|Home | About | Journals | Submit | Contact Us | Français|
The epithelial linings of the small and large intestine are rapidly turned over and provide an ideal system for exploring links between differentiation and regulation of cell cycle exit. We utilized wild type, p21−/−, p27−/− and p21/p27−/− mice to address contributions of the Cdk inhibitors p21 and p27 to proliferation and differentiation in the mouse gastrointestinal tract. We did not detect any significant differences in proliferation, and all differentiated epithelial cell lineages were represented in all four genotypes. These data indicate that p21 and p27 do not play essential roles in the regulation of normal epithelial renewal in the intestine. These Cdk inhibitors are not needed in vivo for either assembly of Cdk/Cyclin complexes that drive active proliferation, or inhibition of Cdk/Cyclin complexes during cell cycle exit. However, expression of Cyclin D2 and to a lesser degree Cyclin D3 was reduced in p27−/− and p21/p27−/− mice, indicating a unique role for p27 in the regulation of these specific D-type Cyclins in vivo. In the absence of p27, reduced levels of Cyclin D2 and D3 may help to counteract increased proproliferative signals in the intestine.
The intestinal epithelium represents a unique model system for studying cell proliferation and differentiation, because it undergoes rapid and continuous renewal throughout life (reviewed in 1). In the small intestine, proliferation is restricted to the crypts that contain anchored stem cells that give rise to rapidly proliferating progenitor cells that produce four different epithelial cell lineages. Three types of epithelial cells (absorptive enterocytes, mucus-producing goblet cells, and peptide hormone secreting enteroendocrine cells) differentiate during an upward migration from the crypt to an adjacent villus. The fourth epithelial lineage, Paneth cells, differentiate as they migrate downwards to the base of the crypt.
Although the small intestine is frequently analyzed in mouse models of intestinal cancer, most human intestinal cancers arise in the colon. In contrast to the small intestine, the colon lacks Paneth cells and villi. Stem cells are located at the base of the crypts in the colon and give rise to colonocytes, goblet cells and endocrine cells that differentiate as they migrate upwards and out of the crypt 2.
Rapid proliferation coupled with continuous differentiation makes the intestine an ideal model for addressing contributions of cell cycle regulators to differentiation. G1 progression is regulated by the D-type Cyclins (D1, D2, D3) that associate with either Cdk4 or Cdk6, and Cdk2 that associates with Cyclin E and A. The Cip/Kip family of Cyclin dependent kinase (Cdk) inhibitors, including p21 (Cip) and p27 (Kip), bind a variety of Cdk/Cyclin complexes with different outcomes 3. Although initially isolated as Cdk inhibitors, p21 and p27 were detected in active Cyclin D/Cdk4 complexes 4–7. p21 was shown to stabilize interactions between Cdk4 and Cyclin D and promote the formation of active complexes 7, 8. In contrast, p21 and p27 association with Cdk2 led to Cdk inhibition, although genetic evidence indicates that Cdk2 is dispensable for p21 and p27 mediated inhibition of cell cycle progression 9. In addition to inhibiting Cdk2, p27 also regulates Cdk1 activities 10. In the last few years novel functions for p21 and p27 in regulating cytoskeletal dynamics and migration have also been proposed 11.
Both p21 and p27 have been implicated as regulators of intestinal epithelial cell differentiation. Expression of p21 is induced as epithelial cells exit the cell cycle and begin the process of terminal differentiation in the small and large intestine 12–15. Similar to p21, p27 expression has been localized to differentiated epithelial cells in the small intestine 14, 16, 17 and colon 18. Although initial reports did not support a role for p21 in regulating differentiation in the gut 19, subsequent studies showed that expression of p21 was induced following expression of dominant negative TCF-4 in colon cancer cells, and that ectopic expression of p21 in the Ls174T colon cancer cell line promoted epithelial cell differentiation associated gene expression 15. A variety of studies, often using human colon cancer cell lines, have suggested that p27 plays a direct role in promoting intestinal epithelial cell differentiation 16, 20–23.
We used wild type and knockout mice to examine contributions of p21 and p27 to intestinal epithelial cell differentiation in vivo. Although disruption of neither p21 nor p27, nor both genes led to significant differences in differentiation or proliferation, disruption of p27 led to downregulation of Cyclin D2 and Cyclin D3 in the colon unveiling a critical role for p27 in regulating these Cyclins in the gastrointestinal tract.
Generation of p21−/− (Cdkn1atm1Tyj) 19, p27−/− (Cdkn1btm1Kiyo; Δ51) 24 and p21−/−; p27 −/− double knockout mice 25 mice was previously described. The genotypes of the mice were confirmed by genomic DNA isolation and PCR. Mice were fed a commercial diet and water ad libitum, and sacrificed at 8–10 weeks of age. When possible littermates were used in separate experiments and results from multiple litters were examined. In the different experiments each lane represents samples from a separate individual mouse (age and sex matched). All studies were approved by the University of Illinois at Chicago Institutional Review Board Animal Care Committee.
Mice were injected with 5-bromo-2-deoxyuridine (BrdU, Sigma-Aldrich Corp. St. Louis, MO) in PBS at 50µg/g bodyweight two hours prior to sacrificing them and tissues were fixed in Carnoy’s (10% glacial acetic acid, 60% ethanol, 30% chloroform) for three hours at 4°C. BrdU incorporation was detected using immunohistochemistry and the M.O.M. Immunodetection Kit (Vector Laboratories Inc., Burlingame, CA). For other immunohistochemical analyses, tissues were fixed in 4% paraformaldehyde and the Vectastain ABC Kit (Vector Laboratories Inc.) was used according to manufacturer’s instructions. BrdU labeling indices were determined as previously described 26. Positive cells in at least ten crypts in well-oriented longitudinal sections were counted in the colons of each animal. In the small intestine, positive cells from 10 well-oriented crypt-villus units from at least 3 quadrants were counted for each animal and used to determine the average of each animal. Differences between means were compared by Student’s two-tailed t-test, and the level of statistical significance was set at P < 0.05.
Rabbit anti-lysozyme, anti-synaptophysin (DAKO Corp., Carpinteria, CA), and rabbit anti-I-FABP 27(a gift from Dr. J. I. Gordon, Washington University St. Louis) were used to examine differentiation as previously described 28. Expression of p27 was examined using a Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), C-19 rabbit polyclonal antibody. Primary antibody was incubated with tissue sections overnight at 4°C at a dilution of 1:2000 followed by tyramide signal amplification using a TSA Biotin Kit (PerkinElmer Life & Analytical Sciences, Shelton, CT). Antibody binding was visualized by incubation of the slides with the chromogenic substrate 3,3’-diaminobenzidine tetrahydrochloride (DAB, Sigma-Aldrich, St. Louis, MO). Controls were performed with equal dilutions of rabbit or mouse IgG (Santa Cruz Biotechnology, Inc.).
RNA was prepared using Trizol Reagent (Invitrogen Corporation, Carlsbad, CA), and assays were performed as described 29. Antisense probes specific for mouse sucrase isomaltase, TFF3, cryptdin-1 and cyclophilin were a gift from Dr. Debra Silberg 30.
In general, 50 µg of total intestinal cell lysate was separated on 12 or 15% SDS-polyacrylamide gels and transferred to PVDF membranes (Millipore, Billerica, MA). All membranes were stained with Ponceau S to confirm equal loading and transfer of protein. Monoclonal antibodies against Cyclin D1 and D2, and rabbit polyclonal anti-Cyclin D3 were purchased from Lab Vision Corporation, Fremont, CA. Antibody specific for p21, (Mouse IgG1, clone SXM30) was obtained from (BD Biosciences, San Jose, CA). Other antibodies used include anti-: p27 (c-19, SC-528), p57 (H-91; SC-8298), Cdk2 (SC-163), Cdk4 (SC-260), PCNA (SC-56) from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, and β-actin (A-5441, Sigma-Aldrich, St. Louis, MO). Peroxidase reactions were visualized with SuperSignal West Dura Extended Duration Substrate for chemiluminescence (Pierce Biotechnology, Rockford, IL), and membranes were exposed to X-ray film.
Expression of p21 and p27 were detected in lysates prepared from small and large intestine (Figure 1A and C). However expression of the related Cdk-inhibitor p57 was not detected by immunoblotting using the same tissue lysates. Since p57 is expressed throughout embryonic development, protein isolated from a 14.5 day mouse embryo was included as a positive control. Although a recent study reported localization of p27 to intestinal crypt cells 31, our immmunohistochemical localization of p27 in the small intestines of wild type mice revealed that nuclear p27 was present primarily in differentiated epithelial cells (Figure 1B). These data are in agreement with previous studies 14, 16, 17. No p21 or functional full-length p27 was detected in lysates prepared from mice bearing disruptions in these genes. In p27 null mice a truncated form of p27 that is not able to bind Cdks is detected (asterisk, 1A, C). Interestingly, increased basal levels of p21 were detected in colons of p27 null mice, as previously reported for the liver and other tissues 25.
The Cdk inhibitors p21 and p27 have been implicated in promoting Cdk/Cyclin assembly and inhibition. To examine the impact of disrupting p21 and p27 expression on proliferation in the small and large intestine, animals were injected with BrdU two hours prior to sacrifice. BrdU incorporation was assessed in at least three different age and sex matched mice per genotype. Labeled cells were counted in well-orientated crypts of similar size/depth in at least three mice per each region of intestine. No significant differences were found in the number of epithelial cells replicating DNA in the small intestine (Figure 2) or colon (Figure 3). All proliferating epithelial cells were localized to the crypt region. Rare BrdU positive cells that appear to be lymphocytes were also detected interspersed within the epithelium or in the lamina propria, the loose connective tissue at the core of each villus that contains fibroblasts, lymphocytes and other hematotpoietic cells, in double knockout mice (Figure 2B).
Since several studies cited above have implicated p21 and p27 in regulating differentiation of intestinal epithelial cells, we examined the differentiation of various cell types in the small intestine, using a variety of different techniques. RNase protection assays did not reveal any significant differences in the expression of genes encoding the following lineage specific markers: sucrase isomaltase (absorptive enterocytes) 32, intestinal trefoil factor TFF3 (goblet cells) 33 or Cryptdin 1 (Paneth cells)34 (Figure 4).
A report that that suggested that differentiation of absorptive cells was impaired in p27 null mice 21 led us to more closely examine differentiation of absorptive cells in all of the genotypes. As a marker for this cell type we examined expression of intestinal fatty acid binding protein (I-FABP), a small intestinal-specific fatty acid binding protein that represents approximately 3% of the cytoplasmic protein of enterocytes 35, 36. Immunohistochemistry and immunoblotting experiments (Figure 5) did not reveal diminished levels of iFABP expression in p27 null mice or any of the genotypes examined. A representative immunoblotting experiment is shown in Figure 5B. Variations in iFABP are greatest between littermates of the same genotypes in this particular experiment, perhaps due to differences in nutritional intake since the mice were allowed free access to food and water. This experiment was repeated with several groups of animals, with similar outcomes. In no instance did absorptive cell differentiation appear impaired in the p27-deficient animals.
Differentiation of mucus secreting goblet cells was examined using Periodic acid-Schiff (PAS) staining that is specific for carbohydrates. PAS positive goblet cells stain pink. PAS positive cells were counted on multiple villi along a fixed length (400 µm) in at least three mice per genotype. We did not detect reduced PAS staining in p27 null mice as recently reported 37. A modest, but significant increase in goblet cell number was detected in mice lacking both p21 and p27 (Figure 6B). However this increase was not sufficient to have an impact on overall TFF3 expression levels (Figure 4). Enteroendocrine cell differentiation was examined with antibodies specific for synaptophysin, a marker for a subset of intestinal enteroendocrine cells 38. Similar numbers of synaptophysin positive cells were present in all genotypes examined (Figure 6C).
Induction of p21 expression in the Ls174T human colon carcinoma cell line was found to induce expression of several differentiation markers including Ephrin B1 15, which may have an impact on Paneth cell differentiation or positioning 39, 40. No differences in expression of Paneth cell specific Cryptdin-1 mRNA were detected by RNase protection assays (Figure 4). In addition, further examination of Paneth cell differentiation examining expression of lysozyme, another Paneth cell marker, did not reveal significant differences between the genotypes. Lysozyme positive crypt cells were generally located at the crypt base (Figure 7A) in all genotypes, although in extremely rare instances a lysozyme positive cell could be found in upper regions of the crypts of p21−/− mice (Figure 7B). Comparable levels of lysozyme were detected in all four genotypes (Figure 7C).
We examined expression of a variety of cell cycle regulated genes in intestines of wild type, p21−/−, p27−/− and double knockout mice. Expression profiles in the colons of multiple age-matched male or female mice was compared and are shown in Figure 8A and B. The primary difference in protein expression detected between the different genotypes in the colon was a striking reduction in Cyclin D2 expression, and to a lesser degree reduced Cyclin D3 expression in the absence of p27 expression.
Epithelial cells in the small and large intestine exit the cell cycle as they migrate and terminally differentiate. Expression of the Cdk inhibitors p21 and p27 in differentiated intestinal epithelial cells raised the possibilities that these proteins are regulators of either cell cycle exit and/or differentiation itself. Here we focused on the examining functions of p21 and p27 during normal intestinal homeostasis and found that regulated proliferation and differentiation is maintained even in the absence of both Cyclin kinase inhibitors. Neither p21 nor p27 is required for Cdk/Cyclin inhibition or assembly in the regenerating epithelial linings of the intestine.
Other studies with p21/p27 double null mice have reported perturbations in cessation of proliferation, but not differentiation in different mouse tissues. Disruption of p21 and p27 was found to perturb hepatocyte quiescence in adult mice 25. Disruption of both p27 and p21 synergistically led to an increase in proliferative life span of ovarian granulosa cells in vivo and in vitro but did not have a major impact on differentiation 41. We did not detect significant increases in proliferation in the intestines of p21/p27 double knockout mice, suggesting that compensatory mechanisms must exist to control proliferation in the small intestine and colon under normal growth conditions. We were unable to detect expression of p57 suggesting that this third member of the Cip/Kip family of Cdk inhibitors does not play a major role in the cessation of proliferation that accompanies intestinal epithelial cell differentiation in these mice. Like our results, an early in situ examination of p57 gene expression in the adult mouse did not reveal expression of this Cdk inhibitor in the intestinal epithelium 42.
One of the first characterizations of p21/p27 double null mouse embryonic fibroblasts suggested that p21 and p27 might play critical roles in Cdk/Cyclin assembly required to promote proliferation 8. However a later study, also using MEFs, indicated that p21 and p27 were not required for Cdk/Cyclin assembly but played a role in regulating stability of Cyclins and thereby the abundance of Cdk/Cyclin complexes 43. Our in vivo data also support a role for p27 in regulating stability of Cyclins in the intestine, and we propose that the decrease in Cyclin D2 and Cyclin D3 levels detected in intestines deficient for p27 may compensate for any increase in active Cdk/Cyclin complexes. We previously demonstrated that Cyclin D2 is expressed in crypt epithelial cells in the colon, where it is positioned to contribute to epithelial cell proliferation, while Cyclin D3 was found primarily in mesenchymal cells 44. A role for p27 in regulating Cyclin D2 and Cyclin D3 levels in other tissues was previously reported 45, and it was found that decreases in Cyclin D levels in p27-deficient mice correlated with the amount of Cyclin D associated with p27 in wild type tissues. Association of p27 with Cyclins D2 and D3 appears to be responsible for maintaining their levels in intestinal cells.
Although a modest increase in goblet cell number was observed in animals deficient for both p21 and p27, all intestinal epithelial lineages are represented in mice lacking both p21 and p27, indicating that these Cdk inhibitors are not required for differentiation of the distinct lineages. Goblet cell hyperplasia has been associated with immune responses in the intestine 46, 47. The presence of BrdU positive intraepithelial and lamina propria cells in double knockout mice (Figure 2B) suggests increased levels of cycling lymphocytes may be present. Both p21 and p27 have been implicated in the regulation of hematopoietic stem cells 48, 49, and p27 null mice have been reported to have increased susceptibility to inflammation 50. Mucosal T cell cycling has been shown to be negatively regulated by p53, which is a major transcriptional activator of the p21 gene 51.
While p21 and p27 appear to have limited roles in the regulation of normal epithelial cell turnover in the intestine, several reports suggest that they have critical functions after stress and after injury. After resection, adaptation is impaired in p21-deficient mice 17. Mice lacking p21 are more sensitive to the colon carcinogen azoxymethane 52, and enhanced tumor development was observed when p21 was disrupted in different mouse models of intestinal tumorigenesis 53, 54. Disruption of the p21 gene also enhanced intestinal tumor metastases in mice 55. Although we did not detect any tumors in the young adult mice used in our experiments, increased numbers of tumors in the intestine have been detected in older p27 deficient mice, and tumorigenesis was enhanced when the animals were maintained on a high fat "western style" diet 56, 57.
Our data suggest that p21 and p27 do not play significant roles in regulating normal intestinal homeostasis. In the studies reported here, only age and sex-matched littermates from multiple litters maintained in a specific antigen free barrier facility were examined to reduce any variability due to sex, age, or strain variation. Since live cell isolation procedures may also influence expression and activities of labile proteins and have a direct impact on cell survival and proliferation, tissues were rapidly removed and homogenized when preparing protein and RNA. Our studies indicate that although p21 and p27 may play critical roles in regulating proliferative responses after injury and stress, redundant mechanisms exist to maintain homeostasis in the rapidly regenerating epithelium of the small and large intestine. Remarkably, disruption of the genes encoding these two key members of the Cip/Kip family has little impact on normal proliferation or differentiation in the mouse intestine.
NIH grants DK44525 and DK56283 (A. L. T.) supported these studies. We are indebted to Dr. Andrei Gartel for helpful discussions, Dr. Jeffrey Gordon for providing the iFABP antibody, Dr. Debra Silberg for sharing RNase protection probes, and Dr. Jennifer Schmidt for providing E14.5 mouse embryos.
Conflict-of-interest and financial disclosure statements: NONE