30 M reads per sample. Validation of selected genes from RNA-Seq data was conducted using RT–qPCR on a CFX96 Bio-Rad thermocycler with Go-Taq SYBR Green master mix (Promega). Analyses were performed based on the Ct values from the target gene and internal control gene (18s rRNA). All primers for PCR analyses are noted in Table S1.
Colon cells were collected as aforementioned and treated for 30 to 40 min at 37 °C with trypsin to obtain single-cell suspensions. Approximately 100,000 cells were used for the transposase reaction as previously described (
). Briefly, cells were resuspended in cold lysis buffer (10 mM Tris–Cl at pH 7.4, 10 mM NaCl, 3 mM MgCl2, and 0.1% Igepal CA-360), and crude nuclear pellets were incubated in transposition mix for 20 to 30 min at 37 °C. DNA was then purified using the MinElute PCR purification kit (Qiagen), PCR amplified using NEBNext High-Fidelity master mix, a common forward primer, and uniquely barcoded reverse primers for each sample. The reaction was purified using MinElute PCR purification kits (Qiagen), and primer dimers were removed using Ampure XP beads (Beckman Coulter). Library construction and paired-end sequencing were performed using an Illumina Hi-Seq 4000 system by Genome Quebec and sequenced to a depth of 55 to 72 M reads.
Western blot and ELISA
Whole-cell lysates were prepared from colon epithelial cells, proteins resolved by SDS-polyacrylamide gel electrophoresis, and then transferred onto nitrocellulose membranes for immunoblot analysis. Primary antibodies used were anti-Cdx2 at 1:1000 dilution (
), anti–caspase-1 (Thermo Fisher Scientific), anti-NLRP3 (Thermo Fisher Scientific), anti–β-tubulin (Santa Cruz), anti–phospho-ser536NF-κB (Cell Signalling), anti–Il-1β, and anti–β-actin (Santa Cruz). ELISA was conducted on colon epithelial lysates using kits from BD Bioscience as recommended by the supplier.
Colon tissue was dissected, fixed in 4% paraformaldehyde at 4 °C overnight, and then transferred to 70% ethanol. Tissue was embedded in paraffin, and 4 μm sections were mounted on slides. Paraffin-embedded tissue sections were deparaffinized in xylene and rehydrated in an ethanol series. Antigen retrieval was carried out using a Decloaking chamber with immersion in citrate buffer (10 mM citric acid, 0.05% Tween-20, pH 6.0). Slides were allowed to cool down while in citrate buffer for 30 min at room temperature, then washed in water for 5 min. Sections were incubated in a blocking solution (5% goat serum, 4% bovine serum albumin, 10% sucrose in 1× Tris-buffered saline [TBS]) for 1 h at room temperature in a humidified chamber and then incubated with anti-NLRP3 primary antibody (Invitrogen; catalog no.: MA5-32255) at a 1:500 dilution in antibody diluent (1% bovine serum albumin, 10% sucrose in 1× TBS) at 4 °C overnight in a humidified chamber followed by three 5-min washes in 1× TBS. Negative control sections were incubated in antibody diluent instead of primary antibody. Sections were then incubated with goat anti-rabbit Alexa-Fluor 488 secondary antibody (Invitrogen; catalog no.: A-11008) for 1 h at room temperature in a dark humidified chamber, washed three times for 5 min each in 1× TBS, and mounted with Fluorescence Mounting Medium (Dako Omnis; code S3023). Images were captured using a Zeiss AxioImager M2.
Cells were washed twice in 2 ml of cold PBS for 5 min, fixed in 4% paraformaldehyde for 20 min at room temperature, pretreated with anti-Fc antibody for 1 h and subsequently for 2 h with fluorophore-conjugated antibodies (EPCAM BioLegend, catalog no.: 118215; CD45 BioLegend, catalog no.: 147711; Ly6C BioLegend, catalog no.: 128015, or Ly6G BioLegend, catalog no.: 127625). Cells were then washed to remove excess antibody, and dead/live staining was performed using zombie yellow (BioLegend) as per the manufacturer’s directions and assessed using a BD LSFRFortessa.
In vivo NLRP3 inhibition
Mice (12–14 weeks old) were treated with either the NLRP3 inhibitor CY-09 (2.5 mg/kg; MilliporeSigma) or saline, following which animals were treated with tamoxifen or vehicle on day 2 and treatment with CY-09 or saline continued to day 6 post-tamoxifen or until the animals became moribund.
Heterologous promoter analysis
The TRIM31 promoter region (1176 bp upstream of the start codon) containing either a WT (TTTATG) or mutated (CCATGG) CDRE was cloned into the PxP2 luciferase reporter vector. C2BBE1 cells, cultured under standard conditions, were transfected in triplicate using Lipofectamine and 1 μg of the appropriate luciferase reporter construct, 1 μg of Cdx2 expression vector (or empty vector control), 0.2 μg of an xgal expression vector, and 100 ng of GFP expression vector to a total of 2.3 μg of DNA per transfection. Cells were harvested 48 h post-transfection, and lysates were analyzed using the Promega Luciferase Assay System, using β-galactosidase activity to correct for transfection efficiency as described previously (
). Multiple testing correction was performed using the Benjamini–Hochberg method, and lists of significantly differentially expressed genes were identified using a q value (i.e., a corrected p value) cutoff of 0.05.
) was used to perform functional enrichment analysis to find classes that are enriched in genes that were differentially expressed in Cdx DKO mutants relative to controls. Using ClusterProfiler, we performed hypergeometric tests to identify GO Biological Processes, Molecular Function terms, and Kyoto Encyclopedia of Genes and Genomes pathways that were overrepresented in these differentially expressed genes, analyzing each annotation set and each direction of differential expression independently.
Genome enrichment patterns
ChIP-Seq and ATAC-Seq reads were aligned to the mm10 genome sequence using bwa, version 0.7.10 (
) bamCoverage, scaling the signal to counts per million. Protein coding gene TSSs were extracted from Encode, v101 database using BioMart. DeepTools computeMatrix was used to calculate the coverage of Cdx2, H3K4me3 ChIP-Seq, and ATAC-Seq reads in the region around each TSS. plotHeatmap was then used to plot the coverage patterns for all factors, separating TSSs into genes significantly up, down, or with no significant fold change in mutants relative to controls.
Statistical analyses were performed using Microsoft Excel and GraphPad Prism 8.0 (GraphPad Software, Inc). Statistical significance was calculated using t test and one-way ANOVA; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, or ∗∗∗∗p < 0.0001.
RNA-Seq data presented in this article have been deposited at Gene Expression Omnibus (NCBI) with the study accession codes; GSE and GSE. ChIP-Seq and ATAC-Seq data were deposited with the study accession codes (to be provided). IBD patient data were obtained from Gene Expression Omnibus accession number GSE107597.
The authors declare that they have no conflicts of interest with the contents of this article.
We extend our thanks to Dr Simon Chewchuk for his assistance in organoid culture and Dr Vera Tang of the Flow Cytometry Core for her support with flow analysis. The graphical abstract was created using BioRender.com.
D. L. conceptualization; S. J., N. A., B. H., and S. S. methodology; B. H. formal analysis; S. J. and S. H. investigation; S. J. writing–original draft; S. S. and D. L. writing–review & editing; D. L. supervision.
Funding and additional information
This work was supported by a grant from the Canadian Institutes of Health Research (grant no.: 146014) to D. L. and a Canadian Institutes of Health Research postdoctoral fellowship to S. J. (reference number: 154639).
Expression of Cdx2 and TRIM31 in mouse organoid culture. (a) Representative organoid derived from colon epithelium. (b, c) Organoid cultures were treated with vehicle or 4-hydroxytamoxifen and harvested after 72 h for RT-qPCR quantification of relative expression of (b) Cdx2 and; (c) TRIM31. Error bars represent the standard deviation from the mean of three independent biological experiments. Statistical significance was calculated using one-way ANOVA∗P<0.05, ∗∗P<0.01
(a) Flow cytometric dot plots of freshly isolated intestinal cells before and after cell separation by anti-CD45 conjugated magnetic beads. Cytometry was based on EpCAM or CD45 in double-stained samples. (b) IL-18 is upregulated in purified DKO colon epithelial cells. Immunoblot analyses for pre IL-18, mature IL-18 and beta actin (left) and relative quantification of compared to control (middle and right hand plots). Error bars represent the standard deviation from the mean of three independent biological experiments. Statistical significance was calculated using two-tailed t test, ∗P<0.05, ∗∗P<0.01
Cytokine gene expression. Mice were treated with either CY-09 or vehicle (saline), tamoxifen-treated on day 2 and gene expression assessed by RT-qPCR analysis for (a) IL-1b; (b) TNF-α; (c) Cxcl1 and; (d) Ccl2. Error bars represent the standard deviation from the mean from three independent biological replicates. Statistical significance was calculated using one-way ANOVA. ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001
(a, b) RT-qPCR quantification of relative expression for; (a) NF-kB1 and; (b) Relb. Error bars represent the standard deviation from the mean from three independent biological samples. Statistical significance was calculated using one-way ANOVA. ns; not significant