DOI: https://doi.org/10.21203/rs.3.rs-2314559/v1
Normal epithelial cells exert their competitive advantage over RasV12-transformed cells and eliminate them into the apical lumen via cell competition. However, the internal or external factors that compromise cell competition and provoke carcinogenesis remains unknown. In this study, we examined the effect of sequential accumulation of gene mutations, mimicking multi-sequential carcinogenesis on RasV12-induced cell competition in intestinal epithelial tissues. Consequently, we found that directionality of RasV12-cell extrusion in Wnt-activated epithelia is reversed, and transformed cells are delaminated into the basal lamina via non-cell autonomous MMP21 upregulation. Subsequently, diffusively infiltrating, transformed cells develop into highly invasive carcinomas. Elevated production of MMP21 is elicited partly through NF-κB signaling, blockage of which restores apical elimination of RasV12 cells. We further found that the NF-κB-MMP21 axis is significantly bolstered in early colorectal carcinoma in humans. Collectively, this study shows that cells with high mutational burdens exploit cell competition for their benefit by behaving as unfit cells, endowing them with an invasion advantage.
Recent advances in cancer genomics have revealed that transformed cells carrying oncogenic insults are frequently produced in all tissues1, 2. Nonetheless, living organisms implement self-defense machinery to suppress oncogenesis, with cell competition being one of the tumor-surveillance systems to remove newly emerging transformed cells3–7. In addition to an anti-tumorigenic roles, it has become increasingly apparent that divergent physiological processes such as tissue repair, stem cell maintenance, and even aging, are governed, at least partially, by cell competition across a range of epithelial tissues8–12. In 1975, Morata et al. discovered that ribosomal-deficient mutants underwent apoptosis when they coexisted with normal epithelial cells in imaginal wing disc of Drosophila13. Following this pioneering cell competition study, it was also discovered in Drosophila, that suboptimal cells with aberrant cell polarity, metabolism, membrane trafficking, and so on, are eliminated by competitive interactions11, 14–16. In addition, recent studies have uncovered that cell competition is highly conserved in mammals, as exemplified by a series of studies using mammalian experimental models17–20. We previously established a cell competition mouse model (LSL-RasV12-IRES-eGFP mice mated with Cre-ERT2 transgenic lines) in which RasV12-transformed cells are produced in a mosaic in epithelial layers by administration of low-dose tamoxifen. Using that model, we demonstrated that RasV12-transformed cells residing between normal cells are apically extruded by cell competition and are excreted along with other body waste21. This model is a valuable platform to study apical elimination of transformed cells in vivo, and is well suited to functional analysis of cell competition in a physiological setting22–25. To date, it has been reported that RasV12 cells are apically expelled in several mouse organs including small intestine21, stomach25, pancreas22, 23, lung22. These findings indicate that removal of transformed cells via cell competition is an innate defense system orchestrated among epithelial cells, which suppresses accumulation of mutated lineages and reduces the risk of oncogenesis.
In general, progressive accumulation of genetic aberrations is associated with onset of carcinogenesis26, 27. For instance, it has been well documented that transition of intestinal normal epithelial cells into malignant cells in human colorectal cancer follows a stepwise series of genetic mutations in order of APC, Ras and p5328, 29. Yet, these studies have not identified the mechanism that determines the sequence of these genetic alterations. In addition, it is still unclear how accumulated oncogenic insults influence cell competition. This caused us to wonder whether genetic mutations might affect the behavior of transformed cells in a competitive environment. We reasoned that it might be possible to exploit human familial adenomatous polyposis (FAP) as a model to evaluate functional perturbation of cell competition during multi-sequential carcinogenesis. Given that mutations in the APC tumor suppressor gene are most prevalent initiating insults in colorectal cancer30, 31, we engineered the cell competition mouse model to sustain infrequent somatic activation of Ras in APC-mutated epithelia in a manner reminiscent of FAP and investigated the fate of transformed cells. In this study, we demonstrated that APC-ablation causes cell competition to malfunction, converting this normal homeostatic mechanism into a potential driver of tumor progression.
Aberrant Wnt activation disturbs apical extrusion and potentiates non-cell autonomous diffuse invasion of RasV12-transformed cells into the basal lamina
To examine the effect of APC-deficiency on apical extrusion of RasV12-transformed cells, we exploited APCMin mice, which harbor a heterozygous loss-of-functional mutation in the APC gene32. Quantitative real-time PCR (q-PCR) analysis using intestinal epithelial cells of wild-type- or APCMin-derived crypt organoid cultures confirmed that expression of Wnt-targeted genes was markedly elevated in APCMin mice compared to wild-type mice (Extended Data Fig. 1a). In order to somatically activate H-Ras protein in a small proportion of intestinal cells in the context of APC ablation, APCMin strains were crossed with Villin-CreERT2 mice and DNMT1-CAG-loxP-STOP-loxP-HRasV12-IRES-eGFP mice (referred to herein as APCMin-Villin-RasV12 mice) or CAG-loxP-STOP-loxP-eGFP mice (APCMin-Villin-GFP mice). Although Villin-CreERT2 mice recombine Cre-responsive elements in both differentiated cells and Lgr5 + stem cells33, low-dose, tamoxifen-dependent, mosaic expression of transgenes was primarily restricted to terminally differentiated enterocytes, but not occur in Olfm4-positive stem cells, Wheat Germ Agglutinin (WGA)-positive paneth and goblet cells or DCLK1-positive tuft cells (Extended Data Fig. 1b). This is presumably due to higher tamoxifen sensitivity of differentiated cells than stem cells.
We next investigated the fate of newly emerging RasV12-transformed cells in intestinal tracts 3 days after tamoxifen administration using 6-8-weeks-old mice to avoid aberrant activation of Wnt signaling caused by loss of heterozygosity (LOH) of the APC locus in old mice34. Consequently, substantial number of RasV12-expressing cells surrounded by normal epithelial cells were apically eliminated in Villin-RasV12 mice. Furthermore, we noticed that a tiny fraction of transformed cells (1.87 ± 0.23%) appeared basally delaminated (Fig. 1a, b). Accordingly, we defined those cells as basally extruded cells, with nuclei oriented toward the basal lamina. Remarkably, the percentage of cells for which the above criteria were met was significantly higher in APCMin-Villin-RasV12 mice (7.67 ± 1.03%) compared with Villin-RasV12 mice, while GFP-expressing cells of APCMin-Villin-GFP mice were either apically or basally extruded no more often than their neighbors (Fig. 1a, b). Furthermore, basally extruded APCMin/RasV12-transformed cells penetrated beyond the basement membrane, and invaded stromal tissues in which they expanded (Extended Data Fig. 2a). Notably, APCMin/RasV12 cells were not positive for Ki-67 staining, a surrogate marker for cell proliferation, excluding the possibility of growth-stimulated overcrowding delamination of APCMin/RasV12 cells (Extended Data Fig. 2b). Collectively, these results suggest that Wnt activation causes deregulation of cell competition, resulting in promotion of basal extrusion of RasV12-transformed cells in vivo. In addition, basal extrusion of APCMin/RasV12 cells was observed at a similar frequency throughout the small intestine, from duodenum to ileum (Extended Data Fig. 3a, b). As with in vivo analysis, experiments using intestinal organoids also revealed basal extrusion of APCMin/RasV12 cells after RasV12 induction by low-dose tamoxifen treatment, while that of APCMin/GFP or RasV12-single mutated cells was rarely observed (Fig. 1c, d). 3D modeling of APCMin/RasV12-transformed organoids highlighted elongating, protruding, transformed cells which were basally exterminated (Fig. 1e). Importantly, when RasV12 mutants were predominantly induced in APCMin organoids by high-dose tamoxifen, the frequency of both apical and basal extrusion was drastically reduced in various cluster-forming transformed cells, in a density-dependent fashion (Fig. 1f, g). These results underscore that surrounding APC-ablated cells are required for APCMin/RasV12 cells to be expelled from epithelia, indicating that transformed cells are extruded non-cell autonomously.
APC Min -Villin-RasV12 mice develop invasive carcinomas
We then tracked the fate of APCMin/RasV12-transformed cells invading the lamina propria long after RasV12 expression. At 21 days after tamoxifen administration, both Villin-RasV12 mice and APCMin-Villin-GFP mice displayed no overt phenotypes, with GFP-expressing cells having disappeared due to rapid turnover of intestinal epithelial cells (Fig. 2a). In marked contrast, APCMin-Villin-RasV12 mice succumbed to intestinal transformation as GFP-positive APCMin/RasV12 cells were confined in the stroma of upper villi (Fig. 2a). Yet, microscopically detectable deformities of tissue architecture in the form of benign adenomatous tumors (adenoma) were not observed in proximity to tumors, implying that cancer cells were directly generated from normal mucous membrane through diffuse infiltration of transformants. Furthermore, stromal cell nests were histopathologically undifferentiated without forming glandular ducts, being different from signet-ring cell carcinomas or neuroendocrine neoplasms, and most nearly resembling the diffuse-type carcinomas in human35, 36. By 36 days, APCMin/RasV12 cells invaded further into submucosa or even into the muscle layer (Fig. 2b). Moreover, those cells invaded cohesively, as evidenced by expression of E-cadherin sustained in the invasive cells (Fig. 2b).
These findings prompted us to characterize the properties of APCMin/RasV12-malignant cells thoroughly by evaluating expression of several markers (Extended Data Fig. 4a). APCMin/RasV12 carcinoma cells were not positive for AB-PAS staining. Moreover, they expressed Sox9 (an intestinal progenitor marker) at the lowest level and maintained expression of CDX2 (an intestinal lineage marker). In contrast, typical adenoma cells produced in elderly APCMin-Villin-GFP mice exhibited the opposite expression pattern, being positive for AB-PAS staining, expressing Sox9 at a high level, as well as diminished CDX2 expression. Immunofluorescent staining of Ki-67 demonstrated that carcinoma cells were moderately growth-stimulated in a sharp contrast to highly proliferative status of adenoma cells, which would reflect a stress-resistant status of APCMin/RasV12 cells to survive in a harsh stromal environment. In addition, APCMin/RasV12 cells did not express synaptophysin or chromogranin, markers for enteroendocrine cells, confirming that they were not derived from enteroendocrine cells (Extended Data Fig. 4b). These results suggest that Wnt-induced perturbation of cell competition generates highly invasive, unique carcinoma cells derived from terminally differentiated cells, and does not follow the histopathogenesis of the well-known adenoma-carcinoma sequence. Importantly, APCMin/RasV12-positive adenomatous structures, in which GFP-positive transformants occupied across from crypts to differentiated tips of villi were also observed, albeit it is rare, plausibly representing stem cell-hit lesions (data not shown). Of particular interest, APCMin/RasV12-transformed cells actively penetrate stromal vessels (Fig. 2c), and 3D imaging of the small intestine by whole mount staining revealed that cancer cells preferentially invade lymphatic vessels (Fig. 2d). Furthermore, APCMin-Villin-RasV12 mice bearing lymphatic invasions invariably exhibited metastasis of tumor cells to mesenteric lymph nodes (Fig. 2e). In summary, cell competition dysregulated by Wnt activation can be exploited by RasV12-transformed differentiated cells to manifest diffuse invasion, resulting in production of highly invasive carcinoma cells, a fate distinct from the adenoma-carcinoma sequence.
With the aim of understanding the molecular mechanism underlying non-cell autonomous basal delamination of transformed cells, we established Madin-Darby canine kidney (MDCK) or MDCK-pTR GFP-RasV12 cells (tetracycline-inducible RasV12-expressing cells) stably expressing a mCherry-conjugated β-catenin mutant (β-cat ΔN) to mimic Wnt activation in vitro (Fig. 3a). This mutant lacks the N-terminal 131 amino acids and fails to undergo GSK3β-mediated degradation, thereby resulting in constitutive activation of Wnt signaling37. Using a TOP FLASH reporter assay, we confirmed that MDCK cells expressing a β-cat ΔN mutant (β-cat ΔN cells) and MDCK-pTR GFP-RasV12 cells expressing the same mutant (β-cat ΔN/RasV12 cells) exhibit comparable Wnt activation (Fig. 3b). When β-cat ΔN/RasV12-transformed cells were co-cultured with β-cat ΔN cells at a ratio of 1:50, a sizeable fraction of β-cat ΔN/RasV12 cells were basally extruded into the collagen matrix over time, while the number of apically extruded cells or dead cells was slightly increased compared to β-cat ΔN/RasV12 cells cultured alone (Fig. 3c, d). The observation that β-cat ΔN/RasV12-transformed cells surrounded by themselves remained within epithelia demonstrates that the presence of surrounding β-cat ΔN cells markedly accelerates basal extrusion of β-cat ΔN/RasV12 cells. Thus, we successfully phenocopied the salient feature of non-cell autonomous basal extrusion of APCMin/RasV12 cells in vitro.
We wondered whether Wnt-activated cells are eliminated by cell competition upon emergence of Wnt-transformed cells within normal epithelia. To address this issue, we established MDCK cells stably expressing a β-cat ΔN mutant in a doxycycline-inducible manner (Extended Data Fig. 5a). At 16 h after doxycycline addition, Wnt signaling was profoundly activated (Extended Data Fig. 5b). No obvious loser phenotypes such as cell extrusion or cell death were observed when β-cat ΔN cells were surrounded by normal MDCK cells, indicating that Wnt activation per se does not confer loser status, at least in this experimental setting (Extended Data Fig. 5c, d). However, it has been reported that cells with activated Wnt signaling are expelled from tissues in zebrafish or hair follicles of mice38, 39, suggesting that removal of Wnt mutants by cell competition may be context-dependent.
To uncover the underlying molecular mechanism whereby Wnt activation potentiates basal extrusion of β-cat ΔN/RasV12-transformed cells, we conducted microarray analyses to search for molecules whose expression is changed in β-cat ΔN/RasV12 cells when co-cultured with β-cat ΔN cells. Among the top-listed upregulated genes, we identified one member of the matrix metalloproteases (MMPs) superfamily, MMP21 (Fig. 3e, f). MMPs are essential for tumor progression by degrading and remodeling extracellular matrix (ECM), enabling malignant cells to invade stomal tissues40, 41. This led us to assess expression levels of other MMP molecules by q-PCR analysis, which revealed that MMP21 was by far the most abundant MMP expressed in β-cat ΔN/RasV12 cells co-cultured with β-cat ΔN cells (Fig. 3g). These results imply that MMP21 has a unique role among MMPs and would represent a particular target for treatment of diffuse-type carcinomas. Immunofluorescence analysis, however, did not show overt alteration in endogenous MMP21 expression (data not shown). MMP21 is a 62-kDa proprotein that is activated via cleavage of the prodomain by furin or related proteases, and is secreted into the extracellular milieu as a 49-kDa active protease42. For this reason, we surmised that synthesized MMP21 is rapidly secreted out of cells, making it difficult to observe MMP21 intracellularly. We therefore hampered the transport machinery by treatment with brefeldin A (BFA). This manipulation resulted in profound accumulation of MMP21 in β-cat ΔN/RasV12 cells when surrounded by β-cat ΔN cells, while that in β-cat ΔN/RasV12 cells cultured alone was elevated to a lesser extent (Fig. 3h, i). Human MMP21 is the most recently cloned MMP gene43, therefore, its physiological function, including substrate specificity, is not fully understood. Based on its amino acid sequence, MMP21 cannot be classified as a collagenase, gelatinase, stromelysin, matrilysin, or membrane-type MMP43. To set out to characterize proteolytic capabilities of MMP21, a recombinant protein of the catalytic domain of MMP21 was produced using E. coli since the latent MMP21 protein is catalytically inert (Extended Data Fig. 6a). This recombinant protein was incubated with several ECMs, which are principal constituents of connective tissues. The MMP21 catalytic domain extensively degraded collagen type Ⅳ, and digested collagen type I and fibronectin, generating both small and large fragments. Intriguingly, the α- and β-chains of laminin were resistant to hydrolysis, while the γ-chain was barely detectable in the presence of MMP21, suggesting that MMP21 is a competent Zn2+-dependent endoproteinase with unique specificity for the γ-chain of laminin (Extended Data Fig. 6b). To examine the functional involvement of MMP21 in basal extrusion of transformants, we established β-cat ΔN/RasV12 cells stably expressing MMP21 short-hairpin RNA (shRNA) (Fig. 3j). MMP21-knockdown profoundly prevented basal extrusion of β-cat ΔN/RasV12 cells (Fig. 3k, l), highlighting an active role of MMP21 in non-cell autonomous basal extrusion of transformed cells. To bolster evidence for the requirement of MMP21 in this process, we evaluated the effect of a pan-MMP inhibitor, GM6001, and found that the frequency of basal extrusion was compromised to the same degree as MMP21-knockdown (Extended Data Fig. 6c, d).
We next explored the functional relevance of MMP21 in basal invasion of APCMin /RasV12 cells in vivo. For this purpose, we first evaluated MMP21 expression in intestinal epithelia and found that MMP21 was profoundly upregulated in RasV12-expressing cells produced mosaically within the APCMin-mutated epithelia, but not in GFP-expressing cells nor in single RasV12-transformed cells (Fig. 4a). To investigate the non-cell autonomous effect on MMP21 expression, intestinal organoids derived from APCMin-Villin-RasV12 mice were treated with either low-dose or high-dose tamoxifen. Mosaic expression of RasV12 mutants under the APCMin-background induced elevated expression of MMP21, but not in APCMin/RasV12 cells surrounded by one another, nor in APCMin/GFP cells or RasV12 cells (Fig. 4b), recapitulating non-cell autonomous upregulation of MMP21 in intestinal epithelia of mice. To examine the functional significance of MMP21 in basal delamination from normal mucosa, we utilized MMP21-deficient mice (Fig. 4c-f). Although most of progeny (93.9%; n = 132 mice) died immediately after birth, probably due to congenital heart disease caused by MMP21-abrogation44–46, we analysed the APCMin-Villin-RasV12-MMP21KO mice that were born healthy, grew normally, and did not exhibit apparent diseases. Consequently, MMP21-deficient APCMin/RasV12 cells were incompetent to drive basal extrusion of transformed cells (Fig. 4g, h). Collectively, these results demonstrate that MMP21 is one of the molecules that govern the diffuse invasion of transformed cells.
To identify the upstream regulator(s) that enhance(s) MMP21 expression, we first examined involvement of the Notch and TGF-β pathways as possible upstream regulators of MMP2147, 48. Yet, no alterations in activities of either pathway were evident by examining the transcriptional signature of β-cat ΔN/RasV12 cells co-cultured with β-cat ΔN cells, armed with the microarray datasets (Fig. 3e). Therefore, we conducted gene-set enrichment analyses (GSEA) to comprehensively overview differentially expressed genes, and found that production of IL-6 and chemokines substantially increased (Fig. 5a). This result is suggestive of activation of the NF-κB signaling pathway49, we thereby conducted a reporter assay using a plasmid with luciferase expression under the control of tandem repeats of NF-κB transcriptional response element. Accordingly, we confirmed the non-cell autonomous activation of NF-κB signaling (Fig. 5b). Notably, Wnt activation was negligible (Extended Data Fig. 7), implying that further enhancement of Wnt signaling in β-cat ΔN/RasV12 cells does not occur. To study the functional relevance of the NF-κB pathway in MMP21-mediated diffuse invasion of transformed cells, we treated cells with an inhibitor of NF-κB, BAY 11-7082, which inhibits IκB kinase (IKK). As a result, addition of BAY 11-7082 substantially diminished non-cell autonomous upregulation of MMP21, suggesting that NF-κB signaling acts as an upstream regulator to enhance MMP21 expression (Fig. 5c, d). We next queried whether the NF-κB complex directly regulates MMP21 transcription. To this end, we performed a chromatin immunoprecipitation (ChIP) assay followed by q-PCR with primers encompassing the promoter region of the MMP21 locus, which includes the p65 binding sequence, as predicted by the JASPAR2014 program (Fig. 5e). ChIP q-PCR experiments showed that activation of NF-κB signaling by TNF-α treatment significantly enhanced association of p65 with the promoter region of MMP21, indicating direct regulation of MMP21 expression by NF-κB signaling pathway (Fig. 5e). Furthermore, inhibition of NF-κB signaling profoundly suppressed basal extrusion of β-cat ΔN/RasV12 cells co-cultured with β-cat ΔN cells, and even enhanced their apical elimination in a dose-dependent fashion (Fig. 5f, g). These results indicate that NF-κB signaling is critical to dictate the direction in which cells are extruded, and its activation redirects transformed cells into basal delamination, with MMP21 being one of downstream targets.
To evaluate NF-κB signaling in vivo, we observed p65 expression and found that nuclear p65 protein profoundly accumulated in APCMin/RasV12 cells in intestinal epithelia, which was otherwise unaffected in APCMin/GFP- or single RasV12-mutated cells (Fig. 5h). Moreover, p65 elevation occurs in a non-cell autonomous fashion, as we found that APCMin/RasV12 cells surrounded by APCMin cells exhibited nuclear accumulation of p65 in intestinal organoids. In contrast, p65 intensity in APCMin/RasV12 mutants surrounded by themselves was no higher than that in RasV12 or APCMin/GFP cells (Fig. 5i). We next examined the effect of inhibition of NF-κB signaling on MMP21 expression and basal extrusion of APCMin/RasV12 cells. Accordingly, BAY 11-7082 treatment drastically diminished non-cell autonomous MMP21 upregulation and counteracted basal extrusion of APCMin/RasV12 cells in intestinal organoids (Fig. 5j-l). Furthermore, we challenged APCMin-Villin-RasV12 mice with SN50, a cell-permeable NF-κB inhibitory peptide, and found that the number of basally extruded cells was significantly reduced compared with the PBS-administered group (Fig. 5m, n). These in vivo results signify our conclusion that the NF-κB signaling positively regulates basal invasion of transformed cells through MMP21 upregulation (Extended Data Fig. 8).
Finally, we investigated whether the molecular mechanism underlying the onset of invasive carcinomas developed in APCMin-Villin-RasV12 mice is involved in human colorectal cancer. Cancer histopathology of APCMin-Villin-RasV12 mice does not mimic any pathological classes of regular human colorectal cancer, thereby, it is hard to obtain the corresponding human clinical samples. Moreover, cell competition is the event occurring in the initial phase of carcinogenesis. We therefore collected 9 tissue specimens from patients diagnosed with early colorectal cancer who received endoscopic treatments and stained those samples with MMP21, p65, β-catenin or p-ERK. β-catenin and p-ERK were analysed to evaluate activity of Wnt signaling and MAPK signaling, respectively. Histological scores (H-scores) of each protein were calculated based on percentages of positive cells showing, negative, weak, moderate, or strong staining intensity (Fig. 6a). Consequently, we found that higher expression of MMP21, nuclear β-catenin, p-ERK and nuclear p65 was observed in tumor tissues compared with normal tissues (Fig. 6b). Furthermore, the H-score of MMP21 was positively corelated with that of nuclear β-catenin, p-ERK or nuclear p65 in tumor tissues (Fig. 6c). Collectively, these results suggest that the NF-κB-MMP21 pathway is bolstered in Wnt- and MAPK-activated early colorectal cancers and may participate in early infiltration of malignant cells in humans.
There has been a surge of interest in cell competition for developing novel anti-cancer therapy. For this purpose, we need to fully understand whether and how cell competition malfunctions during the onset of carcinogenesis. In this study, we showed that preceding Wnt activation tips the balance in cell competition-induced cellular extrusion, which leads to an adverse outcome, illuminating an unanticipated outcome of cell competition.
As for molecular mechanisms underlying APCMin-induced functional perturbation of cell competition, we carefully examined any possibility of cellular aberrations that could be attributed to disoriented extrusion of RasV12-transformed cells. For instance, spindle misorientation associated with irregular extrusion has been described in dividing crypt cells of APCMin mice50–52. Nonetheless, basally extruding APCMin/RasV12 cells observed in this study were terminally differentiated non-proliferative enterocytes, arguing against the possibility that abnormal cell division provokes basal extrusion. Also, APC is involved in regulation of cellular polarity53, 54. This led us to investigate localization of junctional molecules, but we could not find any disturbance in either APCMin/RasV12 cells or β-cat ΔN/RasV12 cells (data not shown). From a molecular standpoint, we found that NF-κB signaling is potentiated when APCMin/RasV12 cells are surrounded by APCMin cells, resulting in elevated MMP21 expression, which is required for basal delamination. Moreover, we found that NF-κB signaling functions as a key determinant to switch the predominant direction of extrusion from apical to basal. Our findings are corroborated by the fact that in various forms of chronic inflammation, NF-κB signaling serves as a master regulator, comprising a key link between inflammation and cancer and robustly accelerates cancer cell invasion55. Since MMP21 abrogation did not recapitulate a phenotype similar to that provoked by NF-κB inhibition, MMP21 would be a downstream molecule governed by NF-κB signaling to conduct diffuse invasion. Hence, it remains to be clarified how the molecular landscape is remodeled to alter cell fates of APCMin/RasV12 compound mutants, which could be harnessed for therapeutic applications.
MMPs comprise a family of topologically related zinc endoproteases that degrade components of the extracellular matrix, or cause shedding of membrane-anchored proteins, and are involved in cancer, arthritis, multiple sclerosis, and cardiovascular disease56, 57. Although a catalytic domain containing the zinc-binding motif HEXXHXXGXXH is highly similar among MMPs, substrate specificity differs widely. In this study, we demonstrated that MMP21 exerts catalytic activity against collagen type Ⅰ, type Ⅳ, fibronectin, and the laminin γ-chain. Although upstream regulators for MMP21 expression had not been identified previously, our data indicate that NF-κB signaling directly regulates transcription of MMP21. The fact that inhibition of NF-κB signaling does not completely abolish the non-cell autonomous upregulation of MMP21 and expression of several MMPs is also under the control of NF-κB activity implies that factors beyond NF-κB signaling are likely involved in MMP21 expression58. Previous studies have revealed that expression of MMP21 is associated with a poor clinical outcomes in several types of cancers59. Moreover, we found that MMP21 is significantly upregulated and its expression is positively correlated with NF-κB signaling, Wnt signaling, and MAPK pathways in early human colorectal cancers. Thus, the present study is indicative of MMP21 as a novel therapeutic target to suppress expansion of early phase tumors.
The adenoma-carcinoma sequence model for colorectal cancer is broadly acknowledged60, 61. Nevertheless, our mouse model delineates a distinct mode of tumorigenesis in that malignant cells are diffusively produced within otherwise histologically normal tissues without adenoma components in proximity to tumors. What causes the difference in adenoma-carcinoma cancer or histopathogenesis of APCMin-Villin-RasV12 mice is of particular interest. One possible explanation is the different origin of cancer cells. In this study, we engineered mice to induce RasV12-mutation in terminally differentiated enterocytes. However, it is well known that cancer arises from an incipient transformation event occurring primary, but not exclusively, in the stem cell compartment. This is based on the facts that stem cells in any organs are intrinsically vulnerable to genetic mutations due to rapid cell division62, 63. Therefore, the current paradigm is that cancer arises from stem cells bearing sporadic mutation(s) that are inherited by all their progeny64. This renders clonal advantage to transformed cells and induces adenomatous transition, followed by incidental production of invasive cells at later phases. Although studies of human colorectal cancer originally put forward stem cells as the origin of disease, several models have shown that differentiated cells can also become principal contributors to tumor development65–68. Thus, it is tempting to speculate that differentiated cells with deleterious mutations cause different cancerous phenotypes, such as diffuse-type cancer, albeit rarely, due to the low incidence of somatic mutations occurring in those cells. Supporting this notion, it has been estimated that patients with the de novo diffuse-type colorectal cancer are quite rare or underestimated since it is barely detectable microscopically69. Furthermore, diffuse-type cancers generated in the digestive tract are difficult to treat, due to scarcely noticeable early symptoms and the highly invasive character of cancer cells. Therefore, it is of clinical significance to delineate the nature of disease pathogenesis, and there is a need for sensitive, novel treatments. Whether the NF-κB-MMP21 pathway is generally relevant to other diffuse-type cancers demands future investigation.
In conclusion, we discovered that Wnt activation disturbs proper cell competition, restrains the extrusion of RasV12-transformed cells, and confers invasive properties on transformed cells to escape primary epithelial sites. Previous reports also demonstrate that the fate of transformed cells upon cell competition can differ, depending on the preceding mutated background70, 71. This study further brings forth the prospect that cell competition constrains the order of appearance of mutations during tumor development, highlighting a new link between cell competition and carcinogenesis.
1. Yokoyama, A. et al. Age-related remodelling of oesophageal epithelia by mutated cancer drivers. Nature 565, 312-317 (2019).
2. Martincorena, I. et al. Somatic mutant clones colonize the human esophagus with age. Science 362, 911-917 (2018).
3. de Beco, S., Ziosi, M. & Johnston, L.A. New frontiers in cell competition. Dev. Dyn. 241, 831-841 (2012).
4. Bowling, S., Lawlor, K. & Rodriguez, T.A. Cell competition: the winners and losers of fitness selection. Development 146 (2019).
5. Vishwakarma, M. & Piddini, E. Outcompeting cancer. Nat. Rev. Cancer 20, 187-198 (2020).
6. Baker, N.E. Emerging mechanisms of cell competition. Nat. Rev. Genet. 21, 683-697 (2020).
7. Kon, S. Physiological and pathological relevance of cell competition in fly to mammals. Dev. Growth Differ. 60, 14-20 (2018).
8. Claveria, C. & Torres, M. Cell Competition: Mechanisms and Physiological Roles. Annu. Rev. Cell. Dev. Biol. 32, 411-439 (2016).
9. Merino, M.M., Levayer, R. & Moreno, E. Survival of the Fittest: Essential Roles of Cell Competition in Development, Aging, and Cancer. Trends Cell Biol. 26, 776-788 (2016).
10. Marques-Reis, M. & Moreno, E. Role of cell competition in ageing. Dev. Biol. 476, 79-87 (2021).
11. Amoyel, M. & Bach, E.A. Cell competition: how to eliminate your neighbours. Development 141, 988-1000 (2014).
12. Di Gregorio, A., Bowling, S. & Rodriguez, T.A. Cell Competition and Its Role in the Regulation of Cell Fitness from Development to Cancer. Dev. Cell 38, 621-634 (2016).
13. Morata, G. & Ripoll, P. Minutes: mutants of drosophila autonomously affecting cell division rate. Dev. Biol. 42, 211-221 (1975).
14. Levayer, R. & Moreno, E. Mechanisms of cell competition: themes and variations. J. Cell Biol. 200, 689-698 (2013).
15. Parker, T.M. et al. Cell competition in intratumoral and tumor microenvironment interactions. EMBO J. 40, e107271 (2021).
16. Nagata, R. & Igaki, T. Cell competition: Emerging mechanisms to eliminate neighbors. Dev. Growth Differ. 60, 522-530 (2018).
17. Kon, S. & Fujita, Y. Cell competition-induced apical elimination of transformed cells, EDAC, orchestrates the cellular homeostasis. Dev. Biol. 476, 112-116 (2021).
18. Maruyama, T. & Fujita, Y. Cell competition in mammals - novel homeostatic machinery for embryonic development and cancer prevention. Curr. Opin. Cell Biol. 48, 106-112 (2017).
19. Nichols, J., Lima, A. & Rodriguez, T.A. Cell competition and the regulative nature of early mammalian development. Cell Stem Cell 29, 1018-1030 (2022).
20. Madan, E., Gogna, R. & Moreno, E. Cell competition in development: information from flies and vertebrates. Curr. Opin. Cell Biol. 55, 150-157 (2018).
21. Kon, S. et al. Cell competition with normal epithelial cells promotes apical extrusion of transformed cells through metabolic changes. Nat. Cell Biol. 19, 530-541 (2017).
22. Sasaki, A. et al. Obesity Suppresses Cell-Competition-Mediated Apical Elimination of RasV12-Transformed Cells from Epithelial Tissues. Cell Rep 23, 974-982 (2018).
23. Sato, N. et al. The COX-2/PGE2 pathway suppresses apical elimination of RasV12-transformed cells from epithelia. Commun Biol 3, 132 (2020).
24. Mori, Y. et al. Extracellular ATP facilitates cell extrusion from epithelial layers mediated by cell competition or apoptosis. Curr. Biol. 32, 2144-2159 e2145 (2022).
25. Hata, M. et al. GPR30-Expressing Gastric Chief Cells Do Not Dedifferentiate But Are Eliminated via PDK-Dependent Cell Competition During Development of Metaplasia. Gastroenterology 158, 1650-1666.e1615 (2020).
26. Fearon, E.R. & Vogelstein, B. A Genetic Model for Colorectal Tumorigenesis. Cell 61, 759-767 (1990).
27. Hanahan, D. & Weinberg, R.A. The hallmarks of cancer. Cell 100, 57-70 (2000).
28. Vogelstein, B. & Kinzler, K.W. The multistep nature of cancer. Trends Genet. 9, 138-141 (1993).
29. Schmitt, M. & Greten, F.R. The inflammatory pathogenesis of colorectal cancer. Nat. Rev. Immunol. 21, 653-667 (2021).
30. Jen, J. et al. Molecular determinants of dysplasia in colorectal lesions. Cancer Res. 54, 5523-5526 (1994).
31. Powell, S.M. et al. APC mutations occur early during colorectal tumorigenesis. Nature 359, 235-237 (1992).
32. Moser, A.R., Pitot, H.C. & Dove, W.F. A Dominant Mutation That Predisposes to Multiple Intestinal Neoplasia in the Mouse. Science 247, 322-324 (1990).
33. el Marjou, F. et al. Tissue-specific and inducible Cre-mediated recombination in the gut epithelium. Genesis 39, 186-193 (2004).
34. Shoemaker, A.R., Gould, K.A., Luongo, C., Moser, A.R. & Dove, W.F. Studies of neoplasia in the Min mouse. Biochim Biophys Acta 1332, F25-48 (1997).
35. Lauren, P. THE TWO HISTOLOGICAL MAIN TYPES OF GASTRIC CARCINOMA: DIFFUSE AND SO-CALLED INTESTINAL-TYPE CARCINOMA. AN ATTEMPT AT A HISTO-CLINICAL CLASSIFICATION. Acta Pathol Microbiol Scand 64, 31-49 (1965).
36. Mueller, J. et al. EXPRESSION OF bcl-2 AND p53 INDE NOVO AND EX-ADENOMA COLON CARCINOMA: A COMPARATIVE IMMUNOHISTOCHEMICAL STUDY. The Journal of Pathology 180, 259-265 (1996).
37. Barth, A.I., Pollack, A.L., Altschuler, Y., Mostov, K.E. & Nelson, W.J. NH2-terminal deletion of beta-catenin results in stable colocalization of mutant beta-catenin with adenomatous polyposis coli protein and altered MDCK cell adhesion. J. Cell Biol. 136, 693-706 (1997).
38. Akieda, Y. et al. Cell competition corrects noisy Wnt morphogen gradients to achieve robust patterning in the zebrafish embryo. Nat Commun 10, 4710 (2019).
39. Brown, S. et al. Correction of aberrant growth preserves tissue homeostasis. Nature 548, 334-337 (2017).
40. Egeblad, M. & Werb, Z. New functions for the matrix metalloproteinases in cancer progression. Nat. Rev. Cancer 2, 161-174 (2002).
41. Cathcart, J., Pulkoski-Gross, A. & Cao, J. Targeting Matrix Metalloproteinases in Cancer: Bringing New Life to Old Ideas. Genes Dis 2, 26-34 (2015).
42. Marchenko, G.N., Marchenko, N.D. & Strongin, A.Y. The structure and regulation of the human and mouse matrix metalloproteinase-21 gene and protein. Biochem. J. 372, 503-515 (2003).
43. Ahokas, K. et al. Matrix metalloproteinase-21, the human orthologue for XMMP, is expressed during fetal development and in cancer. Gene 301, 31-41 (2002).
44. Yuan, Z.Z., Fan, L.L., Jiang, Z.C., Yang, Y.F. & Tan, Z.P. A Novel Nonsense MMP21 Variant Causes Dextrocardia and Congenital Heart Disease in a Han Chinese Patient. Front Cardiovasc Med 7, 582350 (2020).
45. Guimier, A. et al. MMP21 is mutated in human heterotaxy and is required for normal left-right asymmetry in vertebrates. Nat. Genet. 47, 1260-1263 (2015).
46. Akawi, N. et al. Discovery of four recessive developmental disorders using probabilistic genotype and phenotype matching among 4,125 families. Nat. Genet. 47, 1363-1369 (2015).
47. Ahokas, K. et al. Matrix metalloproteinase-21 is expressed epithelially during development and in cancer and is up-regulated by transforming growth factor-beta1 in keratinocytes. Lab. Invest. 83, 1887-1899 (2003).
48. Perles, Z. et al. A human laterality disorder caused by a homozygous deleterious mutation in MMP21. J. Med. Genet. 52, 840-847 (2015).
49. Liu, T., Zhang, L., Joo, D. & Sun, S.C. NF-kappaB signaling in inflammation. Signal Transduct Target Ther 2, 17023- (2017).
50. Caldwell, C.M., Green, R.A. & Kaplan, K.B. APC mutations lead to cytokinetic failures in vitro and tetraploid genotypes in Min mice. J. Cell Biol. 178, 1109-1120 (2007).
51. Fleming, E.S., Temchin, M., Wu, Q., Maggio-Price, L. & Tirnauer, J.S. Spindle misorientation in tumors from APC(min/+) mice. Mol. Carcinog. 48, 592-598 (2009).
52. Quyn, A.J. et al. Spindle orientation bias in gut epithelial stem cell compartments is lost in precancerous tissue. Cell Stem Cell 6, 175-181 (2010).
53. Aoki, K. & Taketo, M.M. Adenomatous polyposis coli (APC): a multi-functional tumor suppressor gene. J. Cell Sci. 120, 3327-3335 (2007).
54. McCartney, B.M. & Näthke, I.S. Cell regulation by the Apc protein Apc as master regulator of epithelia. Curr. Opin. Cell Biol. 20, 186-193 (2008).
55. Karin, M. & Greten, F.R. NF-kappaB: linking inflammation and immunity to cancer development and progression. Nat. Rev. Immunol. 5, 749-759 (2005).
56. Nagase, H. & Woessner, J.F., Jr. Matrix metalloproteinases. J. Biol. Chem. 274, 21491-21494 (1999).
57. Page-McCaw, A., Ewald, A.J. & Werb, Z. Matrix metalloproteinases and the regulation of tissue remodelling. Nat. Rev. Mol. Cell Biol. 8, 221-233 (2007).
58. Quintero-Fabian, S. et al. Role of Matrix Metalloproteinases in Angiogenesis and Cancer. Front Oncol 9, 1370 (2019).
59. Zhang, J. et al. Overexpression of MMP21 and MMP28 is associated with gastric cancer progression and poor prognosis. Oncol Lett 15, 7776-7782 (2018).
60. Vogelstein, B. et al. Genetic alterations during colorectal-tumor development. N Engl J Med 319, 525-532 (1988).
61. Leslie, A., Carey, F.A., Pratt, N.R. & Steele, R.J. The colorectal adenoma-carcinoma sequence. Br J Surg 89, 845-860 (2002).
62. Tomasetti, C., Li, L. & Vogelstein, B. Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science 355, 1330-1334 (2017).
63. Tomasetti, C. & Vogelstein, B. Cancer etiology. Variation in cancer risk among tissues can be explained by the number of stem cell divisions. Science 347, 78-81 (2015).
64. Barker, N. et al. Crypt stem cells as the cells-of-origin of intestinal cancer. Nature 457, 608-611 (2009).
65. Shih, I.M. et al. Top-down morphogenesis of colorectal tumors. Proc Natl Acad Sci U S A 98, 2640-2645 (2001).
66. Huels, D.J. & Sansom, O.J. Stem vs non-stem cell origin of colorectal cancer. Br. J. Cancer 113, 1-5 (2015).
67. Schwitalla, S. et al. Intestinal tumorigenesis initiated by dedifferentiation and acquisition of stem-cell-like properties. Cell 152, 25-38 (2013).
68. Westphalen, C.B. et al. Long-lived intestinal tuft cells serve as colon cancer-initiating cells. J. Clin. Invest. 124, 1283-1295 (2014).
69. Machida, H. et al. Narrow-band imaging in the diagnosis of colorectal mucosal lesions: a pilot study. Endoscopy 36, 1094-1098 (2004).
70. Watanabe, H. et al. Mutant p53-Expressing Cells Undergo Necroptosis via Cell Competition with the Neighboring Normal Epithelial Cells. Cell Rep 23, 3721-3729 (2018).
71. Kohashi, K. et al. Sequential oncogenic mutations influence cell competition. Curr. Biol. 31, 3984-3995 e3985 (2021).
72. Kawamoto, S. et al. A novel reporter mouse strain that expresses enhanced green fluorescent protein upon Cre-mediated recombination. FEBS Lett. 470, 263-268 (2000).
73. Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262-265 (2009).
74. Hogan, C. et al. Characterization of the interface between normal and transformed epithelial cells. Nat. Cell Biol. 11, 460-467 (2009).
75. Kajita, M. et al. Filamin acts as a key regulator in epithelial defence against transformed cells. Nat Commun 5, 4428 (2014).
76. Bernier-Latmani, J. & Petrova, T.V. High-resolution 3D analysis of mouse small-intestinal stroma. Nat Protoc 11, 1617-1629 (2016).
Antibodies and materials
The following antibodies were used in this study: chicken anti-GFP (ab13970), rabbit anti-Ki-67 (ab16667), rabbit anti-DCLK1 (ab31704), rabbit anti-LYVE1 (ab14917), mouse anti-chromogranin A (ab80787) and rabbit anti-p65 (ab16502) antibodies from Abcam, mouse anti-E-cadherin (clone 36;610181), mouse anti-β-catenin (clone 14;610153) and mouse anti-Villin (clone 12;610358) antibodies from BD Transduction, mouse anti-β-actin (clone AC-74;A2228), mouse anti-Pan-Ras (clone Ras10;OP40) and rabbit anti-laminin (L9393) antibodies from Sigma, mouse anti-mCherry antibody (632543) from Clontech, rabbit anti-Olfm4 (39141), rabbit anti-p-ERK (9101), rabbit anti-NF-κB (8242) and rabbit anti-IgG (2729) antibodies from Cell Signaling, rat anti-MECA32 (120502) from BioLegend, mouse anti-CDX2 (MU392A-UC) from BioGenex, rabbit anti-Sox9 (AB5535) from Millipore and mouse anti-synaptophysin (413831) from NICHIREI BIOSCIENCES INC. As for anti-MMP21 antibodies, rabbit anti-MMP21 (TA322032) antibody from ORIGENE was used for western blotting of MDCK lysate whereas rabbit anti-MMP21 (55289-1) antibody from Proteintech was used for mice tissues, and rabbit anti-MMP21 (PA1-25234) antibody from Life Technologies was used for immunostaining. Alexa-Fluor-568- and -647-conjugated phalloidin (Life Technologies) were used at 1.0 U ml−1. Alexa-Fluor-568- and -647-conjugated secondary antibodies were from Life Technologies, Alexa-Fluor-488-conjugated anti-chicken IgY antibody was from Abcam and Alexa-Fluor-647-conjugated anti-rat IgG was from Jackson immunoResearch. Tetramethylrhodamine-conjugated WGA was purchased from Life Technologies. Hoechst 33342 (Life Technologies) was used at a dilution of 1:2,000. The following inhibitors, BAY 11-7082 from Tokyo Chemical Industry, SN50 from Cayman Chemical and GM6001 from Calbiochem were used.
Mice
All animal experiments were conducted under the guidelines by the Animal Care Committee of Tokyo University of Science. The animal protocols were reviewed and approved by the Tokyo University of Science Animal Care Committee (Approval number: S21027). We used 6-10 weeks old C57BL/6 mice of either sex. Villin-CreERT2 mice33 were crossed with DNMT1-CAG-loxP-STOP-loxP-HRasV12-IRES-eGFP mice21 or CAG-loxP-STOP-loxP-eGFP mice72, and were further mated with APCMin mice32 to create Villin-RasV12, APCMin-Villin-RasV12 or APCMin-Villin-GFP mice respectively. Mice heterozygous for each transgene were used for experiments. We obtained the MMP21-null mice by CRISPR/Cas9-mediated genome engineering from Cyagen Biosciences. The 5636 bp ranging across Exon 1 – 7 of MMP21 gene located on chromosome 7 was selected as a target site (Fig. 4c). Cas9 and gRNA were co-injected into fertilized eggs for MMP21-knockout mice production. For PCR genotyping of mice, the following primers were used: 5’-CAAGCCTGGCTCGACGGCC-3’ and 5’-CGCGAACATCTTCAGGTTCT-3’ for the Villin-CreERT2 mice, 5’-CACTGTGGAATCTCGGCAGG-3’ and 5’-GCAATATGGTGGAAAATAAC-3’ for the DNMT1-CAG-loxP-STOP-loxP-HRasV12-IRES-eGFP mice, 5’-CAGTCAGTTGCTCAATGTACC-3’ and 5’-ACTGGTGAAACTCACCCA-3’ for the CAG-loxP-STOP-loxP-eGFP mice, 5’-GCCATCCCTTCACGTTAG-3’, 5’-TTCTGAGAAAGACAGAAGTTA-3’ and 5’-TTCCACTTTGGCATAAGGC-3’ for the APCMin mice, and 5’-TAGGAGCAAACCCAATCACTAAAG-3’, 5’-TCCCGCCTATTCTTTCTGCCCAGC-3’ and 5’-ATCAGACAGAACTATGTGTAACTC-3’ for the MMP21 KO mice. The expected sizes of PCR products were 220 bp for Villin-CreERT2, 403 bp for DNMT1-CAG-loxP-STOP-loxP-HRasV12-IRES-eGFP, 390 bp for CAG-loxP-STOP-loxP-eGFP, 619 bp and 331 bp for APCMin, and 565 bp for MMP21 WT allele and 531 bp for MMP21 KO allele. For culturing intestinal organoids, isolated crypts from the mouse small intestine were entrapped in Matrigel (Corning) and plated in a non-coated 35-mm glass bottom dish as previously described73. The crypts embedded in Matrigel were covered with Advanced DMEM/F12 supplemented with N2 (Invitrogen), B27 (Invitrogen), 50 ng ml–1 EGF (Peprotech), 100 ng ml–1 Noggin (Peprotech), 1.25 mM N-Acetylcysteine (Sigma) and R-spondin conditioned medium collected from 293T-HA-Rspol-Fc cells kindly provided by Dr. Calvin Kuo (Stanford University). After 96 h culture, organoids were incubated with 100 nM or 1 mM tamoxifen (Sigma) for 24 h to induce transgenes. Subsequently, tamoxifen was washed out, and organoids were cultured for the indicated times. For in vivo experiments, 6-10 weeks old Villin-RasV12, APCMin-Villin-RasV12 or APCMin-Villin-GFP mice were given a single intraperitoneal injection of 1 mg of tamoxifen in corn oil (Sigma), and were then sacrificed at the indicated days after Cre activation. For inhibition of NF-κB signaling, mice were intraperitoneally injected with 5 mg kg−1 of SN50 daily.
Cell culture
MDCK and MDCK-pTR GFP-RasV12 cells were cultured as previously described74. To establish MDCK mCherry-β-catenin Δ131 or MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells, cDNA of mCherry-β-catenin Δ131 was cloned into BamHI and EcoRI sites of a PB-EF1-MCS-IRES-Neo vector. MDCK or MDCK-pTR GFP-RasV12 cells were then transfected with PB-EF1-MCS-IRES-Neo-mCherry-β-catenin Δ131 by nucleofection (nucleofector 2b Kit L, Lonza), followed by selection in medium containing 800 mg ml-1 G418 (Invitrogen). MDCK cells stably expressing mCherry-β-catenin Δ131 in a doxycycline-inducible manner were established by transfecting MDCK cells with pPB-TRE3G mCherry-β-catenin Δ131, followed by selection in medium containing 5 mg ml-1 blasticidin (Invitrogen). To establish MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells stably expressing MMP21-shRNAs, each shRNA sequences were cloned into AgeI and EcoRI sites of a pLKO.1 vector. The following shRNA sequences were used: MMP21 shRNA1, 5’-CCGGGCATACTGGAAAGTAGTTAACCTCGAGGTTAACTACTTTCCAGTATGCTTTTTG-3’ and 5’-AATTCAAAAAGCATACTGGAAAGTAGTTAACCTCGAGGTTAACTACTTTCCAGTATGC-3’; MMP21 shRNA2, 5’-CCGGGGCAATTTCTATTTTTCAAATCTCGAGATTTGAAAAATAGAAATTGCCTTTTTG-3’ and 5’-AATTCAAAAAGGCAATTTCTATTTTTCAAATCTCGAGATTTGAAAAATAGAAATTGCC-3’. MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells were then infected with lentivirus carrying pLKO.1-MMP21 shRNAs, and were cultured in the 1 mg ml-1 puromycin-containing medium and subjected to limiting dilution. For tetracycline-inducible MDCK cell lines, 2 mg ml-1 of tetracycline (Sigma) was used to induce expression of RasV12 mutant except for MDCK-pTRE3G mCherry-β-catenin Δ131 cells for which 2 mg ml-1 of doxycycline (Sigma) was used. For immunofluorescence, cells were plated onto collagen gel-coated coverslips. Type-I collagen (Cellmatrix Type I-A) was obtained from Nitta Gelatin and was neutralized on ice to a final concentration of 2 mg ml–1 according to the manufacturer’s instructions. The mixture of type-I collagen and Matrigel at a ratio of 1:4 was used for evaluating basal extrusion of β-cat ΔN/RasV12 cells (Figs 3c,d,k,l, 5f,g and Extended Data Figs 6c,d).
Immunofluorescence and western blotting
For immunofluorescence, MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131, MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 MMP21 shRNAs, MDCK-pTRE3G mCherry-β-catenin Δ131 cells were mixed with MDCK or MDCK mCherry-β-catenin Δ131 cells at a ratio of 1:50 and plated onto collagen-coated coverslips as previously described74. The mixture of cells was incubated for 12-24 h, followed by tetracycline or doxycycline treatment for 24 h except for analyses of basal extrusions, which were examined after 48 h of tetracycline addition. Cells were fixed with 4% paraformaldehyde (PFA) in PBS and permeabilized as previously described75. All primary antibodies were used at 1:100, and all secondary antibodies were used at 1:200. Alexa-Fluor-647-conjugated phalloidin was incubated for 1 h at an ambient temperature. For immunofluorescence using intestinal organoids, cells grown in Matrigel were incubated with Cell Recovery Solution (Corning) for 8 min before fixation with 4% PFA. After fixation, cells were permeabilized in 0.5% Triton X-100/PBS for 1 h and blocked in 1% BSA/PBS for 1 h. For immunohistochemical examinations of small intestine, tissues were isolated, fixed with 10% Formalin Solution for 24 h and embedded in paraffin or Tissue-Tek O.C.T Compound (Sakura Finetek Japan Co.,Ltd.). For paraffin-embedded samples, the continuous 5 μm-thick sections were sliced. The antigen retrieval was carried out by heating slides for 40 min in 10 mM citrate (pH 6.0) solution. To carry out immunofluorescent staining, the sections were blocked with Block-Ace (DS Pharma Biomedical) and permeabilized with 0.1% Triton X-100 in PBS. For DAB staining, after primary and secondary antibody reactions, the samples were developed with DAB (NICHIREI BIOSCIENCES INC) for 10 min, and counterstained with hematoxylin (Sakura Finetek Japan Co.,Ltd.) or methyl green (FUJIFILM Wako). Subsequently, sections were dehydrated with alcohol gradient and treated with xylene to render the sliced transparent. To conduct HE staining, paraffin-embedded samples were sliced, deparaffinized and stained with hematoxylin for 3 min and stained by eosin solution (Sakura Finetek Japan Co.,Ltd.) for 30 sec. For frozen samples, 10 μm-thick sections were cut on a cryostat. The sections were blocked with Block-Ace and permeabilized with 0.1% Triton X-100 in PBS. Primary or secondary antibodies were incubated for overnight at 4 °C or 4 h at ambient temperature, respectively. All primary antibodies were used at 1:200, and secondary antibodies were at 1:500. Whole mount immunostaining of mouse small intestinal villi was performed as previously described76. Briefly, small intestine of mice was cut out and put longitudinally on dish to expose the lumen. After several washes with PBS, tissues were pinned on silicon plates and then fixed with 10% Formalin Solution overnight at 4 °C. The samples were then washed with PBS several times and subsequently dehydrated with 10% sucrose in PBS for 2 h, followed by incubation with 20% sucrose and 10% glycerol in PBS overnight at 4 °C afterwards. After blocking with 3% donkey serum (Jackson Immunoresearch Laboratories) in 0.5% Triton-X 100 in PBS for 2 h, samples were incubated with the indicated primary antibodies diluted in the blocking solution overnight at 4 °C, followed by secondary antibodies reaction. After washing with PBS, samples were mounted with Mowiol (Calbiochem). Immunofluorescence images of intestinal tissues and cultured cells were acquired using the Olympus FV1000 system, Nikon A1R system or KEYENCE BZ-x800. Images were quantified using the MetaMorph software (Molecular Devices) or the ImageJ/Fiji software. Western blotting was carried out as previously described74. Primary antibodies were used at 1:1,000. The western blotting data were analysed using ImageQuant LAS-3000 (GE Healthcare).
Microarray analysis
1.2×107 of 10:1 mix culture of MDCK mCherry-β-catenin Δ131 and MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells or a single culture of MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells were cultured in collagen type I-coated 10-cm plastic dishes. After incubation with tetracycline for 24 h, following Accutase (Nacalai tesque, INC.) treatment, GFP-positive cells were collected by an analytical flow cytometer FACS Aria II or Aria Ⅲ (BD Life Sciences). Total RNA was extracted from the isolated cells using ISOGEN II (NIPPON GENE Co., Ltd). The analysis of gene expression profiling was performed using the Canine (V2) Gene Expression Microarray, 4×44K (Agilent Technologies).
Reverse Transcription and Quantitative real-time PCR
MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells or a 10:1 mix culture of MDCK mCherry-β-catenin Δ131 and MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells were cultured at a density of 1.2×107 cells on collagen-coated 10 cm dishes (Corning). After incubation with tetracycline for 24 h, GFP-positive β-cat/RasV12 cells was separated with an analytical flow cytometer. Total RNA was extracted from the isolated cells using ISOGENII and reverse transcribed using a PrimeScript II Reverse Transcriptase (TAKARA) and Oligo(dT)15 Primer (TAKARA), Random Primer 80 nmol (TAKARA), Deoxynucleotide Mix, 10 mM Molecular Biology Reagent (Sigma), RNase Inhibitor, Recombinant (TOYOBO LIFE SCIENCE). Luna Universal qPCR Master Mix (NEW ENGLAND BioLabs) was used to perform q-PCR using Applied Biosystems 7500 Real-Time PCR System (Applied Biosystems) or QuantStudio 1 Real-Time PCR system (Applied Biosystems). Same procedures were conducted to evaluate Wnt-targeted genes using intestinal organoids. We used β-actin as a reference gene to normalize data. The primer pairs used in above analyses are listed in Supplementary Table 1.
Patient samples
Early colorectal tumor specimens were provided by the Department of Pathology at the University of Tokyo Hospital. We collected 9 formalin-fixed paraffin-embedded clinical samples from patients who were diagnosed with early colorectal cancer, and underwent endoscopic treatments by endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD). Clinical data were collected from electronic medical records. All procedures were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. All samples were reviewed by the University of Tokyo Hospital (no. 0542-(7)), and written informed consent was obtained from all of patients.
Reporter assay
To monitor Wnt activity, MDCK, MDCK mCherry-β-catenin Δ131, MDCK-pTR GFP-RasV12, MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 or MDCK-pTRE3G mCherry-β-catenin Δ131 cells were co-transfected with TOP FLASH or FOP FLASH and pRL-TK (Renilla luciferase plasmid). After 24 h, cells were tyrpsinised, and cultured alone or co-cultured with the indicated cells for 8 h, followed by tetracycline or doxycycline treatment for 16 h. Firefly luciferase activity was measured using Dual-Luciferase Reporter assay (Promega) and normalized by Renilla luciferase activity using SpectraMax iD5 (Molecular Devices). To evaluate activity of NF-κB signaling, MDCK-pTR GFP-RasV12 mCherry-β-catenin Δ131 cells were transfected with a pGL4.32 reporter plasmid encoding for the firefly luciferase gene under the control of promoters containing tandem NF-κB elements (Promega) along with a pRL-TK vector. Subsequently, cells were cultured alone or co-cultured with MDCK mCherry-β-catenin Δ131 cells and subjected to the same procedure above.
Purification and enzyme assay of recombinant MMP21 catalytic domain
A 474-bp fragment of the human MMP21 (171 a.a. – 328 a.a.) containing catalytic domain was amplified by a PCR method with primers (In-Fusion cloning sites are underlined) fwd 5’-GCCGCGCGGCAGCCATTCTCCAAGAGGACGCTG- 3’ and rev 5’-GCTTTGTTAGCAGCCGTTAGGAGCCATACAGCTTTTG- 3’ using the human MMP21 cDNA in the pcDNA3.1 vector (Addgene). The amplified fragment was ligated in frame into the pET21b expression vector (Novagen) using NEBuilder HiFi DNA Assembly kit (New England BioLabs), thereby adding a C-terminal His6 tag to the protein. The resulting vector was transformed into E. coli BL21 (DE3) cells grown in 2YT medium. MMP21 expression was induced by the addition of 0.5 mM isopropyl β-D-thiogalactopyranoside (IPTG) at OD600 = 0.6, followed by the further incubation for 12-16 h at 30 ℃. E. coli cells were collected by centrifugation and lysed with 50 mM Tris buffer (pH7.5) containing 5 mM CaCl2, 100 mM NaCl, 1 mM ZnSO4 and protease inhibitor cocktail (Roche). Inclusion bodies were solubilized in the same buffer containing 6 M GdnHCl and 5 mM DTT, and loaded on Ni-NTA agarose resin (QIAGEN). After washing with buffer containing 15 mM imidazole, the protein bound via the C-terminal His6 tag was eluted with a 50 mM Tris buffer (pH 7.5) containing 5 mM CaCl2, 100 mM NaCl and 500 mM imidazole. Refolding of the recombinant protein was achieved by two-step dilution (1:32) into a 50 mM Tris buffer (pH7.5) containing 10 mM CaCl2, 100 mM NaCl, 1 mM ZnSO4, 0.1% Brij-35 and 10% glycerol at 4 ℃. Refolded MMP21 catalytic domain (20 μg) was incubated with collagen type I (2.25 μg), collagen type Ⅳ (3 μg), fibronectin (2.4 μg) and laminin (3.33 μg) in 50 mM Tris (pH 7.5) buffer containing 150 mM NaCl, 5 mM CaCl2 and 0.05% Brij-35 for overnight at 37 ℃. The samples were analysed by SDS-PAGE to evaluate the proteolytic activity of MMP21 catalytic domain.
ChIP assay
Chromatin immunoprecipitation (ChIP) was performed using the SimpleChIP Plus Enzymatic Chromatin IP Kit (Magnetic Beads, Cell Signaling Technology, 9005 S) following the manufacture’s instructions. Briefly, MDCK-pTR GFP-RasV12 mCherry-β-cateninΔ131 cells were treated with or without 10 ng ml–1 TNF-α for 2 h. Subsequently, cells were cross-linked with 1% formaldehyde in PBS for 10 min at room temperature and quenched with 2.5 M glycine for 5 min. Nuclei were prepared, and chromatin was incubated with micrococcal nuclease at 37 °C for 20 min, and subjected to sonication using Sonifier 250A (BRANSON) to produce chromatin smears with an average size of 150-900 bp. The soluble chromatin supernatants (Input samples) were immunoprecipitated with the rabbit anti-p65 antibody (Cell signaling) alongside rabbit IgG control (Cell signaling) overnight at 4 °C. The immunocomplexes were then rotationally incubated with ChIP-Grade Protein G Magnetic Beads for 2 h at 4 °C. ChIP DNA was eluted in ChIP elution buffer (IP sample) and used for q-PCR analyses. The primer sequences used are listed in Supplementary Table 1. Cycle threshold (Ct) values were normalized to the 2% input sample (Percentage of input=2% × 2(C[T] 2%Input sample-C[T] IP sample)).
Statistical and reproducibility
For data analyses, unpaired two-tailed Student’s t-tests (Figs 1b,d,g, 3b,d,g,i,l, 4h, 5b,d,e,g,l,n and Extended Data Figs 1a, 3b, 5b, 6d, 7) were used to determine P-values using Microsoft Excel. P-values less than 0.05 were considered to be statically significant. For animal studies, experiments were not randomized, and investigators were not blinded to allocation during experiments. All results were reproduced in at least three mice for each experimental condition. Representative figures are shown in Figs 1a,c,e,f, 2a,b,c,d,e, 3c,h,k, 4a,b,e,f,g, 5c,f,h,i,j,k,m, 6a and Extended Data Figs 1b, 2a,b, 3a, 4a,b, 5a,c, 6a,b,c.
Data availability
All transcriptomic datasets generated in this study have been deposited into the Gene Expression Omnibus (GEO) under accession Number GSE217830. All other data supporting the findings of this study are available within the article and its Supplementary Information or from the corresponding author on reasonable request.
ACKNOWLEDGEMENTS
We thank C. Kuo (Stanford University) for the R-spondin-producing cell line. S.Kon was supported by the AMED Practical Research for Innovative Cancer control 19217462, Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research on (B) 20H03166, JSPS Grant-in-Aid for Scientific Research on Innovative Areas 21H00441, the Princess Takamatsu Cancer Research Fund, the MSD Life Science Foundation, and The Uehara Memorial Foundation.
AUTHOR CONTRIBUTIONS
K.N., Y.F. and S.Kon conceived and designed the experiments, and K.N. and S.Kon generated most of the data. H.L., S.Y., S.Kitamoto, E.A., M.K., and M.M. assisted immunohistochemical and pathological analyses of mouse tissue samples. S.T. established cell lines used in this study. J.Koseki conducted the GSEA analysis. K.S., J.Kurauchi, H.T. and H.Y. generated the recombinant MMP21 catalytic domain. Y.S., A.E., S.A., Y.H., T.U. assisted immunohistochemical analyses of clinical samples. The manuscript was written by K.N. and S.Kon with assistance from the other authors.
COMPETING INTERESTS
The authors declare no competing financial interests.