p27 in Intestinal Tumorigenesis and Chemoprevention in the Mouse

Targeted inactivation of p27 was sufficient for intestinal tumor formation in mice, but this was strictly a function of diet: tumors formed in p27 or p27 / mice fed control AIN76A diet and were increased by a western-style diet but did not develop in mice fed standard chow diet. When crossed with the Apc1638N mouse, Apc ,p27 or Apc ,p27 / mice not only formed twice as many tumors than the sum of the tumors from mutation at either locus alone, but on AIN76A diet also developed intestinal intussusception, a tumorassociated pathology in patients leading to intestinal blockage that has not been reported for intestinal cancer in mouse models. Moreover, the frequency of intussusception was increased when the compound mutant mice were maintained on the western diet, leading to early death. Despite this more aggressive tumor phenotype generated by inactivation of p27 than by inactivation of another cyclin-dependent kinase inhibitor, p21 , the nonsteroidal anti-inflammatory drug sulindac was still effective in inhibiting intestinal tumor formation in Apc ,p27 or Apc ,p27 / mice, which contrasts with the abrogation of the effects of sulindac in Apc ,p21 or Apc ,p21 / mice, indicating that p27 is not necessary for tumor inhibition by sulindac. Furthermore, tumor inhibition by sulindac was linked to the induction of p21 expression by the drug, regardless of p27 status, leading to suppression of cell proliferation and promotion of cell differentiation and apoptosis in the intestinal mucosa. (Cancer Res 2005; 65(20): 9363-8) Introduction p27 and p21 are both cyclin-dependent kinase (cdk) inhibitors (1, 2), but they do not have identical functions in regulating multifactor complexes that regulate cell maturation. For example, p27 has been shown to have a significant role in colonic cell differentiation outside of its role in regulating cell proliferation, with forced increases and decreases in expression linked to stimulation and inhibition, respectively, of differentiation along the absorptive cell lineage (3). In addition, there is much better evidence in the literature for level of expression of p27 as a marker of prognosis in human colon cancer than there is for p21 (3–9). Therefore, it was important that we found that, unlike targeted inactivation of p21, targeting of p27 was sufficient for tumor formation, not only in the pituitary, as had been reported (10–12), but also for intestinal adenomas and adenocarcinoma (13). This intestinal tumorigenesis by inactivation of p27 was linked to disruption of intestinal cell maturation caused by elevated expression of c-myc and cyclin D1, as well as the c-myc target gene, cdk4 (13). Although it was also reported that the targeted inactivation of p27 enhanced adenomatosis polyposis coli (Apc)–initiated tumor formation in compound mutant mice, that report did not address tumor formation by inactivation of p27 in the absence of an Apc mutation (14). Here, we show that this is likely because p27-initiated tumor formation is strictly dependent on the diet fed the animals. Moreover, we show that compound Apc ,p27 or Apc ,p27 / mutant mice develop intussusception, an unusual, aggressive tumor-associated pathology seen in patients that has not been previously reported in genetically initiated intestinal cancer in mice. Not only are the frequency and size of the tumors substantially increased in the Apc,p27 compound model by feeding of a western-style, high-risk diet but also the incidence of intussusception, leading to a severe shortening of life span. Finally, the nonsteroidal anti-inflammatory drug sulindac has a substantial effect in inhibiting Apc-initiated tumors in the intestine of either the human (15, 16) or the mouse (17, 18). We show that the inactivation of p27 does not interfere with this tumor inhibition by sulindac, which contrasts markedly with the abrogation of the effects of sulindac on tumor formation in Apc mice that have a targeted inactivation of the gene encoding another cdk inhibitor, p21 (17, 19). This inhibition of tumorigenesis in the Apc,p27 compound mice was associated with the ability of sulindac to induce p21, which led to reduced proliferation and apoptosis in both the intestinal mucosa and tumors of Apc mice in which one or both p27 alleles had been inactivated. All of these responses were lost in the Apc ,p21 and Apc ,p21 / mice. Materials and Methods The Apc1638N and p27 / mouse models and methods for genotyping have been reported (13, 20). Apc1638N mice (C57Bl/6 background) were mated with p27 / mice (mixture of 129S1/sv C57Bl/6) to generate Apc ,p27 offspring (F1). F1 mice were mated to produce desired genotypes: Apc ,p27, Apc ,p27 , or Apc ,p27 / . At weaning (f3-4 weeks), littermates of different genotypes were randomized to different dietary groups and fed, ad libitum , either AIN76A control–defined diet, AIN-76A diet supplemented with 0.02% sulindac, or a western-style diet based on AIN-76A that is formulated on the principle of nutrient density to mimic the intake of major risk factors for colon cancer in the diet of populations in developed countries: high in fat and phosphate and low in calcium and vitamin D (21, 22). Diets were from Research Diets (New Brunswick, NJ). The mice with targeted inactivation of p27 without an Apc mutation were also maintained on standard chow diet (LabDiet, Somerville, NJ), AIN-76A diet, or western-style diet. Mice were weighed weekly and maintained on diet for 16 or 36 weeks, or until they exhibited significant weight loss or other signs of extensive tumor formation. Mice were killed by CO2 overdose and rapidly dissected for evaluation of tumors and fixation of tissues, as described previously (13, 17, 23). Total RNA and protein were isolated from the frozen tissues using TRIzol reagent (Invitrogen Life Technology, Carlsbad, CA), as previously described Requests for reprints:WanCai Yang, Department of Oncology, Montefiore Medical Center, Albert Einstein Cancer Center, 111 East 210th Street, Bronx, NY 10467. Phone: 718-920-7620; Fax: 718-882-4464; E-mail: [email protected]. I2005 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-05-2113 www.aacrjournals.org 9363 Cancer Res 2005; 65: (20). October 15, 2005 Research Article Research. on June 4, 2015. © 2005 American Association for Cancer cancerres.aacrjournals.org Downloaded from (13, 19). The quantity of RNA and protein were measured spectrophotometrically, as we described previously (13, 19). cDNA was synthesized from DNase-treated total RNA using Taqman Multiscribe Reverse Transcriptase (Applied Biosystems, Inc., Foster City, CA). Quantitative PCR analysis was done using the ABI Prism 7900-HT Sequence Detection System (96 wells). The primers for p21, p27, and b-actin; the amplification conditions for the quantitative real-time reverse transcription-PCR; and methods of data analysis have been reported (13, 19). Western blot analysis of specific proteins was done by standard methods (13, 19) using the following primary antibodies for detection: anti-p21, antip27 (Santa Cruz Biotechnology, Santa Cruz, CA), and anti-h-actin (Sigma, St. Louis, MO). Signal was detected by the enhanced chemiluminescence technique (Amersham Life Science, Piscataway, NJ).