Polyethylene Glycol Mediated Colorectal Cancer Chemoprevention: Roles of Epidermal Growth Factor Receptor and Snail

Abstract

Polyethylene glycol (PEG) is a clinically widely used agent with profound chemopreventive properties in experimental colon carcinogenesis. We previously reported that Snail/β-catenin signaling may mediate the suppression of epithelial proliferation by PEG, although the upstream events remain unclear. We report herein the role of epidermal growth factor receptor (EGFR), a known mediator of Snail and overepressed in ~80% of human colorectal cancers (CRC), on PEG-mediated anti-proliferative and hence anti-neoplastic effects in azoxymethane (AOM)-rats and HT-29 colon cancer cells. AOM-rats were randomized to either standard diet or one with 10% PEG 3350 and euthanized 8 weeks later. The colonic samples were subjected to immunohistochemical or Western blot analyses. PEG decreased mucosal EGFR by 60% (p<0.001). Similar PEG effects were obtained in HT-29 cells. PEG suppressed EGFR protein via lysosmal degradation with no change in mRNA levels. To show that EGFR antagonism per se was responsible for the antiproliferative effect, we inhibited EGFR by either pre-treating cells with gefitinib or stably transfecting with EGFR-shRNA and measured the effect of PEG on proliferation. In either case PEG effect was blunted suggesting a vital role of EGFR. Flow cytometric analysis revealed that EGFR-shRNA cells, besides having reduced membrane EGFR also expressed low Snail levels (40%), corroborating a strong association. Furthermore, in EGFR silenced cells PEG effect on EGFR or Snail was muted, similar to that on proliferation. In conclusion, we show that EGFR is the proximate membrane signaling molecule through which PEG initiates antiproliferative activity with Snail/β-catenin pathway playing the central intermediary function.

INTRODUCTION

Colorectal Cancer (CRC) is second most common cause of cancer-related deaths in the United States with 49,960 deaths estimated in 2008 (1). Since early stage colorectal cancer (curable) is generally clinically silent, screening the entire at-risk (age >50) asymptomatic population has been advocated. However, the power of tests such as colonoscopy in reducing mortality and even incidence of CRC is juxtaposed with the reticence of the asymptomatic population to undergo invasive screening tests. From a patient preference perspective, chemoprevention represents an attractive approach for CRC prevention with numerous agents showing efficacy in epidemiological and preclinical studies. However, the results of randomized-controlled studies with several promising agents have been disappointing owing to suboptimal efficacy and/or higher toxicity of the agents utilized. For instance, although the evidence supporting aspirin related CRC inhibition is convincing (2, 3), the U.S. Preventive Service Task Force concludes that the risk of toxicity with aspirin outweighs it chemopreventive benefit (4). This was the rationale for the employment of Cyclooxygenase (COX) 2 inhibitors which cause less GI toxicity. However, the randomized-controlled trial while demonstrating efficacy also noted a 2–3 fold increase in cardiac toxicity (5, 6). Furthermore, although the epidemiological evidence for a protective role of folate seems compelling, recent clinical data suggests that folic acid supplementation may enhance tumor progression (7). Likewise, dietary calcium supplementation by and large seems to be safe but its chemopreventive efficacy is modest (10–15% risk reduction) (8). Thus, safe and more effective chemopreventive agents are urgently needed.

Over the past several years, Corpet and colleagues have indicated that polyethylene glycol (PEG) has remarkable efficacy as a chemopreventive agent (9, 10). Indeed, the ability of this novel agent to suppress tumors or aberrant crypt foci (ACF) in the azoxymethane (AOM)-treated rat model was >90%, generally outperforming reported efficacies of NSAIDS or that of other known chemopreventive agents (11). Our laboratory has both confirmed these findings (12) and extended it to the MIN mouse, another well-validated model of experimental colon carcinogenesis (13). Preliminary epidemiological data support the potential of PEG to suppress adenomas in humans (14). This is coupled with the excellent safety data from the long term use of PEG as a cathartic agent (15). Indeed, PEG is available over-the-counter and is widely clinically used.

One important issue which has stymied the acceptance of PEG as a chemopreventive agent is the lack of understanding of its mechanism of action. During early colon carcinogenesis diffuse epithelial hyperproliferation along with suppression of apoptosis in the colonic mucosa is a key event which is often driven by alterations in the β-catenin signaling pathway (16). Downregulation of β-catenin signaling, with a consequent suppression of proliferation, is a hallmark of a variety of chemopreventive agents including NSAIDS, green tea etc (17, 18). PEG has previously been shown to suppress colonic epithelial proliferation (12, 19). In both cell culture and the AOM-treated rat model, our laboratory has further shown that the antiproliferative effect of PEG was accompanied by marked suppression of Snail, leading to induction of E-cadherin and subsequent sequestration of β-catenin away from the nucleus (12). This leads to decreased β-catenin-dependent transcriptional activity and cyclin D1 expression, an important regulator of cellular proliferation (12). The relevance of Snail in tumorigenesis is evident from our earlier studies showing that selective downregulation of this transcriptional repressor, using antisense phosphorodiamidate morpholino oligomer, decreased epithelial proliferation in the normal histological mucosa with a concomitant reduction in intestinal tumorigenesis in the MIN mouse model (20). Thus, there is compelling rationale to assume that inhibition of Snail signaling by PEG is critical for its chemopreventive effect.

PEG being a large molecule, the central issue is how is it accessible to the cell to regulate the expression of Snail which is predominantly located in the nucleus. We reasoned that PEG may likely be influencing Snail-related signaling pathways by interacting with upstream targets at the cell membrane level. In this context, epidermal growth factor receptor (EGFR), a transmembrane glycoprotein, has been implicated in the control of intracellular Snail levels (21, 22). Moreover, EGFR is overexpressed in most CRCs (23) making it an excellent candidate for a cell surface based PEG target. In the present studies, we demonstrate that PEG downregulates EGFR in both AOM-treated rat and colon cancer HT-29 cells. This downregulation appears to be a result of increased endocytic lysosmal degradation. Moreover, using a ShRNA approach, we demonstrate that EGFR downregulation is central to PEG responsiveness with regards to its role in anti-proliferation and Snail regulation.

MATERIALS AND METHODS

Experimental Animal Protocols

All animal studies were performed in accordance with the Institutional Animal Care and Use Committee (IACUC) of Evanston-Northwestern Healthcare. Twenty four male Fisher 344 rats (100–120g; Harlan, Indianapolis, IN) were maintained on defined (AIN-76A) diet for 2 weeks and then randomized into 3 equal groups. Group 1 was injected with saline (AOM vehicle) and groups 2 and 3 with AOM 20mg/kg body weight/week for 2 weeks (i.p.). Two weeks later, group 3 rats were switched to a PEG-3350 supplemented-diet (10g/100g diet, Harlan Teklad, Madison, WI) and continued until sacrifice 8 weeks after AOM. In another experiment, to study the short-term dose effect of PEG-3350 on EGFR expression, sixteen additional AOM-treated rats were randomized to receive 5 or 10% gavages of PEG (w/v) for a week before euthanizing. Rats were provided water ad libitum and housed in polycarbonate cages in an environmentally controlled room (daily12-h fluorescent light and dark cycles at 24°C and a relative humidity of 70%). All rats were euthanized in a non-fasted state and the colons were isolated and flushed with phosphate buffered saline (PBS) pH 7.4. While small distal segments were removed and fixed in formalin for immunohistochemical studies, the reminder of the colon was opened longitudinally to expose the luminal mucosa. The colonic mucosa was collected by gentle scrapping using glass slides, homogenized in Laemmli buffer and subjected to Western blot analysis as previously described (24).

Immunohistochemical (IHC) Staining

Formalin-fixed colonic tissue sections were examined by immunostaining to assess changes in the expression levels of Snail and EGFR. Briefly, paraffin embedded colonic segments were sliced (4 microns thick) along the vertical axis of the crypts and the sections mounted on Vectabond-coated Superfrost + slides. These slides were then baked for 1 h at 60–70°C, deparaffinized in xylene and rehydrated by graded ethanol washes. The antigen epitope retrieval for Snail and EGFR were achieved by pressure microwaving (NordicWare, Minneapolis, MN) in antigen unmasking solution for 9 min × 2 cycles (Vector Laboratories). Endogenous peroxidase activity was quenched by treating with 3% H₂O₂ for 5 min and nonspecific binding was blocked by 5% horse serum for 2 h. Sections were then incubated overnight (at 4°C) with appropriate primary antibodies [anti-Snail (SNAI1; T-18, 1:100) and anti-EGFR (1:150) (Santa Cruz Biotechnology, CA)] followed by incubation with suitable biotinylated secondary antibodies (1:2000). After washing, the samples were incubated with avidin-biotin peroxidase complex using Vectastatin Elite ABC kit (Vector Laboratories, CA) and the stain developed with DAB (Vector Laboratories, CA). Only complete longitudinal crypts, extending from the muscularis mucosa to colonic lumen, were included for IHC analysis (8 crypts/colon and six rats in each group). Staining intensity was quantified on a 5 point scale by a gastrointestinal pathologist (JH) blinded to the treatment group.

Cell culture and PEG Treatment

The human colon cancer cell line HT-29 (American Type Culture Collection, VA) was cultured in McCoy’s 5A medium with 10% fetal bovine serum (FBS). The cells were seeded in 100 mm Petri-dishes (10⁵ cells/ml), washed twice with phosphate buffered saline (PBS) and serum starved (0.5% FBS) for 72 h before treating with PEG for the indicated time. Cells were then harvested and subjected to flow cytometric analysis, Western blotting and RT-PCR.

Flow Cytometry Analyses

This technique was employed to measure the surface expression of EGFR as well as levels of intracellular Snail in HT-29 cells treated with PEG. Cells (70% confluent) were treated with PEG as described above, and then fixed in 4% buffered paraformaldehyde for 30 min. For EGFR surface staining, the cells were washed twice in a PBS buffer (pH 7.4) containing 2% FBS, 0.2% BSA and 0.02% sodium azide and for intracellular staining of Snail 0.2% saponin was also included in this buffer. Cells were later incubated in their respective PBS buffers with either anti-EGFR 528 (1:200; kindly provided by Dr. H Band, Evanston, IL) or anti-Snail T-18 (1:200) antibodies at room temperature for 1 h. After washing 3 times with their respective buffers, cells were incubated with secondary antibody Alexa 488 Green FI1 labeled anti goat (for Snail) or anti-mouse (for EGFR) (Invitrogen) for 40 min. Cells were subsequently washed 3 times with their respective buffers and subjected to flow cytometric analysis (Becton Dickinson Labware, Franklin Lakes, NJ).

Western Blot Analysis

Western blotting was performed using standard techniques as previously described (24). Briefly, samples were prepared for SDS-PAGE analysis by adding Laemmli sample buffer to cleared whole cell lysates and then heated for 5 min at 95°C prior to loading. An equal amount (40μg) of the protein samples from rat colonic mucosa or HT-29 cells were subjected to SDS-polyacrylamide gel electrophoresis, transferred to PVDF (polyvinylidene difluoride) membranes (Millipore, MA), blocked with 5% non-fat milk and probed with appropriate antibodies (anti-EGFR, 1:200; anti-PCNA, 1:250) using standard techniques. Xerograms were developed with enhanced chemiluminescence (Santa Cruz Biotechnology, Santa Cruz, CA) and image analysis was performed using image acquisition analysis software (Labworks, 4.6; UVP, Upland, CA). Expression levels were normalized to the levels of β-actin or α-tubulin as controls after stripping and re-probing with anti-β-actin (1:300) or anti-α-tubulin (1:200) antibody.

Reverse- Transcriptase Polymerase Chain Reaction (RT-PCR)

HT-29 cells were treated with 10% PEG-3350 for 24h and RNA extracted with TRI Reagent (Sigma) as previously described (25). The cDNA was synthesized using 5μg of RNA and Superscript RT (Invitrogen Life Technologies, CA). Amplification of Snail mRNA was performed using nested PCR protocols (26). Cyclophilin was used as a control for RNA loading (25).

Cell proliferation assay

To demonstrate that EGFR modulation by PEG is important in its anti-proliferative activity, we pretreated HT-29 cells with a known EGFR inhibitor to study if it blunts PEG’s antiproliferative activity. For this experiment HT-29 cells were pre-treated with EGFR-inhibitor, Gefitinib (25μM; Iressa; Astra Zeneca, Cheshire, England) for 2h at 37°C before treating with 10% PEG 3350 for 24 h. For control, cells were pretreated with equal volume of DMSO (Gefitinib vehicle) and then with PEG. For these assays cells were seeded in 96-well plates and at the end of the incubation the media was replaced with fresh media (100 μl) containing 5μl of WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1, 3-benzene disufonate) reagent (Roche Diagnostics, Indianapolis, IN). After 5–15 min incubation, the absorbance was read at 440nm in a Spectramax Plus Spectrophotometer plate reader (Molecular Devices, Sunnyvale, CA).

EGFR knockdown assay

To further explore the importance of EGFR in PEG-mediated anti-proliferation, we studied the effect of PEG on cellular proliferation in EGFR deficient HT-29 cells. We knocked down EGFR gene expression in these cells using short haipin-loop RNA (shRNA) (Origene, Rockville, MD). The EGFR-shRNA and control vectors were transfected in HT-29 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. After transfection, cells were incubated at 37°C in a humidified 5% CO₂ incubator for 72 h. Transfectants were selected by puromycin (0.5 μg/ml). The cells were then incubated with 10% PEG for 24 h before assaying expression of EGFR and proliferating cell nuclear antigen (PCNA).

Statistical Methods

Values were expressed as mean ± SE as indicated. Quantitative densitometry values were compared by unpaired Student’s t test. Differences with p< 0.05 were considered statistically significant.

RESULTS

PEG downregulates colonic epithelial EGFR in AOM-induced rats

PEG supplementation was well tolerated by the rats with no apparent toxicity as reflected by normal body weight gain, reported previously (12). Since cellular hyperproliferation has been found to be associated with overexpression of EGFR in the premalignant colonic mucosa of humans as well as AOM-rat model of colon cancer (27, 28), we reasoned if PEG’s ability to inhibit proliferation (12) is mediated via modulation of the mucosal EGFR expression. For these experiments we used the AOM-rat model and as shown in Figure 1A (representative Western Blot), there was a significant induction of EGFR expression (p< 0.001) at premalignant time point (8 weeks post-AOM treatment) compared to saline-treated controls. In contrast, PEG supplementation in AOM-treated rats (10g PEG-3350/100g diet) significantly decreased EGFR expression by 60% (p< 0.001) almost to the control levels of saline-treated rats. In another AOM-rat experiment (Figure 1B) to study if relatively shorter PEG treatments could also downregulate EGFR, 7 week post AOM treated rats were gavaged with vehicle or 5 or 10% (w/v) PEG-3350 for a week prior to sacrifice. As shown, only a week’s oral PEG gavages caused significant decline in EGFR expression. While 5% PEG showed a trend, 10% PEG caused a highly significant (p<0.001) decrease in EGFR levels.

Figure 1.

Effect of PEG on AOM-induced rat colonic mucosal EGFR protein expression. For these experiments we used AOM-rat model as described in the “Materials and Methods” section. The rats were sacrificed 8 weeks of AOM/saline and colonic mucosa was scrapped and subjected to SDS-PAGE electrophoresis and EGFR expression determined by Western Blot analysis. As shown in Panel A, there was a significant induction of EGFR (p< 0.001) expression at premalignant time point (8 weeks of AOM treatment) compared to saline-treated controls. On the other hand in AOM-treated rats PEG-3350 (10g/100g diet for 8 weeks) supplementation remarkably decreased the EGFR expression (by 60%; p< 0.001 compared to AOM alone) to almost saline-treated control levels. In a separate experiment (Panel B) to study if relatively shorter PEG treatments could also achieve EGFR attenuation, 7 week post AOM rats were gavaged with either 5 or 10% (w/v) PEG-3350 for a week before sacrifice. Only a week of oral PEG gavages to these animals produced astounding decline in EGFR expression. While 5% PEG showed a trend, 10% PEG dose caused a significant (*; p<0.001) decrease in EGFR levels. Housekeeping β-actin was used as a loading control for determining protein expression.

PEG inhibits EGFR expression in HT-29 cells

Since we previously showed that PEG inhibits proliferation of human colon cell line HT-29, we investigated if PEG downregulated EGFR expression in these cells. As depicted by a representative Western blot in Figure 2A, there was a dose and time dependent inhibition of EGFR protein expression in these cells. Both 5 and 10% PEG treatments for 6h or 24h inhibited EGFR expression, however the upper dose (previously shown to be antiproliferative in these cells (12)) caused a significantly greater effect (p<0.05 at 6h and p<0.001 at 24h). Analysis of EGFR expression by flow cytometric analysis confirmed these results (Figure 2B). We further investigated if the effect of PEG occurred at the translational or transcriptional level. As can be seen in a representative RT-PCR blot (Figure 2C), PEG (10% for 24 h) treatment did not cause any significant difference in the mRNA levels in these cells. This suggests that the downregulation of EGFR by PEG may be a result of the post-transcriptional modulation of the receptor.

Figure 2.

Mechanism of EGFR downregulation in HT 29 cells treated with PEG-3350. The human colon cancer cells, HT 29, were seeded in 100 mm Petri-dishes (10⁵ cells/ml) using McCoy’s 5A media containing 10% FBS. After 24 h the cells were washed off the media with PBS before further incubation in serum starved media (0.5% FBS) for 72 h and then treated with 5 or 10% PEG for 6 or 24 h. As shown in the Western blot analysis (Panel A), both 5 and 10% PEG treatment for 6 or 24 h inhibited EGFR expression, however, the effect was statistically significant at 10% dose for both short (p<0.05) and long (p<0.001) exposure times. As can be seen in Panel B, similar effects on membrane EGFR inhibition (55%) could be ascertained by flow cytometric analysis. Further as depicted in the representative RT-PCR blot (Panel C), PEG exposure did not cause any significant differences in the EGFR mRNA levels. To further study if PEG may induce EGFR downregulation by inducing lysosomal degradation, cells were pretreated with 250 nM bafilomycin A1 (BFL; specific lysosomal inhibitor) for 2 h before treating with 10% PEG for 24 h. As can be seen in the Panel D, block in lysosomal degradation almost completely prevented PEG to inhibit EGFR.

PEG downregulates EGFR via Lysosomal degradation in HT-29 Cells

EGFR is recognized to be downregulated by receptor internalization from the plasma membrane into endosomes and other cytoplasmic vesicles where it is degraded by lysosomal/proteosomal processes and may undergo ubiquitination (29). To study the role of lysosomal degradation in PEG induced decrease in EGFR expression, HT-29 cells were pre-treated with specific lysosomal inhibitor, bafilomycin A1 (Calbiochem) prior to treatment with PEG for 24 h. The levels of EGFR expression was analyzed by Western blotting. As shown in Figure 2D, PEG in the absence of the inhibitor resulted in expected decrease in EGFR expression. When the cells were pretreated with bafilomycin for 2h at 37°C there was a small stabilizing effect on the EGFR, however under these conditions PEG’s ability to downregulate EGFR was almost completely lost. There is some evidence to suggest that EGFR degradation could also be blocked by proteosomal inhibitors (29). In our studies when we used lactacystin (Calbiochem) to inhibit proteosomal degradation, there was only a non-significant partial block (by 17%) in EGFR degradation by PEG. Taken together these studies suggest that EGFR downregulation by PEG may be a consequence of lysosomal degradation of EGFR, and only partially facilitated by proteosomal degradation.

EGFR antagonist (Gefitinib) blocks PEG-induced antiproliferative activity

To establish the relevance of EGFR as the primary membrane target of PEG-mediated antiproliferative effects, we investigated if gefitinib (iressa), a known EGFR antagonist, could blunt the effect of PEG on HT-29 cell proliferation. The cells were either pretreated with DMSO (as vehicle; 0.05%) or with 25μM gefitinib for 120 min before incubating with or without PEG for 24 h at 37°C. Proliferation was assayed using WST-1 reagent. As shown in Figure 3A, in control (DMSO) treated cells PEG caused a significant decrease in rate of proliferation; however in gefitinib treated cells the antiproliferative effect of PEG was lost. These studies strongly suggest that PEG’s antiproliferative effects may indeed be mediated by its effect on the EGFR downregulation.

Figure 3.

Causal relationship between PEG induced EGFR downregulation and its antiproliferative activity. HT 29 cells were either pretreated with well known EGFR kinase inhibitor, gefitinib, or transfected with EGFR sh RNA transfections as discussed in “Materials and Methods”. Serum starved cells in 96 well plates were pretreated with DMSO or 25μM gefitinib for 120 min before PEG treatment for 24 h and then assayed for proliferation using standard WST-1 assay. Panel A shows that while in DMSO treated cells, PEG caused more than 50% decrease in proliferation rate, this effect was abolished in cells pretreated with EGFR inhibitor gefitinib. Similarly as shown in Panel B, shRNA that caused ~50% inhibition in EGFR expression as measured by Western blotting, blunted the PEG effect on PCNA (Panel C), a marker of cellular proliferation.

EGFR gene knockdown in HT-29 cells blunts PEG’s antiproliferative activity

To further establish a direct causal relationship between EGFR downregulation and antiproliferative activity of PEG; we determined if specific EGFR knockdown could blunt the effect of PEG on the expression of PCNA in HT-29 cells. The expression of PCNA is known to be stimulated by EGFR (27) and is widely used as a marker of cell proliferation (30, 31). As shown in Figure 3B, transfection of EGFR shRNA into HT-29 cells caused 50% reduction in the expression of EGFR compared to vector transfected cells as control. Whereas in control cells, PEG (10% for 24 h) caused 60% inhibition in the PCNA expression (Figure 3C), in EGFR knockdown cells it produced only 25% inhibition. Taken together these studies provide evidence that EGFR downregulation may be an important mediator of PEG’s antiproliferative effects in colon cancer cells.

EGFR an upstream effecter of PEG-induced inhibition of transcriptional repressor Snail

Our laboratory has previously shown that Snail is activated in colon cancer (32) and PEG treatment inhibits it (12). This downregulation of Snail leads to increased plasma membrane E-cadherin, which in turn limits β-catenin transcriptional activity resulting in cyclin D1 inhibition and hence decreased proliferation (12). Since EGFR is also known to induce Snail (21) and increase cyclin D1 (29), we tested if this membrane receptor may be modulated by PEG and mediate its antiproliferative effects downstream of the Snail/β-catenin pathway. As depicted in Figure 4A (representative immunohistochemical images of colonic sections from AOM treated rats with or without PEG-3350 supplementation), PEG clearly reduced the expression of Snail as well as that of EGFR. To demonstrate a causal relationship between PEG-induced EGFR and Snail expression, we used the EGFR shRNA knockdown approach to study the effect of PEG on both EGFR and Snail expressions in HT-29 cells. Both control and EGFR-shRNA transfected HT-29 cells were treated with 10% PEG-3350 for 24 h. Cells were subjected to flow cytometric analysis to measure EGFR and Snail levels as described in the “Material and Methods” section. As shown in Figure 4B, PEG significantly decreased both EGFR and Snail levels (p<0.001) in the control cells. However, the effect of PEG was blunted for both EGFR and Snail in shRNA-EGFR transfected cells. These studies provide strong evidence that the effect of PEG on Snail and related antiproliferative signaling may be initiated from its upstream modulation of membrane EGFR.

Figure 4.

A causal relationship between PEG-mediated reduced expression of EGFR and Snail. Since EGFR is known to induce Snail which we have previously reported to be inhibited by PEG, we tested if this receptor may be the upstream membrane target for PEG to interact and initiate antiproliferative signaling via Snail. As shown in Panel A (representative immunohistochemical images of colonic sections from AOM treated rats with or without PEG-3350 supplementation), PEG treatment clearly reduced the expression of both Snail and EGFR. To demonstrate a relationship between EGFR and Snail expression, we transfected HT-29 cells with EGFR shRNA and studied the effect of PEG on both EGFR and Snail expressions. Both parental and EGFR-shRNA HT-29 cells were treated with 10% PEG-3350 for 24 h. Cells were subjected to flow cytometric analysis to measure EGFR and Snail levels as described in the “Material and Methods” section. As shown in Panel B, PEG caused a significant decrease in both EGFR and Snail levels (p<0.001) in the parental cells. However, reductions by PEG were blunted for both EGFR and Snail in shRNA-EGFR cells.

DISCUSSION

We extend our previous finding that PEG suppresses epithelial proliferation via modulation of Snail/β-catenin signaling by identifying upstream regulators of this process. In the present study, we for the first time demonstrate an essential role of EGFR downregulation in the PEG mediated antiproliferative and hence chemopreventive effects. The present study provides evidence that PEG downregulates membrane EGFR possibly via lysosomal degradation and this event may control the Snail/β-catenin pathway. The significance of EGFR in PEG chemoprevention is further highlighted by our results indicating that not only is EGFR downregulated by PEG in the premalignant mucosa but by blocking EGFR in cell culture there is a marked decrease in the antiproliferative efficacy of PEG.

Previous studies have suggested that PEG is a remarkably potent chemopreventive agent with effects seen throughout the spectrum of carcinogenesis. Specifically, PEG has been shown to cause regression of established lesions such as ACF (33) and also inhibit the earliest stages of colon carcinogenesis including at the pre-dysplastic mucosa (12). It is well established that there is an increased epithelial proliferation in the histologically normal mucosa of the AOM-treated rats prior to development of neoplasia. Importantly, in humans, analysis of the proliferation in the histologically normal mucosa can predict which patients harbor neoplasia elsewhere in the colon (34). The proliferation rate has been used as a biomarker in several chemopreventive studies.

Proliferation in the colonic mucosa is governed by a numerous genetic and epigenetic events. Arguably the most important is dysregulation of the β-catenin signaling cascade which, by transactivation of Tcf-1/Lef-1, induces transcription of myriad of genes responsible for proliferation (cyclin D1, c-myc etc). β-catenin signaling can be augmented through increased protein stability (via mutations in adenomatous polyposis coli or in CTNNB1) or its altered cellular distribution (35, 36). The latter occurs through loss of E-cadherin, a membrane protein that avidly binds β-catenin to the cell membrane and hence away from the nucleus (37). The importance of E-cadherin is underscored by the observation that its loss triggers increased tumor initiation in the APC-driven murine model of intestinal tumorigenesis (38). Moreover, several reports have indicated that E-cadherin induction may be important mechanism for chemoprevention with agents such as COX-2 inhibitors, NSAIDS, ursodeoxycholic acid (18, 24). Thus corroborating our earlier findings that PEG induced E-cadherin was biologically plausible as a means of chemoprevention (12).

We have recently demonstrated that the induction of E-cadherin by PEG is related to downregulation of the transcriptional repressor Snail (39). We and others demonstrated that Snail was overexpressed in CRCs (32, 40). Importantly, we demonstrated that using antisense oligonucleotides to Snail induced E-cadherin expression in the uninvolved MIN mouse mucosa lead to decreased proliferation and tumorigenesis (20). Based on those studies we presented a paradigm that in the uninvolved mucosa, PEG leads to↓Snail→ ↑E-cadherin→ ↓ β-catenin transcriptional activity → ↓ cell proliferation→ ↓tumorigenesis (Figure 5).

Figure 5.

This figure illustrates the proposed model for the role of EGFR in the chemopreventive activity of PEG. EGFR downregulation leads to decrease in Snail expression with a consequent decrease in β-catenin signaling and consequently suppression of proliferation. PEG may downregulate EGFR by lysosomal degradation.

One challenge of this model is linking the bulky PEG molecule with Snail, generally found in the nucleus. Intuitively, given the size of PEG, it would appear more probable that PEG is controlling Snail levels via membrane-receptor events rather than directly accessing the nucleus. The candidate cell surface target of PEG would need to meet several criterion such as: 1) should be able to regulate Snail expression, 2) should be overexpressed early in colon carcinogenesis (that is when PEG begins to show effectiveness) and 3) its inhibition must lead to a profound decrease in tumorigenesis (which is consonant with the effect of PEG). In this context, EGFR seems to meet all these requisite conditions. For instance, Lu and colleagues published a seminal report linking EGFR to Snail regulation in breast cancer cells (21). Our data from the EGFR knockdowns confirmed this by indicating an increase in basal Snail protein. Moreover, Mann et al. demonstrated that inhibition of EGFR signaling blocked Snail and hence Snail-induced degradation of the pro-carcinogenic prostaglandin E2 (22). Similarly, reduced radiotherapy-induced apoptosis in tumors and cancer cell lines expressing high levels of EGFR has been recently linked to altered Snail/slug expression (41).

Secondly, emerging evidence indicates that increased expression of EGFR is not only relevant in established CRCs but also plays an important role early on at the premalignant stage of colon carcinogenesis (42). For instance, increased expression of EGFR has been reported to be involved in the development of large human ACF (27) and formation of microadenoma in an animal model of colon cancer (28). Our present study clearly demonstrates EGFR upregulation in the histologically normal mucosa of the AOM-treated rat at a pre-dysplastic time-point as a marker of field carcinogenesis. The relevance to human disease is supported by the observation that EGFR activation in the histologically normal appearing rectal mucosa is a marker of neoplasia in the proximal colon (43). Thus, EGFR clearly meets the relevance/timing criteria as a molecular target of PEG.

Thirdly, biological importance of EGFR in CRC is underscored by the demonstration that direct targeting of EGFR, achieved with either monoclonal antibodies or pharmacological inhibitors (44, 45), is a effective strategy in treating CRC. With regards to prevention of neoplasia by the EGFR inhibitors, EKB-569 caused a profound reduction in the polyp number in the APC-MIN mouse model (45). Similarly, EGFR inhibitor gefitinib reduced tumorigenesis in the AOM-treated rat model by 69% with a concomitant reduction in ACF (46). It is therefore; clear that EGFR is a plausible candidate PEG target during CRC chemoprevention. Our data shows that PEG treatment decreases EGFR expression in vivo (AOM-treated rat) and in vitro (HT-29 cells). Furthermore, we show that EGFR is critical in PEG chemopreventive effect by demonstrating that a modest decrease of EGFR using Sh-RNA approach in HT-29 markedly blunted the responsiveness of these cells to PEG as assessed by either the anti-proliferative effect or ability to decrease Snail

The mechanism of EGFR downregulation by PEG remains largely unexplored. Our studies rule out regulation at the transcriptional level but suggest post-translational modulation. EGFR protein levels are regulated through a complex process of endocytosis, targeting the vesicles for either destruction via the lysosomes or recycled back to the membrane. The lysosomal inhibitor experiments suggest that PEG induces endocytosis and hence degradation as has previously been reported for sulindac (47). The mechanism by which PEG induces EGFR endocytosis is not clear. EGFR may involve clathrin-dependent pathway that includes early and late endosomes (48), or clathrin-independent pathways that involve membrane invaginations (caveolae) (49). The latter pathway has been implicated in the proto-oncogene decorin regulation of EGFR (49). In addition, mechanisms involving non-caveolar lipid rafts (50) and formation of transient, circular dorsal ruffles or “waves” are some other recently recognized processes that have been shown to internalize EGFR. More studies are planned to elucidate the mechanisms through which PEG treatment results in internalization and degradation of EGFR

In summary, we have described for the first time that downregulation EGFR is an important mechanism for the chemopreventive activity of PEG. EGFR downregulation leads to decrease in Snail expression with a consequent decrease in β-catenin signaling (as we previously described) and hence suppression of proliferation (Figure 5). This study provides fundamental insights into the mechanism of action of PEG and will bolster further clinical studies. In that context, we are embarking on a phase 2 trial of PEG in human colon carcinogenesis. While it is possible that the required chemopreventive dose may lead to increased bowel movement, this would still allow PEG to be used in the 20–30% of the population with chronic constipation (a putative CRC risk factor). Thus, given the safety and efficacy data, PEG has the potential of being an outstanding chemopreventive agent. Moreover, this novel mechanism of action may allow development of inexpensive, non-toxic topical agents that could target EGFR-dependent malignancies.