Mutations in the adenomatous polyposis coli (APC) gene are pivotal in colorectal tumorigenesis. Existing mouse intestinal tumor models display mainly small intestinal lesions and carcinomas are rare. We defined human CDX2 sequences conferring colon epithelium-preferential transgene expression in the adult mouse. Mice carrying a CDX2P-NLS Cre recombinase transgene and a loxP-targeted Apc allele developed mainly colorectal tumors, with carcinomas seen in 6 of 36 (17%) of mice followed for 300 days. Like human colorectal lesions, the mouse tumors showed biallelic Apc inactivation, β-catenin dysregulation, global DNA hypomethylation, and aneuploidy. The predominantly distal colon and rectal distribution of tumors seen in mice where one Apc allele was inactivated in epithelial cells from distal ileum to rectum suggests that regional differences in the intestinal tract in the frequency and nature of secondary genetic and epigenetic events associated with adenoma outgrowth have a contributing role in determining where adenomas develop. The presence of large numbers of small intestine tumors seemed to inhibit colorectal tumor development in the mouse, and gender-specific effects on tumor multiplicity in the distal mouse colon and rectum mimic the situation in humans where males have a larger number of advanced adenomas and carcinomas in the distal colon and rectum than females. The mouse model of colon-preferential gene targeting described here should facilitate efforts to define novel factors and mechanisms contributing to human colon tumor pathogenesis, as well as work on tumor-promoting environmental factors and agents and strategies for cancer prevention and treatment. [Cancer Res2007;67(20):9721–30]

Colorectal cancer is a leading cause of cancer-related death in much of the industrialized world (1, 2). Despite much progress, we lack a full accounting of the molecular pathogenesis of colorectal cancer or key dietary and environmental factors in colorectal cancer development. A current view is that most colorectal cancers arise from adenomatous precursors, and accumulated gain-of-function mutations in proto-oncogenes and loss-of-function mutations in tumor suppressor genes underlie initiation of adenomatous lesions and progression to carcinoma (24). The identification of germ-line mutations predisposing to tumors in man and/or the mouse has highlighted genes with roles in sporadic tumor development and has clarified mechanisms via which cancers arise. For instance, whereas familial adenomatous polyposis (FAP) syndromes account for <1% of colorectal cancer cases, the identification of heterozygous germ-line mutations in the adenomatous polyposis coli (APC) gene in FAP patients led to efforts showing that somatic APC defects are present in ∼75% of sporadic colorectal adenomas and carcinomas (2, 4). Work on the APC gene also spurred studies that showed that a nonsense mutation at codon 850 of the murine Apc gene was responsible for predisposition to a multitude of small intestinal tumors in a mouse genetic model of multiple intestinal neoplasia (Min; refs. 4, 5). Other studies established that the APC protein regulates β-catenin, a key down-stream mediator of Wnt signals, with oncogenic effects as a transcriptional coactivator of T-cell factor proteins (reviewed in ref. 6).

C57BL6/J mice carrying the ApcMin mutation develop ∼50 adenomas and infrequent carcinomas in the small intestine by 120 to 140 days of age, along with rare colon adenomas and carcinomas (7, 8). The ApcMin mouse has been of keen interest for studying intestinal tumorigenesis mechanisms and potential cancer prevention and treatment strategies (8). Nevertheless, the ApcMin mouse has notable limitations as a model for human colorectal tumorigenesis. The ApcMin mouse manifests chiefly small intestinal lesions, but the vast majority of gastrointestinal tumors in FAP patients are found in the colon and rectum. Overall, small intestine carcinomas are rare in humans, with a prevalence that is about one thirtieth of that of colon carcinomas (9) Perhaps because ApcMin mice usually die by 140 days of age due to anemia and/or intestinal obstruction, few adenomas progress to carcinoma in ApcMin mice (8). Mice with other Apc germ-line mutations (1014) and mice in which Apc has been targeted for somatic inactivation by Cre recombinase (15) have been described. However, whereas some of these mice develop more colonic lesions than ApcMin mice, mice with constitutional Apc mutations develop predominantly small intestinal tumors. As such, they offer modest value over the ApcMin mouse for modeling adenoma-carcinoma progression in the colon. In fact, most genetically engineered mouse models of intestinal tumorigenesis manifest largely or exclusively small intestinal lesions (13), and adenoviral delivery of Cre to target Apc in distal colon tissues is a technically challenging approach described in only a single article (15).

Given this background, we sought to define sequence elements that might confer colon epithelium-preferential transgene expression in the mouse. Our previous studies revealed that the human CDX2 homeobox gene, which is expressed in epithelial cells throughout adult small intestine and colon, contained key cis elements for regulating transcription in cultured colon cancer cells (16). We show here that the CDX2 elements confer colon epithelium-preferential transgene expression in the adult mouse. Mice carrying a CDX2P-NLS Cre recombinase transgene and a floxed Apc allele developed colorectal adenomas, and carcinomas arose in ∼15% to 20% of mice. Morphologic and molecular studies of the mouse tumors revealed their similarity to human colorectal tumors, suggesting that mice in which the CDX2P-NLS Cre transgene is used to target Apc and other genes of interest will be valuable for studies in colorectal cancer prevention, diagnosis, and therapy.

Plasmid and transgene construction. Plasmid vectors for the CDX2P 9.5-NLS lacZ and CDX2P 0.5-NLS lacZ transgenes were constructed using the pnlacF vector (R. Palmiter, University of Washington), which provided the β-gal (lacZ) expression cassette with nuclear localization signal (NLS) sequences and a polyadenylation addition site from mouse protamine gene (mP1pA). The 9.5- and 0.5-kb fragments, containing various 5′-flanking sequences from the human CDX2 promoter region, have previously been described (16). These fragments were subcloned into the pnlacF vector. The CDX2P-NLS Cre transgene was generated by subcloning the 9.5-kb CDX2 promoter fragment along with a polyadenylation cassette from the bovine growth hormone gene (BGHpA) into α Cre-1 plasmid (S. Camper, University of Michigan, Ann Arbor, MI), which provides the NLS Cre recombinase cassette.

Mouse colony. We cared for mice and carried out experimental procedures under approval from the University Committee on Use and Care of Animals, University of Michigan and according to the Michigan state and U.S. federal regulations. Transgenic founders on a mixed background [C57BL6/J × SJL/J] were backcrossed with C57BL6/J mice. CDX2P 9.5-NLS Cre mice or Villin-Cre mice [B6.SJL-Tg(Vil-cre)997Gum/J] were intercrossed with mice carrying loxP-flanked Apc alleles homozygously (ApcloxP/loxP, 580S; ref. 15) or the loxP-Stop-loxP lacZ [Gt(ROSA)26Sor, R26R] reporter allele (17). Progeny from the cross between CDX2P 9.5-NLS Cre and ApcloxP/loxP were intercrossed to derive an outbred colony, segregating for the C57BL6/J and SJL genomes at a ratio of 93.75% and 6.25%, respectively. All the mice were housed in specific pathogen-free conditions. Teklad Mouse Breeder Diet 8626 (Harland-Teklad) and automatically supplied water were provided ad libitum to mice for tumorigenesis experiments. R26R reporter mice were kindly provided by S. Camper (University of Michigan, Ann Arbor, MI).

Tissue fixation and staining and quantitative analysis for β-galactosidase. Dissected gastrointestinal tissues were opened and washed with PBS containing 0.01% Triton X-100 at 4°C with agitation. After brief fixation with 4% paraformaldehyde containing 1.25 mmol/L EGTA, 2 mmol/L MgCl2 in PBS, tissues were placed in 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal) staining solution [1 mg/mL X-gal in N,N-dimethylformamide, 5 mmol/L K3Fe(CN)6, 5 mmol/L K4Fe(CN)6, 2 mmol/L MgCl2, 1.25 mmol/L MgCl2 in PBS] for 4 to 12 h at 37°C (18). The tissue was fixed with 0.2% glutaraldehyde for 10 min and with 4% paraformaldehyde for 4 h at 4°C. Embryos and tissues were then stored in 70% ethanol. For quantitative analysis of β-gal expression, 30 to 50 μg of total protein from homogenate were incubated in β-gal buffer (60 mmol/L Na2HPO4, 40 mmol/L NaH2PO4, 1 mmol/L MgCl2, 50 mmol/L β-mercaptoethanol, 2.2 mmol/L o-nitrophenyl-β-d-galactopyranoside) at 37°C for 3 h, and absorbance at 405 nm was measured.

Immunohistochemical staining. We carried out immunohistochemical analysis as described (19). The antibodies against β-catenin (BD Transduction Laboratories), CDX2 (19), p53 (Novocastra Laboratories, NCL-p53-CM5p), and 5-methylcytosine (Calbiochem, NA81) were used at dilutions of 1:200, 1:1,000, 1:1,000, and 1:200 (final concentration, 5 μg/mL), respectively.

Microdissection, Apc allele analysis, and p53 and KRAS analysis. Formalin-fixed, paraffin-embedded tissues were sectioned at 5 μm and mounted on glass slides, then weakly stained with hematoxylin. Specific regions (neoplastic versus non-neoplastic tissue) were carefully laser-capture microdissected using PixCell IIe apparatus (Arcturus Engineering). The number of cells that were captured ranged from 100 to 500 cells. DNAs were extracted in 30 to 50 μL of proteinase K buffer using PicoPure DNA extraction kit (Arcturus). Analysis for loss of heterozygosity of Apc alleles was done by multiplex PCR using Apc-P3 primer 5′-GTTCTGTATCATGGAAAGATAGGTGGTC-3′, Apc-P4 primer 5′-CACTCAAAACGCTTTTGAGGGTTGATTC-3′, Apc-P5 primer 5′-GAGTACGGGGTCTCTGTCTCAGTGAA-3′, targeted allele (580S), deleted allele (580D), and wild-type allele amplified as 314-bp (P3 and P4), 258-bp (P3 and P5), and 226-bp (P3 and P4) PCR products, respectively (15). For p53 mutational analysis, Trp53E5-S (5′-TCCAATGGTGCTTGGACAATGTG-3′), Trp53E5-AS (5′-CCTAAGAGCAAGAATAAGTCAG-3′), Trp53E6-S (5′-CTGCTCCGATGGTGATGGTAAG-3′), Trp53E6-AS (5′-CTCTAAGCCTAGCTAGCACTCAG-3′), Trp53E7-S (5′-AACTTATTAGAGGCTATAGCCAG-3′), Trp53E7-AS (5′-CATTTAGGTAGATAGGGTAGGAACC-3′), Trp53E8-S (5′-CCTTTGGCTGCAGATATGACAAG-3′), and Trp53E8-AS (5′-TGTGGAAGGAGAGAGCAAGAGGTG-3′) were used. For KRAS codon 12 or 13 mutational analysis, KRAS 1S (5′-TTATTGTAAGGCCTGCTGAA-3′) and KRAS 1AS (5′-GCAGCGTTACCTCTATCGTA-3′) were used.

Microsatellite stability analysis. Microsatellite analysis of mouse tumors was done as described using six microsatellite repeat markers previously shown to be informative in tumors from DNA mismatch repair–deficient mice (2022): A33, GA29, D7Mit91, D17Mit123, mBat-26, and mBat-37.

Primary culture of tumors. Primary culture of colon tumor cells was largely done as described (23, 24). Freshly isolated tumor specimens were minced with sterile razor blades, digested in HBSS containing 60 units/mL collagenase, 0.02 mg/mL dispase I, 0.2 mg/mL soybean trypsin inhibitor, and 2% bovine serum albumin for 1 h at 37°C. Tumor cells were separated and collected as pellet after sedimentation in DMEM containing 10% sorbitol and 5% fetal bovine serum. The pellets were washed with DMEM containing 2% sorbitol twice, and cells were cultured in 60-mm plates with NIH 3T3 feeder cells grown on Matrigel-coated surface and treated with 10 μg/mL mitomycin C for 2 h using DMEM/Ham's F-12 (1:1) with l-glutamate and 15 mmol/L HEPES containing 5 μmol/L dexamethasone, 10 μg/mL insulin, 5.5 μg/mL transferrin, 6.7 ng/mL selenium, 2 μg/mL ethanolamine, 5 μmol/L triiodothyronine, 10 ng/mL epidermal growth factor, 0.2% d-glucose, 50 μg/mL gentamicin, 100 units/mL penicillin, 100 μg/mL streptomycin, and 2% fetal bovine serum for 48 h.

Spectral karyotyping analysis. Spectral karyotyping analysis was done by treating the cells with 80 nmol/L calycullin A in Ham's F-10 medium for 1 h at 37°C to induce premature chromosome condensation (25) instead of inducing mitotic arrest. Chromosomal aberrations were quantified with an Olympus BX-61 microscope equipped with an Applied Spectral Imaging interferometer and 40× and 63× objectives. Chromosome counts were determined by analysis of 10, 8, 19, 10, and 13 independent chromosome spreads from one short-term cultured cell line from each of five independent adenomas.

CDX2 elements regulate gene expression in colon epithelium of adult mice. Previously, we found that a 9.5-kb fragment from the CDX2 proximal promoter region (CDX2P 9.5) strongly activated reporter gene expression in cultured colon cancer cells that expressed CDX2 transcripts (16). The CDX2P 9.5 reporter construct had weak activity in cells with reduced or absent CDX2 endogenous expression. Based on these data, we generated transgenic mice in which the CDX2P 9.5 fragment or a much shorter CDX2 promoter fragment (CDX2P 0.5) was present upstream of sequences encoding β-galactosidase (β-gal; Fig. 1A, CDX2P 9.5-NLS lacZ and CDX2P 0.5-NLS lacZ). Analysis of CDX2P 0.5-NLS lacZ transgenic mice revealed no specific and robust β-gal expression in the gastrointestinal tract, and they were not studied further. In adult mice from three independent transgenic lines carrying the CDX2P 9.5-NLS lacZ transgene, strong homogeneous β-gal expression was seen in epithelium of the cecum and proximal colon, with more variable expression in epithelium from the terminal ileum and distal colon in low-copy-number transgenic lines (Fig. 1B). Of the three lines, the highest copy line showed β-gal expression through a larger region of the distal ileum and in all colon epithelium (Fig. 1B). Restricted expression of the transgene in terminal ileum, cecum, and colon in adult high-copy transgenic CDX2P 9.5-NLS lacZ mice was confirmed with a quantitative assay (Fig. 1C).

Figure 1.

CDX2P 9.5-NLS lacZ, CDX2P 0.5-NLS lacZ, and CDX2P-NLS Cre transgene constructs and β-galactosidase expression in CDX2P 9.5-NLS lacZ in mice. A, schematic diagram of the endogenous CDX2 locus and the three transgene constructs used in this study. The start (ATG) codon for CDX2 protein translation is located in exon 1 (E1) and coding sequences for exons 2 and 3 (E2, E3) are indicated by filled boxes. BamHI (B) and XhoI (X) sites are indicated, along with the size of the BamHI fragment containing CDX2 probe sequence. The NLS lacZ and NLS Cre sequences cloned downstream of the CDX2P regulatory elements are shown. B, β-gal activity in the transgenic mice. Left, studies of mice carrying CDX2P 9.5-NLS lacZ (L, low-copy line; H, high-copy line) or CDX2P 0.5-NLS lacZ (a representative of five lines) indicated high transcriptional activity of CDX2P 9.5-kb fragment mainly in proximal colon in low-copy line and from terminal ileum to rectum in high-copy line, whereas no β-gal activity was observed in mice transgenic for the CDX2P 0.5-kb fragment. es, esophagus; st, stomach; je, jejunum; ce, cecum; co, colon; an, anus. Bar, 10 mm. Right, histologic findings of X-gal–stained gastrointestinal tract from CDX2P 9.5-NLS lacZ (H) mice for the stomach, jejunum, and colon, showing β-gal activity specifically in colon epithelium from crypt to the surface. Bar, 1 mm. C, analysis of β-gal activity in various tissues in CDX2P 9.5-NLS lacZ mice. No β-gal activity above background was observed in organs/tissues other than the gastrointestinal tract from terminal ileum to colon in adult transgenic CDX2P 9.5-NLS lacZ mice. White columns, low-copy line; black columns, high-copy line; t ileum, terminal ileum; a gland, adrenal gland; s muscle, skeletal muscle; fl, foreleg; hl, hind leg. D, analysis of β-gal expression during embryogenesis in CDX2P 9.5-NLS lacZ transgenic mice. β-gal expression in 8.5 dpc embryo (E8.5) viewed laterally, 10.5 dpc embryo (E10.5) viewed laterally and ventrally, and 12.5 dpc embryo (E12.5; second row) viewed laterally, ventrally, and dorsally. The expression of β-gal in gastrointestinal tract from esophagus to rectum was also analyzed at 12.5 dpc (third row). White arrowhead, position of cecum. Tissue on the left is from a CDX2P 9.5-NLS lacZ transgenic mouse and tissue on the right is from a wild-type mouse. Bar, 1 mm. Bottom, immunohistochemical staining of sagittal section of 14.5 dpc whole embryo after X-gal staining with mouse monoclonal anti-CDX2 antibody. Image of higher magnification (inset) indicates that β-gal is expressed in terminal ileum to colon (arrows; blue and brown color) but not, as shown in the inset, in other areas of intestine with CDX2 expression (brown color).

Figure 1.

CDX2P 9.5-NLS lacZ, CDX2P 0.5-NLS lacZ, and CDX2P-NLS Cre transgene constructs and β-galactosidase expression in CDX2P 9.5-NLS lacZ in mice. A, schematic diagram of the endogenous CDX2 locus and the three transgene constructs used in this study. The start (ATG) codon for CDX2 protein translation is located in exon 1 (E1) and coding sequences for exons 2 and 3 (E2, E3) are indicated by filled boxes. BamHI (B) and XhoI (X) sites are indicated, along with the size of the BamHI fragment containing CDX2 probe sequence. The NLS lacZ and NLS Cre sequences cloned downstream of the CDX2P regulatory elements are shown. B, β-gal activity in the transgenic mice. Left, studies of mice carrying CDX2P 9.5-NLS lacZ (L, low-copy line; H, high-copy line) or CDX2P 0.5-NLS lacZ (a representative of five lines) indicated high transcriptional activity of CDX2P 9.5-kb fragment mainly in proximal colon in low-copy line and from terminal ileum to rectum in high-copy line, whereas no β-gal activity was observed in mice transgenic for the CDX2P 0.5-kb fragment. es, esophagus; st, stomach; je, jejunum; ce, cecum; co, colon; an, anus. Bar, 10 mm. Right, histologic findings of X-gal–stained gastrointestinal tract from CDX2P 9.5-NLS lacZ (H) mice for the stomach, jejunum, and colon, showing β-gal activity specifically in colon epithelium from crypt to the surface. Bar, 1 mm. C, analysis of β-gal activity in various tissues in CDX2P 9.5-NLS lacZ mice. No β-gal activity above background was observed in organs/tissues other than the gastrointestinal tract from terminal ileum to colon in adult transgenic CDX2P 9.5-NLS lacZ mice. White columns, low-copy line; black columns, high-copy line; t ileum, terminal ileum; a gland, adrenal gland; s muscle, skeletal muscle; fl, foreleg; hl, hind leg. D, analysis of β-gal expression during embryogenesis in CDX2P 9.5-NLS lacZ transgenic mice. β-gal expression in 8.5 dpc embryo (E8.5) viewed laterally, 10.5 dpc embryo (E10.5) viewed laterally and ventrally, and 12.5 dpc embryo (E12.5; second row) viewed laterally, ventrally, and dorsally. The expression of β-gal in gastrointestinal tract from esophagus to rectum was also analyzed at 12.5 dpc (third row). White arrowhead, position of cecum. Tissue on the left is from a CDX2P 9.5-NLS lacZ transgenic mouse and tissue on the right is from a wild-type mouse. Bar, 1 mm. Bottom, immunohistochemical staining of sagittal section of 14.5 dpc whole embryo after X-gal staining with mouse monoclonal anti-CDX2 antibody. Image of higher magnification (inset) indicates that β-gal is expressed in terminal ileum to colon (arrows; blue and brown color) but not, as shown in the inset, in other areas of intestine with CDX2 expression (brown color).

Close modal

We analyzed β-gal expression in CDX2P 9.5-NLS lacZ transgenic embryos at 8.5, 10.5, and 12.5 days postconception (dpc; Fig. 1D). Consistent with prior data on endogenous Cdx2 expression during development (26), β-gal expression was seen in the tail bud and caudal part of the neural tube at 8.5 and 10.5 dpc. At 12.5 dpc, when Cdx2 expression in gut endoderm has been reported, β-gal expression was seen in the presumptive colon region of CDX2P 9.5-NLS lacZ transgenic embryos (Fig. 1D). We further documented β-gal expression in colon, but not small intestine, of embryos by analyzing endogenous CDX2 protein expression simultaneously via immunohistochemistry (Fig. 1E). The findings indicate that the CDX2P 9.5 fragment regulates gene expression in the caudal region of the embryo during early development, but, later in development and in the adult, the fragment is active predominantly in colon epithelium.

CDX2P 9.5-NLS Cre transgene–mediated targeting of loxP sites in vivo. We generated transgenic mice carrying a Cre recombinase gene controlled by the CDX2P 9.5 fragment (Fig. 1A; CDX2P-NLS Cre transgene) to assess the utility of the mice for somatic targeting of loxP-containing genes. We used a line with ∼50 copies of the transgene for subsequent studies. By crossing the CDX2P-NLS Cre mice with R26R reporter mice (17), we found that Cre recombinase had been active in epithelium from the distal ileum to the distal colon and from the crypt base to the luminal surface (Fig. 2A). Consistent with our findings from analysis of the embryonic expression pattern of the CDX2P 9.5-NLS lacZ transgene, the CDX2P-NLS Cre transgene mediated recombination that led to R26R reporter expression in extraintestinal cell populations of adult mice (e.g., tissues of the hind legs; Fig. 2B). The patterns of extraintestinal β-gal expression in adult mice were predictable based on CDX2P-NLS Cre transgene activity in the caudal region of the embryo (Fig. 2C).

Figure 2.

Expression of β-gal reporter gene in CDX2P-NLS Cre;R26R mice. A, left, expression of β-gal from terminal ileum to rectum. Bar, 10 mm. Right, expression of β-gal in colon epithelium from crypt to surface. Bar, 1 mm. B, quantitative analysis of β-gal expression in tissues/organs from CDX2P-NLS Cre;R26R mice. C, β-gal gene expression during embryogeneis in CDX2P-NLS Cre;R26R mice (12.5 dpc).

Figure 2.

Expression of β-gal reporter gene in CDX2P-NLS Cre;R26R mice. A, left, expression of β-gal from terminal ileum to rectum. Bar, 10 mm. Right, expression of β-gal in colon epithelium from crypt to surface. Bar, 1 mm. B, quantitative analysis of β-gal expression in tissues/organs from CDX2P-NLS Cre;R26R mice. C, β-gal gene expression during embryogeneis in CDX2P-NLS Cre;R26R mice (12.5 dpc).

Close modal

CDX2P-NLS Cre;Apc+/loxP mice develop colorectal adenomas and carcinomas. Mice carrying an Apc allele in which loxP sites flank exon 14 were crossed with CDX2P-NLS Cre mice to generate CDX2P-NLS Cre;Apc+/loxP (CPC;Apc) mice. Cre targeting of the ApcloxP allele deletes exon 14, leading to a frameshift mutation at codon 580 and a truncated APC protein (15). We also targeted Apc for conditional inactivation using another transgenic strain expressing Cre in the gastrointestinal tract (i.e., Villin-Cre mice; ref. 18). Using this strain, Cre targeting of loxP-containing genes has been reported in epithelium throughout the small intestine and colon. As shown in Table 1, 36 CPC;Apc mice, 6 Villin-Cre;Apc mice, and 28 Apc+/loxP (control) mice were studied for up to 300 days. Thirteen (36%) CPC;Apc mice survived up to the study end point. Three (8.3%) CPC;Apc mice died, likely due to intestinal obstruction by tumor, with the remaining 20 CPC;Apc mice euthanized on signs of distress. The average observation period for CPC;Apc mice was ∼200 days (Table 1), and the Villin-Cre;Apc and Apc+/loxP mice were euthanized at similar ages to compare phenotypes. The CPC;Apc mice manifested only mild anemia, with hematocrit values in the range of 33% to 36% (Table 1), in spite of rectal prolapse with intermittent bleeding in 19 (53%) mice. The growth (body weight increase) of male, but not female, CPC;Apc mice was inhibited relative to control mice after 120 days (Fig. 3A). Male and female Villin-Cre;Apc mice suffered more severe anemia than CPC;Apc mice and their growth was inhibited compared with control mice (Table 1). ApcMin mice in our facility (data not shown) and in published reports (7, 8) live only 120 to 140 days, due largely to severe anemia, with hematocrit values of 10% to 20% by 100 to 110 days.

Table 1.

Phenotypic characteristics of CPC;Apc, Villin-Cre;Apc, and Apc+/loxP mice

nAge (d)Body weight (g)Hematocrit (%)Total polyp number (per mouse)Small intestine
Cecum
Colon*
Size (mm)
Size (mm)
Size (mm)
<55–10<55–10>10<55–10>10
CPC;Apc                 
    Total 36 203 ± 26 26.2 ± 3.4 34.8 ± 9.7 380 (10.6) 109 (3.0) 105 40 (1.1) 27 11 231 (6.4) 181 49 
    Male 18 196 ± 22 28.2 ± 3.1 32.8 ± 7.9 212 (11.8) 49 (2.7) 48 22 (1.2) 16 141 (7.8) 107 33 
    Female 18 207 ± 29 24.0 ± 2.3 36.1 ± 10.8 168 (9.3) 60 (3.3) 57 18 (1.0) 11 90 (5.0) 74 16 
Villin-Cre;Apc                 
    Total 203 ± 2 28.4 ± 4.6 24.2 ± 7.7 217 (36.2) 185 (30.8) 175 10 3 (0.5) 29 (4.8) 26 
    Male 202 ± 0 30.7 ± 3.6 21.0 ± 6.3 177 (44.3) 150 (37.5) 142 2 (0.5) 25 (6.3) 24 
    Female 205 ± 0 23.8 ± 1.7 30.5 ± 7.8 40 (20.0) 35 (17.5) 33 1 (0.5) 4 (2.0) 
Apc+/loxP                 
    Total 28 202 ± 24 31.2 ± 7.8 45.1 ± 6.9            
    Male 15 204 ± 22 36.7 ± 6.0 44.2 ± 5.4            
    Female 13 199 ± 27 24.3 ± 1.9 46.1 ± 8.4            
nAge (d)Body weight (g)Hematocrit (%)Total polyp number (per mouse)Small intestine
Cecum
Colon*
Size (mm)
Size (mm)
Size (mm)
<55–10<55–10>10<55–10>10
CPC;Apc                 
    Total 36 203 ± 26 26.2 ± 3.4 34.8 ± 9.7 380 (10.6) 109 (3.0) 105 40 (1.1) 27 11 231 (6.4) 181 49 
    Male 18 196 ± 22 28.2 ± 3.1 32.8 ± 7.9 212 (11.8) 49 (2.7) 48 22 (1.2) 16 141 (7.8) 107 33 
    Female 18 207 ± 29 24.0 ± 2.3 36.1 ± 10.8 168 (9.3) 60 (3.3) 57 18 (1.0) 11 90 (5.0) 74 16 
Villin-Cre;Apc                 
    Total 203 ± 2 28.4 ± 4.6 24.2 ± 7.7 217 (36.2) 185 (30.8) 175 10 3 (0.5) 29 (4.8) 26 
    Male 202 ± 0 30.7 ± 3.6 21.0 ± 6.3 177 (44.3) 150 (37.5) 142 2 (0.5) 25 (6.3) 24 
    Female 205 ± 0 23.8 ± 1.7 30.5 ± 7.8 40 (20.0) 35 (17.5) 33 1 (0.5) 4 (2.0) 
Apc+/loxP                 
    Total 28 202 ± 24 31.2 ± 7.8 45.1 ± 6.9            
    Male 15 204 ± 22 36.7 ± 6.0 44.2 ± 5.4            
    Female 13 199 ± 27 24.3 ± 1.9 46.1 ± 8.4            
*

Colon and rectal lesions were scored collectively as colon.

Figure 3.

Tumorigenesis in CPC;Apc and Villin-Cre;Apc mice. A, body weights (in grams) of male or female CPC;Apc mice and Apc+/loxP control mice. Red, CPC;Apc male; orange, CPC;Apc female; blue, Apc+/loxP male; green, Apc+/loxP female. Points, mean; bars, SD. B, analysis of Apc alleles in the gastrointestinal tract of CPC;Apc mice and Villin-Cre;Apc mice. The wild-type Apc allele (WT) is represented by a 226-bp fragment. The floxed Apc allele (580S) is represented by a 314-bp fragment, whereas the Cre-targeted (inactive) allele (580D) is represented by a band of 258 bp. C, distribution of tumors in CPC;Apc male (n = 18) and female (n = 18) mice and Villin-Cre;Apc male (n = 4) and female (n = 2) mice. Circles, tumors <5 mm; triangles, tumors 5 to 10 mm; squares, tumors >10 mm. Black symbols indicate adenomas, orange indicate intramucosal carcinomas, and red indicate invasive carcinomas. D, examples of intestinal and colorectal tumors arising in 10-month-old female CPC;Apc (top) and Villin-Cre;Apc (middle) mice. Arrowheads, distal intestinal tumors; arrow, a cecal tumor. The distal colon tumors in CPC;Apc mice are boxed. Bar, 10 mm. Bottom, distal colorectal tumors seen in a male CPC;Apc mouse at 119 d of age; five colorectal tumors ranging in size from 2 to 5 mm (arrows) are seen.

Figure 3.

Tumorigenesis in CPC;Apc and Villin-Cre;Apc mice. A, body weights (in grams) of male or female CPC;Apc mice and Apc+/loxP control mice. Red, CPC;Apc male; orange, CPC;Apc female; blue, Apc+/loxP male; green, Apc+/loxP female. Points, mean; bars, SD. B, analysis of Apc alleles in the gastrointestinal tract of CPC;Apc mice and Villin-Cre;Apc mice. The wild-type Apc allele (WT) is represented by a 226-bp fragment. The floxed Apc allele (580S) is represented by a 314-bp fragment, whereas the Cre-targeted (inactive) allele (580D) is represented by a band of 258 bp. C, distribution of tumors in CPC;Apc male (n = 18) and female (n = 18) mice and Villin-Cre;Apc male (n = 4) and female (n = 2) mice. Circles, tumors <5 mm; triangles, tumors 5 to 10 mm; squares, tumors >10 mm. Black symbols indicate adenomas, orange indicate intramucosal carcinomas, and red indicate invasive carcinomas. D, examples of intestinal and colorectal tumors arising in 10-month-old female CPC;Apc (top) and Villin-Cre;Apc (middle) mice. Arrowheads, distal intestinal tumors; arrow, a cecal tumor. The distal colon tumors in CPC;Apc mice are boxed. Bar, 10 mm. Bottom, distal colorectal tumors seen in a male CPC;Apc mouse at 119 d of age; five colorectal tumors ranging in size from 2 to 5 mm (arrows) are seen.

Close modal

Somatic inactivation of the floxed Apc allele in the gastrointestinal tract of CPC;Apc and Villin-Cre;Apc mice was assessed with an Apc allele–specific multiplex PCR (15). In CPC;Apc mice, we found that the Cre-targeted Apc (580D) allele was present along with the wild-type Apc allele in normal mucosa from terminal ileum to the rectum, whereas in Villin-Cre;Apc mice, the 580D Apc allele was present from duodenum to the rectum (Fig. 3B). We found, on average, 11.8 tumors in male CPC;Apc mice and 9.3 tumors in female CPC;Apc mice (Table 1). On average, 5 to 8 tumors were found in the colon and rectum of a CPC;Apc mouse (scored as “colon” in Table 1, due to uncertainty about the exact anatomic demarcation between colon and rectum) and 1 tumor was found in the cecum, along with 3 tumors in distal small intestine (Table 1; Fig. 3C and D). Colorectal tumors could be seen even at early time points. A CPC;Apc mouse euthanized at 119 days of age because of unresolved rectal prolapse had 5 distal tumors ranging in size from 2 to 5 mm (Fig. 3E). Interestingly, whereas male and female CPC;Apc mice had similar numbers of small intestinal and cecal tumors, male CPC;Apc mice had ∼60% more colon tumors (Table 1; Fig. 3C). In Villin-Cre;Apc mice, we found ∼36 tumors per mouse, the vast majority of which were located in the small intestine (Table 1; Fig. 3C and D). The cecal and colon tumors observed in Villin-Cre;Apc mice were smaller than those in similarly aged CPC;Apc mice (Table 1; Fig. 3C and D). Because few Villin-Cre;Apc mice were studied, the significance of the apparent gender differences in tumor multiplicity and distribution in Villin-Cre;Apc mice is uncertain. No tumors were observed in similarly aged control Apc+/loxP mice (Table 1) or in CDX2P-NLS Cre or Villin-Cre mice (data not shown). Embryonic lethality was seen in CDX2P-NLS Cre and Villin-Cre transgenic mice homozygous for the ApcloxP allele.

Colon tumors in CPC;Apc mice manifest features akin to those in human tumors. Microscopic examination of the 231 colorectal tumors that arose in the 36 CPC;Apc mice revealed that 90% of the lesions resembled pedunculated-type tumors in man, in which a relatively well-differentiated, epithelial-rich lesion with tubular-type glands is attached to the mucosa by a stroma-rich stalk (Fig. 4A). In contrast, in 10% of colorectal lesions and 54% of the 40 cecal lesions identified, the tumors displayed a sessile appearance, where the neoplastic glands were minimally, if at all, raised above the mucosal surface (Fig. 4B). Of the 109 small intestine lesions from CPC;Apc mice studied, 71% had a sessile appearance and 29% were pedunculated. In the cohort of 36 CPC;Apc mice, we defined the most advanced lesion histologically, with 22 (66.1%) mice harboring at least one tumor with intramucosal carcinoma features. Six of the 36 (16%) CPC;Apc mice had a total of eight lesions defined as invasive carcinoma into submucosa, with associated desmoplasia and multiple invading irregular glands seen (Fig. 4C). Two of the eight carcinomas were in the cecum and three were in the colon. Intramucosal carcinoma was far more common in tumors >5 mm, but there was no clear relationship between tumor size and the presence of invasive carcinoma into submucosa (data not shown). None of the six CPC;Apc mice with invasive carcinoma had detectable lymph node or distant metastases. Even in the case of a 20-month-old CPC;Apc mouse with a large, deeply invasive distal colorectal carcinoma (Fig. 4D), lymph node or distant metastases were not found. Mammary gland tumors, with histologic features like those reported in ApcMin mice (27), were seen in 2 of 36 CPC;Apc mice. In one Villin-Cre;Apc mouse, a hepatic adenoma was found. No other extracolonic tumors were observed in the CPC;Apc or Villin-Cre;Apc mice studied.

Figure 4.

Histologic analysis of tumors in CPC;Apc mice. H&E staining of a distal colon tumor of pedunculated type (A) and a cecal lesion of sessile type (B). Bar, 1 mm. C, H&E staining (left) and immunohistochemical staining with anti–β-catenin antibody in a tumor showing invasive adenocarcinoma. Bar, 1 mm. Insets, invasion of multiple malignant glands into the submucosa (bar, 100 μm). D, example of a large distal colon carcinoma in a 20-month-old female CPC;Apc mouse. Dissection microscopic image (bar, 5 mm) and H&E staining (bar, 100 μm; m, mucosae; sm, submucosal mp, muscularis propria) reveal sessile appearance with invasive carcinoma, along with desmoplasia and inflammatory cells, in the muscular wall. Arrowheads, invasive front of the neoplastic cells. Immunostaining for β-catenin and p53 in serial sections of the tumor indicates nuclear and cytosol accumulation of β-catenin and strong nuclear accumulation of p53 in the neoplastic elements.

Figure 4.

Histologic analysis of tumors in CPC;Apc mice. H&E staining of a distal colon tumor of pedunculated type (A) and a cecal lesion of sessile type (B). Bar, 1 mm. C, H&E staining (left) and immunohistochemical staining with anti–β-catenin antibody in a tumor showing invasive adenocarcinoma. Bar, 1 mm. Insets, invasion of multiple malignant glands into the submucosa (bar, 100 μm). D, example of a large distal colon carcinoma in a 20-month-old female CPC;Apc mouse. Dissection microscopic image (bar, 5 mm) and H&E staining (bar, 100 μm; m, mucosae; sm, submucosal mp, muscularis propria) reveal sessile appearance with invasive carcinoma, along with desmoplasia and inflammatory cells, in the muscular wall. Arrowheads, invasive front of the neoplastic cells. Immunostaining for β-catenin and p53 in serial sections of the tumor indicates nuclear and cytosol accumulation of β-catenin and strong nuclear accumulation of p53 in the neoplastic elements.

Close modal

Somatic gene defects in CPC;Apc mouse colon tumors. Prior studies in mice carrying constitutional, heterozygous inactivating mutations in the Apc gene (ApcMin, Apc+/Δ714, Apc+/Δ1638N, and Apc+/Δ14) have shown that the wild-type Apc allele is inactivated in small intestine and colon adenomas, most often by allelic loss, at least in the case of C57BL6/J mice (11, 12, 14, 28). From a total of 15 CPC;Apc mice, we isolated DNA from neoplastic cells of 129 tumors, along with DNA from adjacent normal epithelium. In 109 of 129 (84.5%) tumors analyzed, the intensity of the wild-type Apc allele was greatly reduced, relative to the pattern for the wild-type and Cre-targeted (580D) Apc alleles in normal mucosa (Fig. 5A). Based on inefficiencies in eliminating nonneoplastic elements, even in laser-capture microdissection approaches, and alternative mechanisms of inactivating Apc besides allele loss, our analysis likely underestimated the frequency of wild-type Apc allele inactivation in tumors. Indeed, consistent with the view that loss of the APC protein function in regulating β-catenin levels and localization contributed to the development of nearly all tumors, immunohistochemical analysis of 97 tumors from 19 different CPC;Apc mice revealed strong β-catenin accumulation in the nucleus and cytosol of neoplastic cells in all 97 tumors, as compared with the membrane-associated β-catenin staining seen in adjacent normal epithelium (Fig. 5B). Alterations in the murine Cdx2 gene have been suggested to cooperate with Apc defects in intestinal tumorigenesis (29, 30). We found reduced or absent Cdx2 protein expression in the neoplastic cells of 88 of the 97 (91%) tumors analyzed (Fig. 5B).

Figure 5.

Molecular analysis of colorectal tumors arising in CPC;Apc mice. Analysis of mutant and wild-type Apc alleles. Microdissected DNAs from normal epithelium and four independent tumors were analyzed by multiplex PCR. WT indicates wild-type Apc allele at 226 bp; 580S is the floxed Apc allele of 314 bp; and 580D is the Cre-targeted Apc allele of 258 bp. Morphology by H&E staining, accompanied by analysis of β-catenin, CDX2, p53, and 5-methylcytosine (5-MeCyd) in serial sections of a representative colon tumor and adjacent normal colon epithelium. Nuclear and cytosol accumulation of β-catenin is seen in neoplastic elements relative to membrane staining in normal epithelium. Left to right, reduced CDX2 expression, moderate accumulation of p53, and reduced 5-methylcytosine staining, respectively, are seen in neoplastic glands relative to normal epithelium. Bar, 50 μm. Spectral karyotype analysis (SKY) and DAPI staining of primary colon tumor cells from CPC;Apc mice. Average number of chromosomes was 58.56 (±17.68, SD) with no translocations.

Figure 5.

Molecular analysis of colorectal tumors arising in CPC;Apc mice. Analysis of mutant and wild-type Apc alleles. Microdissected DNAs from normal epithelium and four independent tumors were analyzed by multiplex PCR. WT indicates wild-type Apc allele at 226 bp; 580S is the floxed Apc allele of 314 bp; and 580D is the Cre-targeted Apc allele of 258 bp. Morphology by H&E staining, accompanied by analysis of β-catenin, CDX2, p53, and 5-methylcytosine (5-MeCyd) in serial sections of a representative colon tumor and adjacent normal colon epithelium. Nuclear and cytosol accumulation of β-catenin is seen in neoplastic elements relative to membrane staining in normal epithelium. Left to right, reduced CDX2 expression, moderate accumulation of p53, and reduced 5-methylcytosine staining, respectively, are seen in neoplastic glands relative to normal epithelium. Bar, 50 μm. Spectral karyotype analysis (SKY) and DAPI staining of primary colon tumor cells from CPC;Apc mice. Average number of chromosomes was 58.56 (±17.68, SD) with no translocations.

Close modal

In the case of human colon tumors, missense mutations of codon 12 of the KRAS gene are present in 40% to 45% of sporadic adenomas and carcinomas and KRAS codon 13 missense mutations are found in a fraction of tumors (2, 3). In our sequence-based studies of 97 tumors from the CPC;Apc mice, we did not identify any mutations of KRAS codon 12 or 13. Missense p53 mutations are found in 50% to 60% of human colorectal carcinomas, but such mutations are seen in <10% of colorectal adenomas (2, 3). To address the potential role of p53 defects in tumors in CPC;Apc mice, we first studied p53 expression by immunohistochemistry. We found that 93 of the 97 tumors studied showed moderately elevated p53 expression in the nucleus of neoplastic cells relative to adjacent normal mucosa (Fig. 5B), with the remaining 4 tumors showing p53 expression equivalent to levels in normal mucosa. Consistent with the immunohistochemistry results suggesting that endogenous wild-type p53 expression might be elevated in tumors perhaps due to cellular stresses (e.g., hypoxia), sequence-based analysis of exons 5 to 8 of the Trp53 gene were carried out in 86 of the 93 tumors with moderate p53 expression, and no missense substitutions were seen. The only lesion with robust p53 nuclear staining was the large deeply invasive distal colon carcinoma found in a 20-month-old CPC;Apc mouse (Fig. 4D). This case harbored a p53 missense mutation, resulting in substitution of valine for alanine at mouse p53 codon 156 (corresponding to human codon 159).

A number of prior studies in the literature have established that whereas focal hypermethylation of certain CpG-rich sequence elements (e.g., the promoters of selected genes) can often be seen in colorectal tumors (reviewed in refs. 3, 31), global DNA hypomethylation is a common feature of human colorectal adenomas and carcinomas (3133). Immunohistochemistry studies with an anti–5-methylcytosine antibody in each of 17 tumors studied yielded data indicating that decreased global DNA methylation in neoplastic cells relative to adjacent normal epithelium was a common feature of CPC;Apc tumors (Fig. 5B).

Genomic instability has been suggested to play a critical role in the initiation and/or progression of the majority of colorectal tumors in man (34). To assess microsatellite instability in the CPC;Apc tumors, we studied three mononucleotide repeat and three dinucleotide repeat markers that have previously been shown to manifest instability in a subset of tissues and tumors from mice defective for mismatch repair (2022). Of 66 CPC;Apc tumors analyzed with these six markers, one tumor showed variant alleles at one mononucleotide repeat marker (data not shown). The results imply that the vast majority of tumors had a microsatellite stable phenotype. To investigate potential chromosomal instability of CPC;Apc tumors, we carried out spectral karyotyping analysis on one short-term primary culture line of epithelial cells from each of five independent large colorectal adenomas. Spectral karyotyping analysis and 4′,6-diamidino-2-phenylindole (DAPI) staining revealed that the primary cultured cells had aneuploidy, with 58.56 ± 17.68 chromosomes seen (Fig. 5C and data not shown), although no specific structural rearrangements were noted. The findings suggest that chromosomal instability has a contributing role in tumor progression in the CPC;Apc mouse model.

Data from clinical, pathologic, and molecular studies indicate that most human colorectal cancers arise from adenomatous precursors (14). However, only a fraction of adenomas progress to cancer. Adenomas >1 cm in size have a roughly 10% chance of progressing to carcinoma over a 10-year period (35). APC defects seem to have a preeminent role in the pathogenesis of colorectal adenomas and carcinomas in man: (a) biallelic inactivating mutations in the APC gene are found in ∼75% of sporadic adenomas and carcinomas, and (b) APC inactivation seems to play a critical initiating role in adenoma development in the inherited setting of FAP and in sporadic cases (24). Mouse strains carrying germline Apc mutations, particularly the ApcMin mouse, have been of keen interest for insights they can offer into intestinal tumor pathogenesis, for their potential value in defining environmental factors affecting tumor development, and for identifying novel agents and strategies for cancer prevention and treatment (8). Nonetheless, ApcMin mice have limitations for modeling human colorectal cancer, particularly because they manifest predominantly small intestinal lesions (7, 8, 13), whereas sporadic and inherited intestinal tumors in man arise overwhelmingly in the colon and rectum. In addition, perhaps due to their death from severe anemia by 140 days of age, few ApcMin mice develop carcinomas. Thus, mouse models that better recapitulate features of human colorectal tumorigenesis are needed. We have described a mouse model of colorectal adenoma-carcinoma progression in which a CDX2P-NLS Cre transgene is used to target Apc for somatic inactivation preferentially in the colon.

Whereas mice carrying CDX2P 9.5–regulated transgenes expressed the transgenes throughout the caudal region during early development, transgene expression was restricted to distal small intestine, cecum, and colon and rectum later in development and in adult tissues. Consistent with this expression pattern, the CPC;Apc mice reproducibly developed multiple colorectal tumors, with few small intestine lesions. The majority of lesions were adenomas, but carcinomas were seen in 17% of CPC;Apc mice during a 300-day follow-up period. In a 20-month-old CPC;Apc mouse, we found a large carcinoma in the distal colon. More carcinomas and perhaps even metastatic lesions might have been seen in an older cohort of CPC;Apc mice because the cohort of 36 CPC;Apc mice studied in depth here were followed only to 300 days of age and 36% of the CPC;Apc mice were alive at this time point.

Besides displaying morphologic features similar to human colorectal tumors, tumors arising in CPC;Apc mice had molecular features resembling those in human lesions. The vast majority of the mouse lesions had biallelic Apc inactivation and β-catenin dysregulation. We also found that CPC;Apc tumors showed global DNA hypomethylation relative to adjacent normal epithelium. Spectral karyotyping analysis of a short-term primary cell line from each of five independent adenomas revealed aneuploidy in all lines without evidence of specific structural rearrangements. We found no evidence of somatic p53 or KRAS mutations in the vast majority of tumors obtained from CPC;Apc mice. Based on studies of human colorectal tumors, our results are perhaps not unexpected. Only 10% of human adenomas <1 cm in size have KRAS mutations (3, 36), and only 3 of 380 lesions in our cohort of 36 mice were >1 cm. Additionally, because the overwhelming majority of the CPC;Apc tumors studied were adenomas, and p53 mutations are rare in human adenomas (24), it is not surprising that p53 mutations were not found in the vast majority of tumors studied. Of interest, the most advanced CPC;Apc carcinoma studied had acquired a p53 missense mutation. The virtual absence of microsatellite instability in the mouse CPC;Apc tumors studied is also perhaps not surprising when the situation in man is considered. High-frequency instability of microsatellite repeat tracts is rarely seen in sporadic adenomas or in adenomas arising in FAP patients (2, 4). Microsatellite instability is most common in carcinomas arising in the proximal colon (2), and the vast majority of the CPC;Apc mouse lesions were adenomas in the distal colon.

Our studies with the CPC;Apc model yielded some observations that will require further work to understand better the biological significance and underlying mechanisms. First, the distribution of tumors in the intestinal tract of the CPC;Apc mice was not uniform. On average, one tumor was found in the cecum and five to eight tumors were found in the colon and rectum, with most located distally. These findings on the nonuniform distribution of tumors in affected mice, at least relative to the pattern of Cdx2 9.5 promoter activity and Apc gene targeting in the mice, imply that additional somatic defects promoting the clonal outgrowth of epithelial cells with a single Apc-mutant allele may not occur in a stochastic fashion in the intestinal tract. Further studies are needed to determine whether the nonuniform distribution of tumors is largely due to regional differences in the frequency of wild-type Apc allele inactivation, regional differences in the frequency and types of additional somatic events required for progressive adenoma growth, or perhaps both possibilities.

Second, male and female CPC;Apc mice had equivalent numbers of small intestinal and cecal tumors, but male mice had a 50% increase in distal colon and rectal tumors relative to females. This tendency toward a 50% increase in distal tumors was also seen in male Villin-Cre;Apc mice, but the number of Villin-Cre;Apc mice studied was limited. In humans, males have been reported to have a larger number of colon adenomas of clinically significant size (i.e., >9 mm) in the distal half of the colon, compared with females (37). Data also indicate that hormone replacement therapy in women conferred protection against colorectal adenomas (38), and a recent study reported that estrogen receptors α and β are inhibitory modifiers of Apc-dependent tumorigenesis in the proximal colon of ApcMin mice (39).

Third, our data suggest that small intestinal tumor burden may inhibit via an unknown mechanism the development and/or progression of colorectal tumors in the mouse. We reported here that CPC;Apc mice developed, on average, three small intestine tumors and seven to eight colorectal and cecal tumors, whereas Villin-Cre;Apc mice of similar age manifest ∼31 small intestine tumors and five cecal and colorectal tumors, with the colorectal tumors in the Villin-Cre;Apc mice smaller on average than those in CPC;Apc mice. Perhaps even more remarkably, although ApcMin mice carry one inactive Apc allele in all somatic cells, a typical ApcMin mouse in our colony develops 50 or so small intestinal adenomas and only one or no colon adenomas (data not shown). The early death of ApcMin mice by 120 to 140 days of age is not sufficient to account for the difference in colon tumor multiplicity because readily detectable distal adenomas can be found in CPC;Apc mice at 120 days. Insights into mechanisms by which the presence of a large number of small intestine adenomas exerts apparent inhibitory effects on colorectal tumor growth in mice may be of interest for the situation in man.

In closing, whereas we have shown here the utility of the CDX2P-NLS Cre transgene for targeting the ROSA26 reporter and the Apc gene, the CDX2P-NLS Cre transgene should prove valuable for studying other genes in an in vivo model of colon tumorigenesis, including tumor suppressor genes that can be conditionally inactivated or proto-oncogenes that can be conditionally activated by loxP-mediated deletions. Furthermore, assuming that biallelic conditional inactivation of a given gene does not result in a developmental phenotype due to inactivation in the caudal region during embryogenesis, the CDX2P-NLS Cre transgene may also be useful for studying other genes in cell fate specification and the physiology of colonic epithelium.

Note: T. Hinoi and A. Akyol contributed equally to this work.

Grant support: National Cancer Institute grants CA082223 and CA085463 (E.R. Fearon) and grant CA46592 to the University of Michigan Comprehensive Cancer Center Core facilities.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. T. Saunders and the University of Michigan Transgenic Core for outstanding support, and Drs. T. Noda, D. Gumucio, and S. Camper for providing assistance with the mouse studies.

1
Potter JD. Colorectal cancer: molecules and populations.
J Natl Cancer Inst
1999
;
91
:
916
–32.
2
Arnold CN, Goel A, Blum HE, Boland CR. Molecular pathogenesis of colorectal cancer, implications for molecular diagnosis.
Cancer
2005
;
104
:
2035
–47.
3
Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis.
Cell
1990
;
61
:
759
–67.
4
Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer.
Cell
1996
;
87
:
159
–70.
5
Su LK, Kinzler KW, Vogelstein B, et al. Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene.
Science
1992
;
256
:
668
–70.
6
Clevers H. Wnt/β-catenin signaling in development and disease.
Cell
2006
;
127
:
469
–80.
7
Moser AR, Pitot HC, Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse.
Science
1990
;
247
:
322
–4.
8
Shoemaker AR, Gould KA, Luongo C, Moser AR, Dove WF. Studies of neoplasia in the Min mouse.
Biochim Biophys Acta
1997
;
1332
:
F25
–48.
9
Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007.
CA Cancer J Clin
2007
;
56
:
43
–66.
10
Fodde R, Edelmann W, Yang K, et al. A targeted chain-termination mutation in the mouse Apc gene results in multiple intestinal tumors.
Proc Natl Acad Sci USA
1994
;
91
:
8969
–73.
11
Oshima M, Oshima H, Kitagawa K, Kobayashi M, Itakura C, Taketo M. Loss of Apc heterozygosity and abnormal tissue building in nascent intestinal polyps in mice carrying a truncated Apc gene.
Proc Natl Acad Sci USA
1995
;
92
:
4482
–6.
12
Smits R, Kartheuser A, Jagmohan-Changur S, et al. Loss of Apc and the entire chromosome 18 but absence of mutations at the Ras and Tp53 genes in intestinal tumors from Apc1638N, a mouse model for Apc-driven carcinogenesis.
Carcinogenesis
1997
;
18
:
321
–7.
13
Boivin GP, Washington K, Yang K, et al. Pathology of mouse models of intestinal cancer: consensus report and recommendations.
Gastroenterology
2003
;
124
:
762
–77.
14
Colnot S, Niwa-Kawakita M, Hamard G, et al. Colorectal cancers in a new mouse model of familial adenomatous polyposis: influence of genetic and environmental modifiers.
Lab Invest
2004
;
84
:
1619
–30.
15
Shibata H, Toyama K, Shioya H, et al. Rapid colorectal adenoma formation initiated by conditional targeting of the Apc gene.
Science
1997
;
278
:
120
–3.
16
Hinoi T, Loda M, Fearon ER. Silencing of CDX2 expression in colon cancer via a dominant repression pathway.
J Biol Chem
2003
;
278
:
44608
–16.
17
Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain.
Nat Genet
1999
;
21
:
70
–1.
18
Madison BB, Dunbar L, Qiao XT, Braunstein K, Braunstein E, Gumucio DL. cis elements of the villin gene control expression in restricted domains of the vertical (crypt) and horizontal (duodenum, cecum) axes of the intestine.
J Biol Chem
2002
;
277
:
33275
–83.
19
Hinoi T, Lucas PC, Kuick R, Hanash S, Cho KR, Fearon ER. CDX2 regulates liver intestine-cadherin expression in normal and malignant colon epithelium and intestinal metaplasia.
Gastroenterology
2002
;
123
:
1565
–77.
20
Kabbarah O, Mallon MA, Pfeifer JD, Edelmann W, Kucherlapati R, Goodfellow PJ. A panel of repeat markers for detection of microsatellite instability in murine tumors.
Mol Carcinog
2003
;
38
:
155
–9.
21
Yang G, Scherer SJ, Shell SS, et al. Dominant effects of an Msh6 missense mutation on DNA repair and cancer susceptibility.
Cancer Cell
2004
;
6
:
139
–50.
22
Bacher JW, Abdel Megid WM, Kent-First MG, Halberg RB. Use of mononucleotide repeat markers for detection of microsatellite instability in mouse tumors.
Mol Carcinog
2005
;
44
:
285
–92.
23
Paraskeva C, Williams A. The colon, culture of epithelial cells. 1st ed. In: Freshney RI, editor. New York (NY): John Wiley & Sons, Inc.; 1992.
24
Macartney KK, Baumgart DC, Carding SR, Brubaker JO, Offit PA. Primary murine small intestinal epithelial cells, maintained in long-term culture, are susceptible to rotavirus infection.
J Virol
2000
;
74
:
5597
–603.
25
Bezrookove V, Smits R, Moeslein G, et al. Premature chromosome condensation revisited: premature chromosome condensation revisited: a novel chemical approach permits efficient cytogenetic analysis of cancers.
Genes Chromosomes Cancer
2003
;
38
:
177
–86.
26
Beck F, Erler T, Russell A, James R. Expression of Cdx-2 in the mouse embryo and placenta: possible role in patterning of the extra-embryonic membranes.
Dev Dyn
1995
;
204
:
219
–27.
27
Moser AR, Mattes EM, Dove WF, Lindstrom MJ, Haag JD, Gould MN. ApcMin, a mutation in the murine Apc gene, predisposes to mammary carcinomas and focal alveolar hyperplasias.
Proc Natl Acad Sci USA
1993
;
90
:
8977
–81.
28
Luongo C, Moser AR, Gledhill S, Dove WF. Loss of Apc+ in intestinal adenomas from Min mice.
Cancer Res
1994
;
54
:
5947
–52.
29
da Costa LT, He TC, Yu J, et al. CDX2 is mutated in a colorectal cancer with normal APC/β-catenin signaling.
Oncogene
1999
;
18
:
5010
–4.
30
Aoki K, Tamai Y, Horiike S, Oshima M, Taketo MM. Colonic polyposis caused by mTOR-mediated chromosomal instability in Apc+/Δ716 Cdx2+/− compound mutant mice.
Nat Genet
2003
;
35
:
323
–30.
31
Esteller M. Aberrant DNA methylation as a cancer-inducing mechanism.
Annu Rev Pharmacol Toxicol
2005
;
45
:
629
–56.
32
Feinberg AP, Vogelstein B. Hypomethylation distinguishes genes of some human cancers from their normal counterparts.
Nature
1983
;
301
:
89
–92.
33
Feinberg AP, Gehrke CW, Kuo KC, Ehrlich M. Reduced genomic 5-methylcytosine content in human colonic neoplasia.
Cancer Res
1988
;
48
:
1159
–61.
34
Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers.
Nature
1998
;
396
:
643
–9.
35
Stryker SJ, Wolff BG, Culp CR, Libbe SD, Ilstrup DM, MacCarty RL. Natural history of untreated colonic polyps.
Gastroenterology
1987
;
93
:
1009
–13.
36
Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations during colorectal tumor development.
N Engl J Med
1988
;
319
:
525
–32.
37
Lieberman DA, Holub J, Eisen G, Kraemer D, Morris CD. Prevalence of polyps greater than 9 mm in a consortium of diverse clinical practice settings in the United States.
Clin Gastroenterol Hepatol
2005
;
3
:
798
–805.
38
Chlebowski RT, Wactawski-Wende J, Ritenbaugh C, et al. Estrogen plus progestin and colorectal cancer in postmenopausal women.
N Engl J Med
2004
;
350
:
991
–1004.
39
Cho NL, Javid SH, Carothers AM, Redston M, Bertagnolli MM. Estrogen receptors α and β are inhibitory modifiers of Apc-dependent tumorigenesis in the proximal colon of Min/+ mice.
Cancer Res
2007
;
67
:
2366
–72.