Mol Cancer Ther. 2006;5:2060-2069
© 2006 American Association for Cancer Research
Research Articles: Therapeutics
Chemoprevention of colon carcinogenesis by polyethylene glycol: suppression of epithelial proliferation via modulation of SNAIL/ß-catenin signaling
Hemant K. Roy1,
Dhananjay P. Kunte1,
Jennifer L. Koetsier1,
John Hart3,
Young L. Kim2,
Yang Liu2,
Marc Bissonnette4,
Michael Goldberg1,
Vadim Backman2 and
Ramesh K. Wali1
1 Department of Internal Medicine, Evanston-Northwestern Healthcare; 2 Biomedical Engineering Department, Northwestern University, Evanston, Illinois; and Departments of 3 Pathology and 4 Internal Medicine, University of Chicago Medical Center, Chicago, Illinois
Requests for reprints: Ramesh K. Wali, Feinberg School of Medicine, Evanston Northwestern Healthcare, 2650 Ridge Avenue, Evanston, IL 60201. Phone: 847-570-4108; Fax: 847-733-5451. E-mail: rwali{at}enh.org
 |
Abstract
|
|---|
Polyethylene glycol (PEG) is one of the most potent chemopreventive agents against colorectal cancer; however, the mechanisms remain largely unexplored. In this study, we assessed the ability of PEG to target cyclin D1–ß-catenin–mediated hyperproliferation in the azoxymethane-treated rat model and the human colorectal cancer cell line, HT-29. Azoxymethane-treated rats were randomized to AIN-76A diet alone or supplemented with 5% PEG-8000. After 30 weeks, animals were euthanized and biopsies of aberrant crypt foci and uninvolved crypts were subjected to immunohistochemical and immunoblot analyses. PEG markedly suppressed both early and late markers of azoxymethane-induced colon carcinogenesis (fractal dimension by 80%, aberrant crypt foci by 64%, and tumors by 74%). In both azoxymethane-treated rats and HT-29 cells treated with 5% PEG-3350 for 24 hours, PEG decreased proliferation (45% and 52%, respectively) and cyclin D1 (78% and 56%, respectively). Because ß-catenin is the major regulator of cyclin D1 in colorectal cancer, we used the T-cell factor (Tcf)–TOPFLASH reporter assay to show that PEG markedly inhibited ß-catenin transcriptional activity. PEG did not alter total ß-catenin expression but rather its nuclear localization, leading us to assess E-cadherin expression (a major determinant of ß-catenin subcellular localization), which was increased by 73% and 71% in the azoxymethane-rat and HT-29 cells, respectively. We therefore investigated the effect of PEG treatment on levels of the negative regulator of E-cadherin, SNAIL, and observed a 50% and 75% decrease, respectively. In conclusion, we show, for the first time, a molecular mechanism through which PEG imparts its antiproliferative and hence profound chemopreventive effect. [Mol Cancer Ther 2006;5(8):2060–9]
|
Introduction
|
|---|
Colorectal cancer is the second leading cause of cancer deaths in the United States (1). Because therapy for advanced disease is generally poor, there have been continuing efforts to develop effective chemopreventive strategies to reduce the morbidity and mortality. There have been a myriad of agents that have been shown to protect against colorectal cancer in preclinical and epidemiologic studies (2, 3); however, suboptimal efficacy and potential toxicity have limited their widespread clinical exploitation (4). For instance, animal models and clinical trials (case-control, cohort, and randomized placebo-controlled studies) provide compelling evidence about the chemopreventive ability of aspirin against colorectal cancer. However, the modest neoplasia reduction rate (30–50%; refs. 5–7) and risk of toxicity (e.g., increased risk of gastrointestinal bleeding and hemorrhagic strokes) have lead most decision analyses to conclude that aspirin is not cost-effective for colorectal cancer prevention in the general population (8, 9). Additionally, whereas statins may be better tolerated and offers similar risk reductions to aspirin, it is estimated that to prevent a single colorectal cancer, one would need to treat as many as 4,814 subjects (10). Thus, for chemoprevention to be a clinically acceptable strategy, more efficacious and less toxic agents are needed.
Initial studies have indicated that the novel agent polyethylene glycol (PEG) may possess requisite chemopreventive characteristics needed for widespread clinical application (11). Indeed, in the azoxymethane-treated rat model, PEG is the most effective agent reducing neoplasia by 
90% (compared with the 50% decrease seen with nonsteroidal anti-inflammatory agents; ref. 12). PEG is also widely used clinically and has an outstanding safety profile. Although diarrhea can be a side effect, PEG may be very attractive in the large number of patients with chronic constipation (a potential colorectal cancer risk factor; ref. 13). Preliminary epidemiologic data supports the effectiveness of PEG in human colorectal cancer prevention (14). Despite these compelling findings, lack of mechanistic insights has stymied widespread interest in this potent chemopreventive agent.
Chemoprevention can occur by either preventing initiation of carcinogenesis or by arresting tumor progression phase of colon cancer (15). At the cellular level, chemopreventive agents generally increase epithelial apoptosis and/or inhibit proliferation. Suppression of proliferation is critical in preventing clonal expansion of the initiated colonocytes (16–19). These early events in colon carcinogenesis are believed to be driven by activation of the ß-catenin signaling through either truncating mutations of the adenomatous polyposis coli gene or activating mutations in the ß-catenin gene, Ctnnb1 (20). The increased ß-catenin protein levels enables nuclear translocation and transactivation of the lymphoid enhancer factor 1/Tcf-1, leading to transcriptional induction of a number of genes that are important in the early stages of colon carcinogenesis. Our group has previously shown that a number of structurally unrelated chemopreventive agents abrogated the increased ß-catenin signaling through up-regulation of a plasma membrane protein E-cadherin (18, 21). E-cadherin can avidly bind ß-catenin, thereby suppressing the transcriptional activity of ß-catenin by sequestering it away from the nucleus (22–24). The mechanism by which chemopreventive agents up-regulate E-cadherin remains unclear; however, we have growing evidence to show that E-cadherin loss in early colon carcinogenesis may be related to overexpression of its transcriptional repressor SNAIL. For instance, we have shown that targeted down-regulation of SNAIL in the MIN mouse resulted in normalization of mucosal E-cadherin, with a corresponding inhibition of cell proliferation and tumorigenesis (25).
Despite a number of reports attesting to the remarkable chemopreventive efficacy, the mechanisms of action of PEG remain largely unexplored. Previous studies in colon cancer cell lines have suggested that the chemopreventive effect of PEG may, at least partly, be related to cytostatic effects (26). However, no in vivo data has been reported to date. In the present study, we used the azoxymethane-treated rat model as it accurately replicates many of the clinical, genetic, cellular, and morphologic features of human colon carcinogenesis (27, 28). We showed that PEG suppressed morphologic variables [aberrant crypt foci (ACF) and tumors] and restored the microarchitectural organization (as measured by fractal dimension, one of the earliest and most sensitive markers of colon carcinogenesis). Moreover, PEG mitigated the premalignant colonic mucosal hyperproliferation in the azoxymethane-treated rats with an induction of E-cadherin and concomitant decrease in ß-catenin activity (as indicated by decreased nuclear localization) and cyclin D1 (a well-established ß-catenin target that is critical in hyperproliferation). These changes were accompanied by a profound inhibition of SNAIL expression. Our findings in the animal model were mirrored in cell culture, showing that PEG treatment in HT-29 cells suppressed SNAIL with a corresponding decrease in ß-catenin activity. These results provide persuasive evidence that the SNAIL/E-cadherin/ß-catenin axis may be a key molecular target for PEG-mediated colon cancer chemoprevention.
|
Materials and Methods
|
|---|
Experimental Animal Protocol
All animal studies were done in accordance with the institutional animal care and use committee of Evanston-Northwestern Healthcare. Seventy-five male Fisher 344 rats (80–100 g; Harlan, Indianapolis, IN) were maintained on defined (AIN-76A) diet for 2 weeks and then randomized into three equal groups. Group 1 was injected with saline (azoxymethane vehicle) and groups 2 and 3 with azoxymethane 20 mg/kg body weight/wk for 2 weeks (i.p.). Two weeks later, group 3 rats were switched to a PEG-8000 supplemented diet (5 g/100 g diet; ICN, Aurora, OH) and continued until sacrifice at 30 weeks. Rats were sacrificed in a nonfasted state 2 hours after administering BrdUrd (i.p.; 50 mg/kg body weight) to label S-phase cells. Colons were flushed, tumors were scored, and small distal segments were removed and fixed in formalin for immunohistochemical studies. The remaining portions were longitudinally bisected, with one half of the colonic sections fixed flat (70% ethanol) for ACF determination and the other section used to obtain mucosal scrapings for Western blotting as previously described (18). ACF were assessed by staining the segments in 0.2% methylene blue for 2 minutes and scored by an observer blinded to the treatment group. Total as well as ACF with four or more component crypts/ACF were counted. For the studies related to altered fractal dimension, 30 Fisher 344 rats were randomized to azoxymethane or saline treatment as above (20 azoxymethane and 10 saline treated). Seven weeks after the second azoxymethane injection, rats were gavaged with PEG-3350 or vehicle for 1 week and then euthanized with the colons subjected to four-dimensional elastic light-scattering fingerprinting analysis as described below.
Immunohistochemical Staining
Four colonic punch biopsies (2 mm2) per animal taken from large ACF (four or more crypts per foci) and four biopsies per animal taken from the histologically normal colonic mucosa were subjected to immunohistochemical analysis using techniques previously described (18). Briefly, 4 µm paraffin-embedded sections were mounted on Superfrost+ slides (Vector Laboratories, Burlingame, CA), baked at 60°C for 1 hour, deparaffinized in xylene, and then hydrated in graded series of ethanol washes. The antigen retrieval for BrdUrd, cyclin D1, ß-catenin, and SNAIL were accomplished by pressure microwaving (NordicWare, Minneapolis, MN) in antigen unmasking solution (Vector Laboratories), whereas sections for E-cadherin were treated with 10% Triton X-100 for 15 minutes. Endogenous peroxidase activity was quenched by treating with 3% H2O2 for 5 minutes, and nonspecific binding was blocked by 5% horse serum for 2 hours at 4°C. Sections were then incubated overnight with primary antibodies [anti-BrdUrd (1:50; Zymed Laboratories, Inc., San Francisco, CA), anti-cyclin D1 (1:50), and anti-SNAIL (1:100; Santa Cruz Biotechnology, Santa Cruz, CA); anti-E-cadherin (1:100) and anti-ß-catenin (1:50; BD Biosciences, San Jose, CA)], followed by appropriate biotinylated secondary antibodies. After washing, the samples were developed using Vectastatin Elite ABC kit (Vector Laboratories). Only complete longitudinal crypts extending from the muscularis mucosa to colonic lumen were counted for immunohistochemical labeling (eight crypts per colon and six rats in each group). Staining intensity was measured on a five-point scale by a gastrointestinal pathologist (J.H.) blinded to the treatment group.
Fractal Dimension
To determine the stage of carcinogenesis that PEG targeted, we assessed one of the earliest described markers of neoplastic transformation of the colon, fractal dimension (29–32). The fractal dimension of fresh colonic tissue (within 1 hour of sacrifice) was determined using four-dimensional elastic light-scattering fingerprinting, as previously described (31). Briefly, these determinations are based on the fact that Fourier transformation of the angular distribution (at 550 nm wavelength) of the scattered light yield two-point mass density correlation function between local tissue regions separated by distance r (1 µm < r < 50 µm), C (r) 
rD–3, with D being fractal dimension that can be extrapolated from the linear slopes of C(r) in the linear regions of log-log scale of this equation.
Cell Culture and PEG Treatment
The human colon cancer cell line HT-29 (American Type Culture Collection, Manassas, VA) was cultured in McCoy's 5A medium with 10% serum. Before PEG treatment, the cells were subcultured in a low serum medium (0.5%) and seeded in six-well plates (105 cells/mL). Based on previous studies, HT-29 cells were treated with 5% PEG-3350 for 24 hours. Cells were then harvested and subjected to protein and mRNA measurements.
Western Blot Analysis
Western blotting was done using standard techniques as previously described. Briefly, 30 µg protein were subjected to SDS-PAGE, transferred to polyvinylidene difluoride membranes (Amersham Pharmacia, Piscataway, NJ), blocked with 5% nonfat milk and probed with specific antibodies (proliferating cell nuclear antigen, ß-catenin, E-cadherin, and ß-actin) using standard techniques. Xerograms were developed with enhanced chemiluminescence (Amersham Pharmacia) and quantitated with densitometry. Consistency in protein loading was controlled by probing stripped blots for ß-actin/
-Tubulin.
Reverse Transcriptase-PCR
HT-29 cells were treated with 5% PEG-3350 for 2 hours and RNA was extracted with TRI Reagent (Sigma Chemical Co., St. Louis, MO) as previously described (33). The cDNA was synthesized using 5 µg RNA and Superscript RT (Invitrogen Life Technologies, Carlsbad, CA). Amplification of SNAIL mRNA was done using nested PCR protocols (34). Cyclophilin was used as a control for RNA loading (33).
Luciferase Reporter Assay
To determine luciferase reporter activity, Tcf luciferase constructs (0.5 µg), containing the wild-type (pTOPFLASH) or mutant (pFOPFLASH; Upstate, Charlottesville, VA) Tcf binding sites, were transfected into HT-29 cells (5 x 105 per well). Transfection experiments were carried out in triplicate using Lipofectamine 2000 (Invitrogen Life Technologies) following the instructions of the manufacturer. In addition, the cells were cotransfected with an internal control (0.1 µg pRL-TK Renilla luciferase vector; Promega, Madison, WI). The cells were incubated for 48 hours after the transfection and then treated with PEG or vehicle for 24 hours. The cells were trypsinized, washed with PBS, and lysed in 1x passive lysis buffer (Dual Luciferase kit; Promega). To measure the activities of firefly and Renilla luciferase, 20 µL aliquots of the supernatant were transferred into a 96-well plate and assayed in a 1420-Multilabel counter luminometer (Perkin-Elmer Life Sciences, Wellesley, MA) using reagents from the Dual Luciferase kit (Promega). The firefly (TOPFLASH or FOPFLASH) luciferase activity was corrected for Renilla luciferase activity (pRL-TK) to control for transfection efficiency. TOPFLASH activity was also normalized to the FOPFLASH activity. Data expressed as the mean of triplicate values of the normalized TOPFLASH activity of the PEG-treated cells relative to the control.
Statistical Methods
ANOVA for a nested design was used to examine the effects of carcinogen treatment and PEG supplementation on the tumor/ACF burden and BrdUrd labeling. Quantitative densitometery values were compared by unpaired Student's t test. Differences with P < 0.05 were considered statistically significant. Values were expressed as mean ± SE as indicated.
|
Results
|
|---|
PEG Inhibits Initiation of Colon Carcinogenesis in Azoxymethane-Induced Carcinogenesis
PEG supplementation was exceedingly well-tolerated with no evidence of any associated toxicities, and all animal groups showed comparable weight gain (data not shown). As shown in Fig. 1
, PEG significantly decreased tumorigenesis by 76% (P < 0.001), total ACF by 74% (P < 0.001), and complex ACF by 64% (P < 0.001; n = 12 from each group). These results are in agreement with numerous other reports showing remarkable efficacy of PEG in suppressing ACF and tumor development (17, 35).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 1. Effect of azoxymethane treatment and PEG supplementation on ACF and tumor incidence. Twelve animals each from saline- or azoxymethane-treated groups on AIN-76A diet and 12 animals from azoxymethane-PEG–supplemented group were examined for the presence of tumors and ACF as described in Materials and Methods. Rats injected with saline did not develop any ACF or tumors. Effect of PEG treatment on the percentage decline in the number of total and larger ACF ( 4AC/ACF) and tumors compared with unsupplemented azoxymethane rats. Columns, mean; bars, SE. *, P < 0.001 for total ACF, large ACF, and tumors in PEG-supplemented groups compared with respective controls from azoxymethane-alone unsupplemented groups.
|
|
PEG Normalizes Azoxymethane-Induced Altered Colonic Fractal Dimension
We next determined whether PEG suppressed the initiation stages of colon carcinogenesis by assessing the mucosal fractal dimension. Alteration in this microarchitectural variable is one of the earliest events of colon carcinogenesis (30). As shown in Fig. 2
, fractal dimension of the colonic mucosa of azoxymethane-treated rats assessed 8 weeks subsequent to azoxymethane treatment was significantly increased compared with age-matched, saline-treated rats (P = 0.001). In the short-term PEG-supplemented group, the fractal dimension completely normalized to the values of saline-treated group (P = not significant) and was markedly lower than the azoxymethane-treated group that did not receive PEG (P = 0.004). These data suggest that PEG causes reversion of fundamental microarchitectural events that are altered in colon carcinogenesis.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 2. Effect of short-term PEG treatment on mucosal microarchitecture as measured by fractal dimension. The fractal dimension was determined by four-dimensional elastic light-scattering fingerprinting analysis of the uninvolved mucosa within 1 h of sacrifice (see Materials and Methods). From each animal, four-dimensional elastic light-scattering fingerprinting measurements were average recordings from >50 tissue sites (1 mm2 each) that were evenly distributed throughout the colonic surface. This variable was significantly increased in the rats after 8 wks of azoxymethane (AOM) treatment (P = 0.001; n = 10) compared with age-matched, saline-treated rats. In comparison, 1 wk of PEG treatment (2 g gavages/d) resulted in complete normalization of this marker (P = 0.004; compared with azoxymethane-alone group).
|
|
PEG Inhibits Colonocyte Hyperproliferation
Because the diffuse mucosal hyperproliferation is one of the hallmarks of early colon carcinogenesis and a frequent target for chemoprevention, we assessed the effect of PEG on BrdUrd incorporation (marker for cells in S phases). For this, six complete crypts/tissue section (n = 6), extending from the muscularis mucosa to the colonic lumen, were counted and expressed as percentage of crypt cells positive for BrdUrd immunostaining. Consistent with previous reports, we noted that azoxymethane significantly increased BrdUrd labeling within the colonic mucosa (in both normal and aberrant crypts; Fig. 3A
). PEG supplementation markedly reduced the overall BrdUrd incorporation by 78% (P < 0.001; Fig. 3B). Similarly, in HT-29 cells, PEG treatment resulted in a 50% reduction in proliferating cell nuclear antigen expression when compared with vehicle-treated cells (Fig. 3C).
View larger version (39K):
[in this window]
[in a new window]
|
Figure 3. BrdUrd and cyclin D1 incorporation in normal and ACF crypts. Saline- and azoxymethane-treated rats on either AIN-76A diet or PEG-supplemented diet were prepared as described in Materials and Methods. Thirty weeks after diet initiation, rats were injected with BrdUrd and sacrificed 2 h later. Colon segments were fixed in formalin for 4 h, sectioned and cell proliferation measured by BrdUrd incorporation using a detection kit (Zymed Laboratories). A, representative immunhistochemical staining of the normal crypts and ACF from the indicated groups (n = 6 in each group): a, saline-treated rats; b, azoxymethane-treated rats on AIN-76A unsupplemented diet; c, azoxymethane-treated rats on AIN-76A diet supplemented with PEG (x100). Note that ACF crypts are larger with increased BrdUrd labeling in the azoxymethane-alone group compared with the PEG-supplemented group. Expression of cell cycle regulator cyclin D1 in the ACF and associated uninvolved mucosa from the indicated groups: d, saline-treated rats; e, azoxymethane-treated rats on AIN-76A unsupplemented diet; f, azoxymethane-treated rats on AIN-76A diet supplemented with PEG (x200). B, a quantitative representation of the azoxymethane- and PEG-induced changes in the mucosal BrdUrd and cyclin D1 staining (n = 6). Azoxymethane significantly increased the percentage of positive cells labeled with BrdUrd ( , P < 0.001, compared with saline-treated group) or cyclin D1 (*, P < 0.001, compared with saline-treated group), whereas PEG treatment, compared with azoxymethane-alone group, significantly reversed these changes in both BrdUrd (, P < 0.001) as well as cyclin D1 (**, P < 0.001) staining. C, a representative Western blot showing cyclin D1 expression in normal (saline group) and ACF crypts (from azoxymethane and azoxymethane + PEG groups) by Western blotting. PCNA, proliferating cell nuclear antigen. Ethanol-fixed punch biopsies of crypts and of ACF were examined for cyclin D1 expression (n = 6 rats in each group). Blots were reprobed for ß-actin expression to confirm comparable protein loading. Colonic ACF from azoxymethane-treated groups had higher levels of cyclin D1 expression than saline-treated normal crypts (*P < 0.001), whereas PEG treatment decreased the expression of cyclin D1 in azoxymethane-induced ACF (**, P < 0.001; A.U., arbitrary units). D, a representative Western blot and quantitative representation of the effect of PEG on the proliferating cell nuclear antigen (proliferation marker) expression in colon cancer cell line HT-29. HT-29 cells were cultured in McCoy's 5A medium as described in Materials and Methods. The cells were treated with 5% PEG for 24 h and cells were lysed by boiling for 3 min in Laemmli SDS-stop buffer. Compared with control, PEG significantly suppressed proliferating cell nuclear antigen (P < 0.001; n = 6) in these cells.
|
|
PEG Inhibits the Proproliferative Cyclin D1
Because cyclin D1 has been implicated as a major mediator of colonic hyperproliferation (36), we assessed this cell cycle regulator in the microscopically normal mucosa and ACF. Compared with saline-injected rats, azoxymethane treatment resulted in increased mucosal immunostaining of cyclin D1 in both ACF and microscopically normal crypts. This induction of cyclin D1 was abrogated by PEG (by 53% P < 0.001; Fig. 3A and B). As shown in Fig. 3D, these findings were mirrored by immunoblot analysis of colonic mucosal preparations (PEG decreased cyclin D1 by 62%; P < 0.001; n = 6).
PEG Inhibits ß-Catenin Transcriptional Activity
One of the most important regulators of cyclin D1 expression in colon carcinogenesis is ß-catenin (37). ß-Catenin transcriptionally induces cyclin D1 by first translocating to the nucleus, leading to activation of the Tcf/lymphoid enhancer factor transcription factors. Thus, to further elucidate how PEG may be inhibiting cyclin D1 expression, we investigated if PEG could decrease ß-catenin activity (by using a Tcf/lymphoid enhancer factor reporter assay, TOPFLASH) in HT-29 cells. As shown in Fig. 4A
, PEG significantly decreased TOPFLASH activity by 43% (n = 3). For in vivo correlates, we assessed nuclear ß-catenin in azoxymethane-induced ACF and noted a 50% decrease in PEG-supplemented group compared with azoxymethane-treated animals on AIN-76A diet (P < 0.05; Fig. 4B). Thus, our studies strongly suggest that PEG blocks ß-catenin signaling.
View larger version (39K):
[in this window]
[in a new window]
|
Figure 4. A, inhibition of transcriptional activity of ß-catenin/Tcf by PEG in HT-29 cells. As described in Materials and Methods, HT-29 cells were cotransfected with Tcf-4 reporter plasmid (TOPFLASH) or a mutant Tcf-binding site (FOPFLASH), respectively, and pRL-TK Renilla vector as internal control. Twenty-four hours posttransfection, cells were treated with PEG (10%) for 24 h. The values were normalized to the internal control. PEG caused a significant decrease (43%) in the ß-catenin/Tcf activity compared with untreated controls (*, P < 0.001; n = 3). B, immunohistochemical staining of ß-catenin in azoxymethane (a) or azoxymethane + PEG groups (b) also did not reveal any significant differences between the two groups. However, ACF from the azoxymethane-alone group (insets) had mostly nuclear ß-catenin localization, whereas PEG treatment caused its relocalization to the membrane. C, Western blots showing expression of ß-catenin in mucosal sections from azoxymethane-alone and azoxymethane + PEG groups. PEG treatment did not alter total ß-catenin levels in the uninvolved colonic mucosa of azoxymethane-treated animals. Similar effects were seen when HT-29 cells were treated with PEG. Housekeeping ß-actin or -tubulin expressions were used as loading controls for samples from rat or HT-29 cells, respectively.
|
|
PEG Up-Regulates E-cadherin Expression
We assessed lysates from both HT-29 and colonic mucosa from azoxymethane-treated rats and found that total ß-catenin levels seemed unchanged by PEG (Fig. 4C), suggesting that altered subcellular localization was a critical factor in this pathway. E-cadherin is an important modulator of ß-catenin localization that avidly binds, and therefore sequesters, ß-catenin to the plasma membrane (18, 38). To address the role of E-cadherin in the PEG-induced changes in ß-catenin localization, we assessed E-cadherin expression in response to PEG treatment. As measured by immunostaining, compared with saline, azoxymethane significantly decreased E-cadherin expression in ACF (by 72%; P < 0.001; Fig. 5A, a–c
), whereas PEG treatment reversed this loss. Western blot analysis of the ACF revealed similar trends with PEG, preventing azoxymethane-induced down-regulation of E-cadherin (P < 0.01; Fig. 5B). As also shown in Fig. 6
, induction of E-cadherin by PEG was demonstrable in HT-29. Specifically, PEG treatment resulted in a 70 ± 8% induction of E-cadherin protein as measured by immunoblot analysis.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 5. A, E-cadherin and SNAIL immunohistochemical expression in normal and ACF crypts. Colonic biopsies from saline, azoxymethane, and azoxymethane + PEG were immunostained for E-cadherin and its transcriptional repressor SNAIL: a, E-cadherin expression in representative normal crypt (note membrane association of E-cadherin showing increasing expression from crypt base to colonic surface); b, in ACF and uninvolved crypts from azoxymethane alone; c, and in azoxymethane + PEG–supplemented animals. SNAIL expression in normal crypt (d), in crypts from azoxymethane alone (e), and in azoxymethane + PEG–supplemented groups (f). Insets, high-magnification (x400) images of the ACF showing mostly nuclear staining. Compared with normal mucosa, there was a decrease in E-cadherin expression and corresponding increase in SNAIL in ACF crypts from the azoxymethane-treated unsupplemented rats. These changes were partially inhibited by PEG supplementation. B, a representative Western blot showing E-cadherin expression in normal (saline group) and in ACF (from azoxymethane and azoxymethane + PEG groups). Ethanol-fixed biopsies of crypts and of ACF were examined for E-cadherin expression (n = 6 rats in each group). Blots were reprobed for ß-actin expression to confirm comparable protein loading. Compared with saline, azoxymethane significantly decreased E-cadherin (*, P < 0.001), whereas PEG prevented this loss (**, P < 0.01, compared with azoxymethane-alone group).
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 6. Effect of PEG on the E-cadherin and SNAIL levels in colon cancer cell line HT-29. HT-29 cells were cultured in McCoy's 5A medium and treated with 5% PEG for 24 h and then lysed by boiling for 3 min in Laemmli SDS-stop buffer. Protein expression of E-cadherin was quantified by Western blotting. As can be seen compared with control, PEG significantly increased E-cadherin expression (P < 0.05; n = 6) in these cells. mRNA expression of SNAIL was measured by PCR. PEG significantly decreased the SNAIL mRNA levels (P < 0.05).
|
|
PEG Suppresses SNAIL Expression
To elucidate the mechanisms through which PEG induced E-cadherin, we examined its transcriptional repressor, SNAIL. As shown in Fig. 5A (d–f), PEG decreased SNAIL by 40% compared with azoxymethane-treated controls (P < 0.05). Because immunoblotting was not possible given the lability of SNAIL (39), we measured SNAIL mRNA in HT-29 cells and also noted that PEG markedly reduced SNAIL in these cells (52 ± 9, P < 0.05; Fig. 6). This implicates SNAIL down-regulation as a mechanism through which PEG may regulate E-cadherin levels. This, in turn, would allow E-cadherin to sequester ß-catenin and thereby inhibit cyclin D1 expression and abrogate preneoplastic mucosal hyperproliferation.
|
Discussion
|
|---|
These studies, for the first time, provide important insights into the potential mechanisms of the chemopreventive efficacy of PEG. PEG seems to inhibit early events in colon carcinogenesis. From a molecular perspective, PEG treatment resulted in suppression of proliferation through inhibition of ß-catenin–mediated cyclin D1 levels. PEG does not seem to regulate ß-catenin activity by altering the total protein levels but rather by suppression of its nuclear localization. Our data suggests that PEG-induced up-regulation of E-cadherin results in avid binding and sequestration of ß-catenin to the plasma membrane. Finally, we provide evidence that the induction of E-cadherin is related to the decrease in its transcriptional repressor, SNAIL (Fig. 7
).
View larger version (10K):
[in this window]
[in a new window]
|
Figure 7. Proposed paradigm for the molecular mechanism through which PEG suppresses colonic mucosal hyperproliferation.
|
|
The observation that early and late markers of colon carcinogenesis (e.g., ACF and tumors) were suppressed to a similar magnitude suggests that the initiation phases of carcinogenesis were targeted by PEG. Therefore, we looked at one of the earliest markers of colon carcinogenesis, fractal dimension. This variable reflects mucosal architecture at scales ranging from small organelles to groups of cells. Our group has recently reported that alterations in fractal dimension preceded development of ACF in the azoxymethane-treated rat and adenomas in the MIN mouse (30). Moreover, we have shown that fractal dimension is a sensitive early marker in cancer prevention (31). Thus, the demonstration of complete normalization of fractal dimension with short-term PEG treatment in this study supports the contention that PEG reverses the earliest changes of colon carcinogenesis.
From a cellular perspective, our data provides strong evidence that PEG-mediated suppression of colonic neoplasia is mediated, at least partly, by suppression of epithelial proliferation (16, 19). PEG markedly reduced BrdUrd incorporation in both the ACF and the uninvolved mucosa. This was supported by the marked decrease in proliferation in HT-29 cells when treated with PEG. Diffuse mucosal hyperproliferation is the hallmark of colonic carcinogenesis and is a critical early step in enabling rapid expansion of the genetically initiated colonocyte. Indeed, studies have shown that analysis of proliferation from the rectum can accurately predict occurrence of neoplastic lesions throughout the colon (7). Targeting hyperproliferation has been a common theme for a variety of agents, including calcium, ursodeoxycholic acid, etc., in colon cancer chemoprevention (19, 40). Suppressing proliferation has been shown to be a useful biomarker in clinical chemoprevention studies. Based on our results, the antiproliferative actions of PEG seem to be a result of mucosal field effect involving not only the ACF, but also the microscopically normal epithelium. Our results from the cell culture, in keeping with the study from Corpet's laboratory (26), indicate that PEG has a marked cytoprotective effect. Furthermore, antiproliferative effect of PEG complements its proapoptotic effect that we have previously reported in cell culture, azoxymethane-treated rat, and the MIN mouse (17, 41).
To understand the mechanisms by which PEG suppresses colonic hyperproliferation, we focused our attention on ß-catenin as its activation seems to be one of the earliest genetic events critical in development of colon cancer (42). Several studies support the notion that ß-catenin signaling is intimately involved in the colonic hyperproliferation through its regulation of cyclin D1. In our studies, there was an excellent correlation between cyclin D1 and proliferation markers in the uninvolved colonic mucosa of the azoxymethane-treated rat. PEG treatment markedly decreased nuclear ß-catenin in the ACF with a concomitant suppression of cyclin D1. In the uninvolved mucosa, nuclear ß-catenin is generally less evident (24), but we still observed a profound suppression of cyclin D1 with PEG. To further strengthen the link between PEG and suppression of ß-catenin signaling, we assessed this relationship in HT-29 cells. We showed that the suppression in proliferation by PEG was mirrored not only by the decrease in cyclin D1 but also of ß-catenin transcriptional activity as assessed by the TOPFLASH assay. Thus, taken together, our data provides convincing evidence that PEG treatment suppresses ß-catenin signaling as a mechanism for suppressing hyperproliferation.
The importance of the PEG-mediated suppression of ß-catenin signaling is supported by several lines of evidence. As discussed previously, ß-catenin activation is believed to be one of the earliest events in colon carcinogenesis (42). Indeed, activation of ß-catenin is sufficient to induce intestinal tumorigenesis (43). Targeted down-regulation of ß-catenin or cyclin D1 have been shown to have striking antiproliferative and antineoplastic effects (44, 45). Furthermore, ß-catenin or its downstream proliferation target cyclin D1 plays a central role in chemoprevention. For instance, a number of groups, including our own, have shown that ß-catenin is a target for nonsteroidal anti-inflammatory drugs in chemoprevention (24). Importantly, it has been shown that the chemopreventive effect of nonsteroidal anti-inflammatory drugs can be blocked by prevention of ß-catenin down-regulation (46). Thus, there is a strong reason to believe that the suppression of ß-catenin/cyclin D1 signaling by PEG contributes to the antiproliferative and, at least partly, to its chemopreventive effect.
PEG could potentially regulate ß-catenin through either regulating protein levels (generally through alterations in degradation by the ubiquitin-proteosomal system) or subcellular targeting. We noted that in both azoxymethane-treated rat and HT-29 cells, total ß-catenin expression remained unchanged with PEG treatment. However, PEG reduced the nuclear localization of ß-catenin in ACF, suggesting that alterations in subcellular distribution of ß-catenin are central in the ability of PEG to modulate ß-catenin signaling. Alterations in localization of ß-catenin have been seen with a variety of chemopreventive agents, such as nonsteroidal anti-inflammatory drugs (24, 32). Localization of ß-catenin to the plasma membrane is controlled through its avid interaction with E-cadherin. Moreover, E-cadherin up-regulation is a common means of sequestration of the ß-catenin to the plasma membrane (thus preventing nuclear translocation of ß-catenin) and, hence, transcriptional activity. Indeed, our group has shown that E-cadherin induction occurs in response to a wide variety of chemopreventive agents, including the nonsteroidal anti-inflammatory drug nabumetone, ursodeoxycholic acid, and the vitamin D analogue (18, 21) with a concomitant decrease in ß-catenin signaling (21). In this regard, our data with PEG is consistent with these other known chemopreventive agents. Specifically, PEG treatment increased E-cadherin expression in both the azoxymethane-treated rat and HT-29 cells. This mirrored the decrease in ß-catenin signaling as evident by decreased cyclin D1 expression, nuclear ß-catenin localization, and ß-catenin–regulated promoter activity.
We next attempted to elucidate the mechanism(s) by which PEG up-regulated E-cadherin expression. Loss of E-cadherin is a frequent occurrence in colon carcinogenesis generally thought to be through promoter hypermethylation. However, in a number of cancers such as breast, melanoma, gastric, and hepatocellular tumors, an increasingly common motif in E-cadherin silencing is induction of the transcriptional repressor SNAIL (47). We and others have shown that SNAIL is overexpressed in most human colon cancers (32). Furthermore, we have previously shown that targeting SNAIL with antisense phosphorodiamidate morpholino oligomers in the MIN mouse model of intestinal tumorigenesis resulted in a marked induction of E-cadherin in the uninvolved mucosa with a corresponding reduction in tumor incidence. Knockdown of SNAIL was also accompanied by reversion of the abnormalities in fractal dimension (25). In this study, we show that PEG markedly suppressed SNAIL protein expression (as measured by immunohistochemical analysis) in the azoxymethane-treated rat and SNAIL mRNA in HT-29 cells. Although these data would clearly be bolstered by Western blot analysis, it has been shown that SNAIL is so labile that it cannot be detected on Western blot analysis using conventional techniques (39). In many respects, the data obtained with PEG is reminiscent of our anti-SNAIL phosphorodiamidate morpholino oligomer supporting the role of SNAIL down-regulation in PEG-chemoprevention of colon cancer.
Although our data supports the paradigm for the role of PEG in targeting SNAIL/E-cadherin/ß-catenin in inhibiting colon carcinogenesis initiation, several important issues were not addressed in our report. For instance, SNAIL is a transcription factor that is predominantly located in the nucleus, whereas PEG, owing to its large molecular size, is less likely to directly access the nucleus. In this regard, we have preliminarily data that PEG down-regulates plasma membrane–associated epidermal growth factor receptor in the azoxymethane-treated rat and HT-29 cells (48). Epidermal growth factor receptor is known to induce SNAIL (49). Another point that needs to be emphasized is that whereas our article focuses on proliferation, we have previously shown that PEG also induces apoptosis in animal models and cell culture (17, 41). Clearly, both SNAIL and ß-catenin have important effects on apoptosis. Thus, modulation of the SNAIL/E-cadherin/ß-catenin axis may conceivably regulate both the antiproliferative and proapoptotic chemopreventive effects of SNAIL. Finally, our data indicates that PEG inhibits the initiation phase of colon carcinogenesis given that the magnitude of suppression of fractal dimension, small ACF, large ACF, and tumors were similar. However, there may also be an effect on progression although this study did not address this issue.
In summary, we have shown that PEG treatment suppresses the earliest stages of colon carcinogenesis at least partly through its antiproliferative effects. From a signaling perspective, we present a paradigm that PEG down-regulates SNAIL, leading to an increase in plasma membrane E-cadherin, which, in turn, restricts ß-catenin transcriptional activity resulting in a decrease in cyclin D1 and hence proliferation (Fig. 7). Future studies will be designed to elucidate pathways that mediate SNAIL down-regulation by PEG (e.g., epidermal growth factor receptor). In addition, it will be important to directly test whether SNAIL down-regulation is necessary or sufficient for chemopreventive efficacy of PEG.
|
Footnotes
|
|---|
Grant support: National Cancer Institute grants 1RO3CA10549 and U01CA11125.
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.
Note: This article was presented in part in abstract form at Digestive Disease Week Meetings.
Received 1/30/06;
revised 5/ 9/06;
accepted 5/31/06.
|
References
|
|---|
Jemal A, Murray T, Ward E, et al. Cancer statistics 2005. CA Cancer J Clin 2005;55:10–30.[Abstract/Free Full Text] Hawk ET, Umar A, Viner JL. Colorectal cancer chemoprevention—an overview of the science. Gastroenterology 2004;126:1423–47.[CrossRef][Medline] Serrano D, Lazzeroni M, Decensi A. Chemoprevention of colorectal cancer: an update. Tech Coloproctol 2004;8 Suppl 2:248–52.[CrossRef] Arber N, Levin B. Chemoprevention of colorectal cancer: ready for routine use? Curr Top Med Chem 2005;5:517–25.[CrossRef][Medline] Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med 2003;348:891–9.[Abstract/Free Full Text] Sandler RS. Aspirin and other nonsteroidal anti-inflammatory agents in the prevention of colorectal cancer. Important Adv Oncol 1996;123–37. Sandler RS, Baron JA, Tosteson TD, Mandel JS, Haile RW. Rectal mucosal proliferation and risk of colorectal adenomas: results from a randomized controlled trial. Cancer Epidemiol Biomarkers Prev 2000;9:653–6.[Abstract/Free Full Text] Ladabaum U, Chopra CL, Huang G, Scheiman JM, Chernew ME, Fendrick AM. Aspirin as an adjunct to screening for prevention of sporadic colorectal cancer. A cost-effectiveness analysis. Ann Intern Med 2001;135:769–81.[Abstract/Free Full Text] Suleiman S, Rex DK, Sonnenberg A. Chemoprevention of colorectal cancer by aspirin: a cost-effectiveness analysis. Gastroenterology 2002;122:78–84.[CrossRef][Medline] Poynter JN, Gruber SB, Higgins PD, et al. Statins and the risk of colorectal cancer. N Engl J Med 2005;352:2184–92.[Abstract/Free Full Text] Di Palma JA, Smith JR, Cleveland M. Overnight efficacy of polyethylene glycol laxative. Am J Gastroenterol 2002;97:1776–9.[CrossRef][Medline] Corpet DE, Tache S. Most effective colon cancer chemopreventive agents in rats: a systematic review of aberrant crypt foci and tumor data, ranked by potency. Nutr Cancer 2002;43:1–21.[CrossRef][Medline] Roberts MC, Millikan RC, Galanko JA, Martin C, Sandler RS. Constipation, laxative use, and colon cancer in a North Carolina population. Am J Gastroenterol 2003;98:857–64.[CrossRef][Medline] Dorval ED, Viguier J, Bertrand P, et al. Prevention of colorectal adenomas by polyethylene glycol (PEG): a population-based study of 1165 colonoscopies in France. Proc Am Assoc Cancer Res 2003;44:A-879. Janne PA, Mayer RJ. Chemoprevention of colorectal cancer. N Engl J Med 2000;342:1960–8.[Free Full Text] Do K, Barnard GF. Effect of concomitant polyethylene glycol and celecoxib on colonic aberrant crypt foci and tumors in F344 rats. Dig Dis Sci 2005;50:1304–11.[CrossRef][Medline] Roy HK, Gulizia J, DiBaise JK, et al. Polyethylene glycol inhibits intestinal neoplasia and induces epithelial apoptosis in Apc(min) mice. Cancer Lett 2004;215:35–42.[CrossRef][Medline] Wali RK, Khare S, Tretiakova M, et al. Ursodeoxycholic acid and F(6)-D(3) inhibit aberrant crypt proliferation in the rat azoxymethane model of colon cancer: roles of cyclin D1 and E-cadherin. Cancer Epidemiol Biomarkers Prev 2002;11:1653–62.[Abstract/Free Full Text] Wali RK, Stoiber D, Hart J, Bissonnette M. Polyethylene suppresses hyperproliferation and increases apoptosis in AOM-induced colonic aberrant crypts. Gastroenterology 2002;122:A–241. Wong NA, Pignatelli M. ß-catenin—a linchpin in colorectal carcinogenesis? Am J Pathol 2002;160:389–401.[Abstract/Free Full Text] Roy HK, Karoski WJ, Ratashak A, Smyrk TC. Chemoprevention of intestinal tumorigenesis by nabumetone: induction of apoptosis and Bcl-2 downregulation. Br J Cancer 2001;84:1412–6.[CrossRef][Medline] Boon EM, Keller JJ, Wormhoudt TA, et al. Sulindac targets nuclear ß-catenin accumulation and Wnt signalling in adenomas of patients with familial adenomatous polyposis and in human colorectal cancer cell lines. Br J Cancer 2004;90:224–9.[CrossRef][Medline] Narayan S. Curcumin, a multi-functional chemopreventive agent, blocks growth of colon cancer cells by targeting ß-catenin-mediated transactivation and cell-cell adhesion pathways. J Mol Histol 2004;35:301–7.[CrossRef][Medline] Roy HK, Karolski WJ, Wali RK, Ratashak A, Hart J, Smyrk TC. The nonsteroidal anti-inflammatory drug, nabumetone, differentially inhibits ß-catenin signaling in the MIN mouse and azoxymethane-treated rat models of colon carcinogenesis. Cancer Lett 2005;217:161–9.[CrossRef][Medline] Roy HK, Iversen P, Hart J, et al. Down-regulation of SNAIL suppresses MIN mouse tumorigenesis: modulation of apoptosis, proliferation, and fractal dimension. Mol Cancer Ther 2004;3:1159–65.[Abstract/Free Full Text] Parnaud G, Corpet DE, Gamet-Payrastre L. Cytostatic effect of polyethylene glycol on human colonic adenocarcinoma cells. Int J Cancer 2001;92:63–9.[CrossRef][Medline] Ponz de Leon M, Roncucci L, Di Donato P, et al. Pattern of epithelial cell proliferation in colorectal mucosa of normal subjects and of patients with adenomatous polyps or cancer of the large bowel. Cancer Res 1988;48:4121–6.[Abstract/Free Full Text] Risio M, Lipkin M, Candelaresi G, Bertone A, Coverlizza S, Rossini FP. Correlations between rectal mucosa cell proliferation and the clinical and pathological features of nonfamilial neoplasia of the large intestine. Cancer Res 1991;51:1917–21.[Abstract/Free Full Text] Cross SS. Fractals in pathology. J Pathol 1997;182:1–8.[Medline] Roy HK, Kim YL, Wali RK, et al. Spectral markers in preneoplastic intestinal mucosa: an accurate predictor of tumor risk in the MIN mouse. Cancer Epidemiol Biomarkers Prev 2005;14:1639–45.[Abstract/Free Full Text] Roy HK, Liu Y, Wali RK, et al. Four-dimensional elastic light-scattering fingerprints as preneoplastic markers in the rat model of colon carcinogenesis. Gastroenterology 2004;126:1071–81; discussion 1948.[CrossRef][Medline] Roy HK, Smyrk TC, Koetsier J, Victor TA, Wali RK. The transcriptional repressor SNAIL is overexpressed in human colon cancer. Dig Dis Sci 2005;50:42–6.[CrossRef][Medline] Kunte DP, Wali RK, Koetsier JL, et al. Down-regulation of the tumor suppressor gene C-terminal Src kinase: an early event during premalignant colonic epithelial hyperproliferation. FEBS Lett 2005;579:3497–502.[Medline] Yanez-Mo M, Lara-Pezzi E, Selgas R, et al. Peritoneal dialysis and epithelial-to-mesenchymal transition of mesothelial cells. N Engl J Med 2003;348:403–13.[Abstract/Free Full Text] Corpet DE, Parnaud G. Polyethylene-glycol, a potent suppressor of azoxymethane-induced colonic aberrant crypt foci in rats. Carcinogenesis 1999;20:915–8.[Abstract/Free Full Text] Arber N, Lightdale C, Rotterdam H, et al. Increased expression of the cyclin D1 gene in Barrett's esophagus. Cancer Epidemiol Biomarkers Prev 1996;5:457–9.[Abstract] Mann B, Gelos M, Siedow A, et al. Target genes of ß-catenin-T cell-factor/lymphoid-enhancer-factor signaling in human colorectal carcinomas. Proc Natl Acad Sci U S A 1999;96:1603–8.[Abstract/Free Full Text] Conacci-Sorrell M, Simcha I, Ben-Yedidia T, Blechman J, Savagner P, Ben-Ze'ev A. Autoregulation of E-cadherin expression by cadherin-cadherin interactions: the roles of ß-catenin signaling, Slug, and MAPK. J Cell Biol 2003;163:847–57.[Abstract/Free Full Text] Zhou BP, Deng J, Xia W, et al. Dual regulation of Snail by GSK-3b-mediated phosphorylation in control of epithelial-mesenchymal transition. Nat Cell Biol 2004;6:931–40.[CrossRef][Medline] Wali RK, Stoiber D, Nguyen L, et al. Ursodeoxycholic acid inhibits the initiation and postinitiation phases of azoxymethane-induced colonic tumor development. Cancer Epidemiol Biomarkers Prev 2002;11:1316–21.[Abstract/Free Full Text] Roy HK, DiBaise JK, Black J, Karolski WJ, Ratashak A, Ansari S. Polyethylene glycol induces apoptosis in HT-29 cells: potential mechanism for chemoprevention of colon cancer. FEBS Lett 2001;496:143–6.[Medline] Yamada Y, Yoshimi N, Hirose Y, et al. Frequent ß-catenin gene mutations and accumulations of the protein in the putative preneoplastic lesion lacking macroscopic aberrant crypt foci appearance, in rat colon carcinogenesis. Cancer Res 2001;60:7699–701. Harada N, Tamai Y, Ishikawa T, et al. Intestinal polyposis in mice with a dominant stable mutation of the ß-catenin gene. EMBO J 1999;18:5931–42.[CrossRef][Medline] Arber N, Doki Y, Han EK, et al. Antisense to cyclin D1 inhibits the growth and tumorigenicity of human colon cancer cells. Cancer Res 1997;57:1569–74.[Abstract/Free Full Text] Roh H, Green DW, Boswell CB, Pippin JA, Drebin JA. Suppression of ß-catenin inhibits the neoplastic growth of APC-mutant colon cancer cells. Cancer Res 2001;61:6563–8.[Abstract/Free Full Text] Li H, Liu L, David ML, et al. Pro-apoptotic actions of exisulind and CP461 in SW480 colon tumor cells involve ß-catenin and cyclin D1 down-regulation. Biochem Pharmacol 2002;64:1325–36.[CrossRef][Medline] Blanco MJ, Moreno-Bueno G, Sarrio D, et al. Correlation of Snail expression with histological grade and lymph node status in breast carcinomas. Oncogene 2002;21:3241–6.[CrossRef][Medline] Huang JC, Roy HK, Koetsier J, Kunte DP, Wali RK. Polyethylene glycol (PEG) mediated chemoprevention of colorectal cancer (CRC) involves downregulation of epidermal growth factor receptor (EGFR). Am J Gastroenterol 2005;100:S154. Lu Z, Ghosh S, Wang Z, Hunter T. Downregulation of caveolin-1 function by EGF leads to the loss of E-cadherin, increased transcriptional activity of ß-catenin, and enhanced tumor cell invasion. Cancer Cell 2003;4:499–515.[CrossRef][Medline]
Top
Abstract
Introduction
