Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Mar 5;215(3):911-926.
doi: 10.1084/jem.20170934. Epub 2018 Feb 8.

The Cdx2 Homeobox Gene Suppresses Intestinal Tumorigenesis Through Non-Cell-Autonomous Mechanisms

Affiliations
Free PMC article

The Cdx2 Homeobox Gene Suppresses Intestinal Tumorigenesis Through Non-Cell-Autonomous Mechanisms

Camille Balbinot et al. J Exp Med. .
Free PMC article

Abstract

Developmental genes contribute to cancer, as reported for the homeobox gene Cdx2 playing a tumor suppressor role in the gut. In this study, we show that human colon cancers exhibiting the highest reduction in CDX2 expression belong to the serrated subtype with the worst evolution. In mice, mosaic knockout of Cdx2 in the adult intestinal epithelium induces the formation of imperfect gastric-type metaplastic lesions. The metaplastic knockout cells do not spontaneously become tumorigenic. However, they induce profound modifications of the microenvironment that facilitate the tumorigenic evolution of adjacent Cdx2-intact tumor-prone cells at the surface of the lesions through NF-κB activation, induction of inducible nitric oxide synthase, and stochastic loss of function of Apc This study presents a novel paradigm in that metaplastic cells, generally considered as precancerous, can induce tumorigenesis from neighboring nonmetaplastic cells without themselves becoming cancerous. It unveils the novel property of non-cell-autonomous tumor suppressor gene for the Cdx2 gene in the gut.

Figures

None
Figure 1.
Figure 1.
CDX2 gene expression level in 566 human colon cancers and 19 nontumoral samples of the GSE39582 dataset. (A) Boxplot of the level of CDX2 expression in the 443 CRC samples of the discovery set organized in the six subtypes according to Marisa et al. (2013) (C1–C6). (B) Boxplot of the CDX2 expression level in the 123 samples of the validation set organized in six subtypes. Data are given ± SD. (C) Disease-free survival comparing CDX2high versus CDX2low CRC in the GSE39582 dataset. The cutoff for low versus high CDX2 expression is fixed at the median of the C4 group. CDX2low patients exhibit a significantly reduced disease-free survival (P < 0.019). (D) Disease-free survival comparing CDX2high versus CDX2low CRC in the C4 subtype. CDX2low patients exhibit a significantly reduced disease-free survival (P = 0.009). P-values were calculated with the log-rank test.
Figure 2.
Figure 2.
Cecal lesion induced by mosaic gene knockout of Cdx2 in the adult gut epithelium. (A) Histology of cecal lesions in AhCreERT::Cdx2f/f mice 4 mo after βNF+Tam administration. Bar, 500 µm. The boxed region is magnified in the right panel. Bar, 250 µm. (B) Immunodetection of the Cdx2 protein. The protein is almost absent, except in few glands with intact Cdx2 (arrowheads) entrapped in the lesions. Bar, 100 µm. In A and B, n = 20 mice from five crossings. (C) Immunodetection of intestinal proteins (Cdx1, Muc2) and gastric proteins (Cldn18, Tff1, Tff2, Sox2) in the cecum of wild-type mice, cecal lesions of AhCreERT::Cdx2f/f mice, cecal heteroplasia of Cdx2+/− mice, and the stomach of wild-type mice. n = 4 animals per genotype from two crossings. Bars, 100 µm. (D) Heatmap comparison of transcriptomic data for intestinal and gastric genes in the cecum of wild-type mice (Cec), cecal lesions of AhCreERT::Cdx2f/f mice (CL), cecal heteroplasia of Cdx2+/− mice (Het), and the stomach of wild-type mice (Sto). Stomach-1 represents gastric genes up-regulated in CL and Het compared with Cec; Stomach-2 represents gastric genes up-regulated in Het but not in CL; Stomach-3 represents gastric genes up-regulated in neither CL nor in Het.
Figure 3.
Figure 3.
Functional characterization of the cecal lesions. (A) Gene ontology enrichment analysis for terms related to cancer in the transcriptome of the cecal lesions of AhCreERT::Cdx2f/f mice compared with the cecum of wild-type mice. (B) Immunohistochemical staining of Ki67, β-catenin, Sox9, and Olfm4 in the cecal lesions; the insets respectively represent the Sox9 and Olfm4 patterns in wild-type cecal glands. Bars: (β-catenin) 100 µm; (Ki67, Sox9, and Olfm4) 200 µm. (C) Coimmunofluorescence detection of β-catenin and Cdx2, β-catenin and Sox9, β-catenin and Olfm4, Cdx2 and Sox9, and Cdx2 and Olfm4 in serial sections of the stomach, jejunum, and cecum of wild-type mice and in the cecal lesions of AhCreERT::Cdx2f/f mice. In the cecal lesions, open squares show a gland depleted in Cdx2, and asterisks show a gland with intact Cdx2. The arrowheads point to the surface epithelium expressing Cdx2. Bars, 50 µm. Pictures in B and C are representative of the data obtained in n = 4 animals per genotype in two crossings.
Figure 4.
Figure 4.
Comparative expression patterns of the Cdx2, Sox9, and Olfm4 proteins in gastrointestinal lesions. (A) Foregut-type heteroplasia in the cecum of Cdx2+/− mice. Top: region at the border (dotted line) between the normal cecal mucosa (right side) and the heteroplastic tissue (left side). The normal epithelium expresses Cdx2, whereas Sox9 and Olfm4 are present at the bottom of the glands; the heteroplastic tissue, devoid of Cdx2, shows a strong expression of Sox9 but no expression of Olfm4. Bottom: rare clusters of glands in the heteroplasia expressing both Sox9 and Olfm4. Bars, 50 µm. n = 4 mice. (B) Human lesions: hyperplastic polyp in the stomach (n = 3); gastric-type metaplastic polyp in the duodenum (n = 1); and Meckel diverticulum with gastric-type differentiation (n = 3). For the normal small intestinal mucosa (bottom), the crypt-villous axis is from left to right (n = 3). Bars, 50 µm.
Figure 5.
Figure 5.
Mixed tumors developing by mosaic loss of Cdx2 combined with Apc heterozygosity. (A) Survival curve (Kaplan–Meier representation) of wild-type (WT), AhCreERT::Cdx2f/f (C), Apc+/Δ14 (A), and Apc+/Δ14::AhCreERT::Cdx2f/f (CA) mice; n = 15 animals of each genotype; p-values were calculated using the log-rank test; n.s., not significant. (B) Number of polyps in the small intestine, cecum, and colon of AhCreERT::Cdx2f/f (C), Apc+/Δ14::AhCreERT::Cdx2f/f (CA), and Apc+/Δ14 (A) mice; n = 10 mice of each genotype; data are given ± SD; p-values were calculated using the Wilcoxon–Mann–Whitney test; n.s., not significant. (C) Histology of a cecal mixed tumor in Apc+/Δ14::AhCreERT::Cdx2f/f mice with the juxtaposition of metaplastic-type (open square) and dysplastic areas (asterisk). Bar, 500 µm; n = 15 mice in three crossings. (D) Invasion beyond the muscularis mucosae. Bar, 100 µm. (E) Histology and immunostaining of β-catenin and Cdx2 in the metaplastic-type and dysplastic areas. Bars, 50 µm; n = 15 mice in three crossings.
Figure 6.
Figure 6.
Tracing of Cdx2-depleted cells in mixed tumors. Detection of β-catenin, Cdx2, and Tomato in small intestinal and cecal mixed tumors of Apc+/Δ14::AhCreERT::Cdx2f/f::RosaCAGtdTomato mice. The second line of pictures represents a higher magnification of the first one. Tomato was detected by direct fluorescence emission before indirect immunodetection of β-catenin and Cdx2. The mutually exclusive patterns of Tomato and cytoplasmic/nuclear β-catenin were obtained in three mixed tumors from the cecum and three mixed tumors from the small intestine coming from four mice in two independent crossings, in four sections analyzed in each sample. Bars, 100 µm.
Figure 7.
Figure 7.
Modification of the microenvironment in the cecal lesions of AhCreERT::Cdx2f/f mice. (A) Heatmap comparison of transcriptomic data for extracellular matrix genes (left) and cytokine genes (right) up-regulated in the cecal lesions of AhCreERT::Cdx2f/f mice (CL) compared with the cecum of wild-type mice (Cec). (B) Immunodetection of Iba1+ macrophages, CD4+ T lymphocytes, and FoxP3+ Treg lymphocytes in the normal cecal mucosa of wild-type mice and cecal lesions of AhCreERT::Cdx2f/f mice. Pictures correspond to serial sections. The asterisks indicate the lumen of the cecum. Bars: (cecum) 50 µm; (cecal lesions) 100 µm. (C) Erk1/2 and STAT3 signaling in the cecal lesions: the immunostaining illustrates the distribution of phospho-pErk1/2 and phospho-STAT3 in the cecal lesions compared with the normal cecum. Closed and open squares respectively show Cdx2-depleted and Cdx2-intact glands. Bars: (cecum) 50 µm; (cecal lesions) 100 µm. Pictures in B and C were obtained in four mice of each genotype from two crossings.
Figure 8.
Figure 8.
NF-κB activation and iNOS induction in the cecal lesions of AhCreERT::Cdx2f/f mice. Coimmunodetection of Cdx2 and β-catenin, RelA (p65 NF-κB) and β-catenin, iNOS and β-catenin, RelA and Cdx2, and RelA and Cdx2 in the normal cecal mucosa of wild-type mice and the cecal lesions of AhCreERT::Cdx2f/f mice. Open and closed arrows respectively show Cdx2-devoid and Cdx2-expressing surface epithelium. The asterisk points to a Cdx2-depleted gland underneath the surface epithelium. Pictures correspond to serial sections. They were obtained in four mice of each genotype from two independent crossings. Bars, 100 µm.
Figure 9.
Figure 9.
Surface initiation and iNOS-dependence of the dysplastic structures in mixed tumors. (A) Genotyping of microdissected glands for the Apcwt and ApcΔ14 alleles: (1) normal cecal gland and (2) Cdx2-depleted gland in cecal lesions of AhCreERT::Cdx2f/f mice; (3) normal and (4) dysplastic gland in Apc+/Δ14 mice; (5) normal cecal gland, (6) metaplastic-type gland, and (7) dysplastic gland in mixed tumors of Apc+/Δ14::AhCreERT::Cdx2f/f mice. PCR results are representative of the results obtained in n = 3 mice. (B) Immunodetection of β-catenin, Cdx2, Olfm4, and Sox9 in the Cdx2-expressing surface epithelium of mixed tumors. The arrowheads point to surface epithelial cells with cytoplasmic/nuclear β-catenin; the asterisks show Cdx2-depleted glands. Pictures correspond to serial sections. They were obtained in n = 10 mice. Bars, 50 µm. (C) Left: localization of the dysplastic glands in mixed tumors by β-catenin immunostaining, showing cytoplasmic/nuclear accumulation in glands connected to the surface epithelium. Bar, 100 µm. Right: distribution of dysplastic glands. Each bar represents the localization and extent of one dysplastic structure along the surface-to-bottom axis of mixed tumors. Results are expressed as percentage of the height of the mixed tumors. They correspond to data obtained from nine mixed tumors in seven mice. (D) Surface (square millimeters) of the metaplastic-type areas (light gray) and dysplastic areas (dark gray) in the lesions developed in the cecum of Apc+/Δ14::AhCreERT::Cdx2f/f mice either untreated or treated with AG. Values represent the mean surfaces in 12 sections for each sample. They were obtained from n = 4 AG-treated mice and n = 4 untreated mice.

Comment in

Similar articles

See all similar articles

Cited by 10 articles

See all "Cited by" articles

References

    1. Anders S., Pyl P.T., and Huber W. 2015. HTSeq—A Python framework to work with high-throughput sequencing data. Bioinformatics. 31:166–169. 10.1093/bioinformatics/btu638 - DOI - PMC - PubMed
    1. Andoniadou C.L., Matsushima D., Mousavy Gharavy S.N., Signore M., Mackintosh A.I., Schaeffer M., Gaston-Massuet C., Mollard P., Jacques T.S., Le Tissier P., et al. 2013. Sox2(+) stem/progenitor cells in the adult mouse pituitary support organ homeostasis and have tumor-inducing potential. Cell Stem Cell. 13:433–445. 10.1016/j.stem.2013.07.004 - DOI - PubMed
    1. Aoki K., Tamai Y., Horiike S., Oshima M., and Taketo M.M. 2003. Colonic polyposis caused by mTOR-mediated chromosomal instability in Apc+/Delta716 Cdx2+/- compound mutant mice. Nat. Genet. 35:323–330. 10.1038/ng1265 - DOI - PubMed
    1. Bae J.M., Lee T.H., Cho N.-Y., Kim T.-Y., and Kang G.H. 2015. Loss of CDX2 expression is associated with poor prognosis in colorectal cancer patients. World J. Gastroenterol. 21:1457–1467. 10.3748/wjg.v21.i5.1457 - DOI - PMC - PubMed
    1. Bahri R., Pateras I.S., D’Orlando O., Goyeneche-Patino D.A., Campbell M., Polansky J.K., Sandig H., Papaioannou M., Evangelou K., Foukas P.G., et al. 2015. IL-15 suppresses colitis-associated colon carcinogenesis by inducing antitumor immunity. OncoImmunology. 4:e1002721 10.1080/2162402X.2014.1002721 - DOI - PMC - PubMed

Publication types

MeSH terms

Feedback