Polyethylene glycol-mediated colorectal cancer chemoprevention: roles of epidermal growth factor receptor and Snail

Polyethylene glycol-mediated colorectal cancerchemoprevention: roles of epidermal growthfactor receptor and Snail

Ramesh K. Wali,1 Dhananjay P. Kunte,1

Jennifer L. Koetsier,1 Marc Bissonnette,2

and Hemant K. Roy1

1Evanston Northwestern Healthcare, Evanston, Illinois and2University of Chicago, Chicago, Illinois

AbstractPolyethylene glycol (PEG) is a clinically widely used agentwith profound chemopreventive properties in experimentalcolon carcinogenesis. We reported previously that Snail/B-catenin signaling may mediate the suppression ofepithelial proliferation by PEG, although the upstreamevents remain unclear. We report herein the role ofepidermal growth factor receptor (EGFR), a knownmediator of Snail and overepressed in f80% of humancolorectal cancers, on PEG-mediated antiproliferative andhence antineoplastic effects in azoxymethane (AOM) ratsand HT-29 colon cancer cells. AOM rats were randomizedto either standard diet or one with 10% PEG-3350 andeuthanized 8 weeks later. The colonic samples were sub-jected to immunohistochemical or Western blot analyses.PEG decreased mucosal EGFR by 60% (P < 0.001).Similar PEG effects were obtained in HT-29 cells. PEGsuppressed EGFR protein via lysosmal degradation with nochange in mRNA levels. To show that EGFR antagonismper se was responsible for the antiproliferative effect,we inhibited EGFR by either pretreating cells with gefitinibor stably transfecting with EGFR-short hairpin RNA andmeasured 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-short hairpinRNA cells, besides having reduced membrane EGFR, alsoexpressed low Snail levels (40%), corroborating a strong

association. Furthermore, in EGFR silenced cells, PEGeffect on EGFR or Snail was muted, similar to that onproliferation. In conclusion, we show that EGFR is theproximate membrane signaling molecule through whichPEG initiates antiproliferative activity with Snail/B-cateninpathway playing the central intermediary function. [MolCancer Ther 2008;7(9):3103–11]

IntroductionColorectal cancer is second most common cause of cancer-related deaths in the United States with 49,960 deathsestimated in 2008 (1). Because early-stage colorectal cancer(curable) is generally clinically silent, screening the entireat-risk (age >50) asymptomatic population has beenadvocated. However, the power of tests such as colono-scopy in reducing mortality and even incidence ofcolorectal cancer is juxtaposed with the reticence of theasymptomatic population to undergo invasive screeningtests. From a patient preference perspective, chemopreven-tion represents an attractive approach for colorectal cancerprevention with numerous agents showing efficacy inepidemiologic and preclinical studies. However, the resultsof randomized controlled studies with several promisingagents have been disappointing owing to suboptimalefficacy and/or higher toxicity of the agents used. Forinstance, although the evidence supporting aspirin-relatedcolorectal cancer inhibition is convincing (2, 3), the U.S.Preventive Service Task Force concludes that the risk oftoxicity with aspirin outweighs it chemopreventive benefit(4). This was the rationale for the employment of cyclo-oxygenase-2 inhibitors, which cause less gastrointestinaltoxicity. However, the randomized controlled trial whileshowing efficacy also noted a 2- to 3-fold increase in cardiactoxicity (5, 6). Furthermore, although the epidemiologicevidence for a protective role of folate seems compelling,recent clinical data suggest that folic acid supplementationmay enhance tumor progression (7). Likewise, dietarycalcium supplementation by and large seems to be safe, butits chemopreventive efficacy is modest (10-15% riskreduction; ref. 8). Thus, safe and more effective chemo-preventive agents are urgently needed.Over the past several years, Corpet et al. have indicated

that polyethylene glycol (PEG) has remarkable efficacy as achemopreventive agent (9, 10). Indeed, the ability of thisnovel agent to suppress tumors or aberrant crypt fociin the azoxymethane (AOM)-treated rat model was >90%,generally outperforming reported efficacies of nonsteroidalanti-inflammatory drugs or that of other known chemo-preventive agents (11). Our laboratory has both confirmedthese findings (12) and extended it to the MIN mouse,another well-validated model of experimental colon

Received 5/5/08; revised 6/13/08; accepted 6/18/08.

Grant support: NIH grants R42CA130508, 5U01CA111257, andRO3CA119261.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.

Note: Presented in part at the 108th Annual American GastroenterologyAssociation Meeting, Washington, DC, 2007.

Requests for reprints: Ramesh K. Wali, Feinberg School of Medicine atNorthwestern University, Department of Internal Medicine, EvanstonNorthwestern Healthcare, 1001 University Place, Suite 314, Evanston,IL 60201. Phone: 224-364-7657. Fax: 847-733-5451.E-mail: [email protected]

Copyright C 2008 American Association for Cancer Research.

doi:10.1158/1535-7163.MCT-08-0434

3103

Mol Cancer Ther 2008;7(9). September 2008

Research. on February 27, 2016. © 2008 American Association for Cancermct.aacrjournals.org Downloaded from

http://mct.aacrjournals.org/

carcinogenesis (13). Preliminary epidemiologic data sup-port the potential of PEG to suppress adenomas in humans(14). This is coupled with the excellent safety data from thelong-term use of PEG as a cathartic agent (15). Indeed, PEGis 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 ofunderstanding of its mechanism of action. During earlycolon carcinogenesis, diffuse epithelial hyperproliferationalong with suppression of apoptosis in the colonic mucosais a key event that is often driven by alterations in theh-catenin signaling pathway (16). Down-regulation ofh-catenin signaling, with a consequent suppression ofproliferation, is a hallmark of a variety of chemopreventiveagents including nonsteroidal anti-inflammatory drugs,green tea, etc. (17, 18). PEG has been shown previously tosuppress colonic epithelial proliferation (12, 19). In both cellculture and AOM-treated rat model, our laboratory hasfurther shown that the antiproliferative effect of PEG wasaccompanied by marked suppression of Snail, leading toinduction of E-cadherin and subsequent sequestration ofh-catenin away from the nucleus (12). This leads todecreased h-catenin-dependent transcriptional activityand cyclin D1 expression, an important regulator of cellularproliferation (12). The relevance of Snail in tumorigenesis isevident from our earlier studies showing that selectivedown-regulation of this transcriptional repressor, usingantisense phosphorodiamidate morpholino oligomer,decreased epithelial proliferation in the normal histologicmucosa with a concomitant reduction in intestinal tumori-genesis in the MIN mouse model (20). Thus, there iscompelling rationale to assume that inhibition of Snailsignaling 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. Wereasoned that PEG may likely be influencing Snail-relatedsignaling pathways by interacting with upstream targets atthe cell membrane level. In this context, epidermal growthfactor receptor (EGFR), a transmembrane glycoprotein, hasbeen implicated in the control of intracellular Snail levels(21, 22). Moreover, EGFR is overexpressed in mostcolorectal cancers (23), making it an excellent candidatefor a cell surface-based PEG target. In the present studies,we show that PEG down-regulates EGFR in both AOM-treated rat and colon cancer HT-29 cells. This down-regulation appears to be a result of increased endocyticlysosmal degradation. Moreover, using a short hairpinRNA (shRNA) approach, we show that EGFR down-regulation is central to PEG responsiveness with regardsto its role in antiproliferation and Snail regulation.

Materials andMethodsExperimental Animal ProtocolsAll animal studies were done in accordance with the

Institutional Animal Care and Use Committee of Evanston-Northwestern Healthcare. Twenty-four male Fisher 344 rats

(100-120 g; Harlan) were maintained on defined (AIN-76A)diet for 2 weeks and then randomized into three equalgroups. Group 1 was injected with saline (AOM vehicle)and groups 2 and 3 with AOM 20 mg/kg body weight/wkfor 2 weeks (i.p.). Two weeks later, group 3 rats wereswitched to a PEG-3350-supplemented diet (10 g/100 gdiet; Harlan Teklad) and continued until sacrifice 8 weeksafter AOM. In another experiment, to study the short-termdose effect of PEG-3350 on EGFR expression, 16 additionalAOM-treated rats were randomized to receive 5% or 10%gavages of PEG (w/v) for 1 week before euthanizing. Ratswere provided water ad libitum and housed in polycar-bonate cages in an environmentally controlled room (daily12-h fluorescent light and dark cycles at 24jC and a relativehumidity of 70%). All rats were euthanized in a nonfastedstate and the colons were isolated and flushed with PBS(pH 7.4). Whereas small distal segments were removed andfixed in formalin for immunohistochemical studies, thereminder of the colon was opened longitudinally to exposethe luminal mucosa. The colonic mucosa was collected bygentle scrapping using glass slides, homogenized inLaemmli buffer, and subjected to Western blot analysis asdescribed previously (24).

Immunohistochemical StainingFormalin-fixed colonic tissue sections were examined by

immunostaining to assess changes in the expression levelsof Snail and EGFR. Briefly, paraffin-embedded colonicsegments were sliced (4 Am thick) along the vertical axis ofthe crypts and the sections were mounted on Vectabond-coated Superfrost + slides. These slides were then bakedfor 1 h at 60jC to 70jC, deparaffinized in xylene, andrehydrated by graded ethanol washes. The antigen epitoperetrieval for Snail and EGFR were achieved by pressuremicrowaving (NordicWare) in antigen unmasking solutionfor 9 min � 2 cycles (Vector Laboratories). Endogenousperoxidase 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 incubatedovernight (at 4jC) with appropriate primary antibodies[anti-Snail (SNAI1; T-18, 1:100) and anti-EGFR (1:150; SantaCruz Biotechnology)] followed by incubation with suitablebiotinylated secondary antibodies (1:2,000). After washing,the samples were incubated with avidin-biotin peroxidasecomplex using Vectastatin Elite ABC kit (Vector Laborato-ries) and the stain developed with DAB (Vector Laborato-ries). Only complete longitudinal crypts, extending fromthe muscularis mucosa to colonic lumen, were included forimmunohistochemical analysis (8 crypts per colon and 6rats in each group). Staining intensity was quantified on afive-point scale by a gastrointestinal pathologist blinded tothe treatment group.

Cell Culture and PEG TreatmentThe human colon cancer cell line HT-29 (American Type

Culture Collection) was cultured in McCoy’s 5A mediumwith 10% fetal bovine serum. The cells were seeded in100 mm Petri dishes (105 cells/mL), washed twice withPBS, and serum starved (0.5% fetal bovine serum) for72 h before treating with PEG for the indicated time. Cells

EGFR/Snail and PEG-Colorectal Cancer Chemoprevention3104




were then harvested and subjected to flow cytometricanalysis, Western blotting, and reverse transcription-PCR.

Flow CytometryAnalysesThis technique was employed to measure the surface

expression of EGFR as well as levels of intracellular Snail inHT-29 cells treated with PEG. Cells (70% confluent) weretreated with PEG as described above and then fixed in 4%buffered paraformaldehyde for 30 min. For EGFR surfacestaining, the cells were washed twice in a PBS (pH 7.4)containing 2% fetal bovine serum, 0.2% bovine serumalbumin, and 0.02% sodium azide; for intracellular stainingof Snail, 0.2% saponin was also included in this buffer.Cells were later incubated in their respective PBS bufferswith either anti-EGFR 528 (1:200; kindly provided by Dr. H.Band) or anti-Snail (T-18; 1:200) antibodies at roomtemperature for 1 h. After washing three times with theirrespective buffers, cells were incubated with secondaryantibody Alexa 488 Green FI1-labeled anti-goat (for Snail)or anti-mouse (for EGFR; Invitrogen) for 40 min. Cells weresubsequently washed three times with their respectivebuffers and subjected to flow cytometric analysis (BectonDickinson Labware).

Western Blot AnalysisWestern blotting was done using standard techniques as

described previously (24). Briefly, samples were preparedfor SDS-PAGE analysis by adding Laemmli sample buffer

to clear whole-cell lysates and then heated for 5 min at95jC before loading. An equal amount (40 Ag) of the proteinsamples from rat colonic mucosa or HT-29 cells weresubjected to SDS-polyacrylamide gel electrophoresis, trans-ferred to polyvinylidene difluoride membranes (Millipore),blocked with 5% nonfat milk, and probed with appropriateantibodies [anti-EGFR, 1:200; anti-proliferating cell nuclearantigen (PCNA), 1:250] using standard techniques. Xero-grams were developed with enhanced chemiluminescence(Santa Cruz Biotechnology) and image analysis was doneusing image acquisition analysis software (Labworks, 4.6;UVP). Expression levels were normalized to the levels ofh-actin or a-tubulin as controls after stripping and reprobingwith anti-h-actin (1:300) or anti-a-tubulin (1:200) antibody.Reverse Transcription-PCRHT-29 cells were treated with 10% PEG-3350 for 24 h and

RNA was extracted with TRI Reagent (Sigma) as describedpreviously (25). The cDNA was synthesized using 5 AgRNA and Superscript RT (Invitrogen Life Technologies).Amplification of Snail mRNA was done using nested PCRprotocols (26). Cyclophilin was used as a control for RNAloading (25).

Cell Proliferation AssayTo show that EGFR modulation by PEG is important in

its antiproliferative activity, we pretreated HT-29 cells witha known EGFR inhibitor to study if it blunts the

Figure 1. Effect of PEG on AOM-induced rat colonic mucosal EGFR protein expression. For these experiments, we used the AOM rat model as describedin Materials and Methods. The rats were sacrificed 8 wk of AOM/saline and colonic mucosa was scrapped and subjected to SDS-PAGE electrophoresis andEGFR expression was determined by Western blot analysis.A, there was a significant induction of EGFR (P <0.001) expression at premalignant time point(8 wk of AOM treatment) compared with saline-treated controls. On the other hand, in AOM-treated rats, PEG-3350 (10 g/100 g diet for 8 wk)supplementation remarkably decreased the EGFR expression (by 60%; P < 0.001, compared with AOM alone) to almost saline-treated control levels. B, ina separate experiment, to study if relatively shorter PEG treatments could also achieve EGFR attenuation, 7 wk post-AOM rats were gavaged with either5% or 10% (w/v) PEG-3350 for 1 wk before sacrifice. Only a week of oral PEG gavages to these animals produced astounding decline in EGFR expression.Whereas 5% PEG showed a trend, 10% PEG dose caused a significant (*, P < 0.001) decrease in EGFR levels. Housekeeping h-actin was used as aloading control for determining protein expression.

Molecular Cancer Therapeutics 3105




antiproliferative activity of PEG. For this experiment, HT-29 cells were pretreated with EGFR inhibitor gefitinib(25 Amol/L; Iressa; AstraZeneca) for 2 h at 37jC beforetreating with 10% PEG-3350 for 24 h. For control, cells werepretreated with equal volume of DMSO (gefitinib vehicle)and then with PEG. For these assays, cells were seeded in96-well plates. At the end of the incubation, the mediumwas replaced with fresh medium (100 AL) containing 5 AL4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disufonate (WST-1) reagent (Roche Diagnostics).After 5 to 15 min incubation, the absorbance was read at440 nm in a Spectramax Plus Spectrophotometer platereader (Molecular Devices).

EGFRKnockdown AssayTo further explore the importance of EGFR in PEG-

mediated antiproliferation, we studied the effect of PEG oncellular proliferation in EGFR-deficient HT-29 cells. Weknocked down EGFR gene expression in these cells usingshRNA (Origene). The EGFR-shRNA and control vectorswere transfected in HT-29 cells using LipofectAMINE 2000(Invitrogen) according to the manufacturer’s instructions.After transfection, cells were incubated at 37jC in a

humidified 5% CO2 incubator for 72 h. Transfectants wereselected by puromycin (0.5 Ag/mL). The cells were thenincubated with 10% PEG for 24 h before assayingexpression of EGFR and PCNA.

Statistical MethodsValues were expressed as mean F SE as indicated.

Quantitative densitometry values were compared byunpaired Student’s t test. Differences with P < 0.05 wereconsidered statistically significant.

ResultsPEGDown-regulatesColonic Epithelial EGFRinAOM-

Induced RatsPEG supplementation was well tolerated by the rats with

no apparent toxicity as reflected by normal body weightgain as reported previously (12). Because cellular hyper-proliferation has been found to be associated with over-expression of EGFR in the premalignant colonic mucosa ofhumans as well as AOM rat model of colon cancer (27, 28),we reasoned if the ability of PEG to inhibit proliferation(12) is mediated via modulation of the mucosal EGFR

Figure 2. Mechanism of EGFR down-regulation in HT29 cells treated with PEG-3350. The human colon cancer cells, HT29, were seeded in 100 mm Petridishes (105/mL) using McCoy’s 5A medium containing 10% fetal bovine serum. After 24 h, the cells were washed off the medium with PBS before furtherincubation in serum-starved medium (0.5% fetal bovine serum) for 72 h and then treated with 5% or 10% PEG for 6 or 24 h. A, as shown in Western blotanalysis, both 5% and 10% PEG treatment for 6 or 24 h inhibited EGFR expression; however, the effect was statistically significant at 10% dose for bothshort (P < 0.05) and long (P < 0.001) exposure times. B, similar effects on membrane EGFR inhibition (55%) could be ascertained by flow cytometricanalysis. C, further as depicted in the representative reverse transcription-PCR blot, PEG exposure did not cause any significant differences in the EGFRmRNA levels. To further study if PEG may induce EGFR down-regulation by inducing lysosomal degradation, cells were pretreated with 250 nmol/Lbafilomycin A1 (BFL ; specific lysosomal inhibitor) for 2 h before treating with 10% PEG for 24 h. D, block in lysosomal degradation almost completelyprevented PEG to inhibit EGFR.





expression. For these experiments, we used the AOM ratmodel, and as shown in Fig. 1A (representative WesternBlot), there was a significant induction of EGFR expression(P < 0.001) at premalignant time point (8 weeks post-AOMtreatment) compared with saline-treated controls. Incontrast, PEG supplementation in AOM-treated rats (10 gPEG-3350/100 g diet) significantly decreased EGFR expres-sion by 60% (P < 0.001) almost to the control levels ofsaline-treated rats. In another AOM rat experiment (Fig. 1B)to study if relatively shorter PEG treatments could alsodown-regulate EGFR, 7 weeks post-AOM-treated rats weregavaged with vehicle or 5% or 10% (w/v) PEG-3350 for 1week before sacrifice. As shown, only a week of oral PEGgavages caused significant decline in EGFR expression.Whereas 5% PEG showed a trend, 10% PEG caused ahighly significant (P < 0.001) decrease in EGFR levels.

PEG Inhibits EGFRExpression in HT-29 CellsBecause we showed previously that PEG inhibits

proliferation of human colon cell line HT-29, we investi-gated if PEG down-regulated EGFR expression in thesecells. As depicted by a representative Western blot inFig. 2A, there was a dose- and time-dependent inhibition ofEGFR protein expression in these cells. Both 5% and 10%PEG treatments for 6 or 24 h inhibited EGFR expression;however, the upper dose (shown previously to beantiproliferative in these cells; ref. 12) caused a significantlygreater effect (P < 0.05 at 6 h and P < 0.001 at 24 h).Analysis of EGFR expression by flow cytometric analysisconfirmed these results (Fig. 2B). We further investigated ifthe effect of PEG occurred at the translational or transcrip-tional level. As can be seen in a representative reversetranscription-PCR blot (Fig. 2C), PEG (10% for 24 h)treatment did not cause any significant difference in themRNA levels in these cells. This suggests that the down-regulation of EGFR by PEG may be a result of the post-transcriptional modulation of the receptor.

PEG Down-regulates EGFR via Lysosomal Degrada-tion in HT-29 CellsEGFR is recognized to be down-regulated by receptor

internalization from the plasma membrane into endosomesand other cytoplasmic vesicles where it is degraded bylysosomal/proteosomal processes and may undergo ubi-quitination (29). To study the role of lysosomal degradationin PEG induced decrease in EGFR expression, HT-29 cellswere pretreated with specific lysosomal inhibitor, bafilo-mycin A1 (Calbiochem) before treatment with PEG for 24 h.The levels of EGFR expression was analyzed by Westernblotting. As shown in Fig. 2D, PEG in the absence of theinhibitor resulted in expected decrease in EGFR expression.When the cells were pretreated with bafilomycin for 2 h at37jC, there was a small stabilizing effect on the EGFR;however, under these conditions, the ability of PEG todown-regulate EGFR was almost completely lost. There issome evidence to suggest that EGFR degradation could alsobe blocked by proteosomal inhibitors (29). In our studies,when we used lactacystin (Calbiochem) to inhibit proteo-somal degradation, there was only a nonsignificant partialblock (by 17%) in EGFR degradation by PEG. Taken

together, these studies suggest that EGFR down-regulationby PEG may be a consequence of lysosomal degradationof EGFR and only partially facilitated by proteosomaldegradation.

EGFR Antagonist (Gefitinib) Blocks PEG-InducedAntiproliferativeActivityTo establish the relevance of EGFR as the primary

membrane target of PEG-mediated antiproliferative effects,we investigated if gefitinib, a known EGFR antagonist,could blunt the effect of PEG on HT-29 cell proliferation.The cells were pretreated either with DMSO (as vehicle;0.05%) or with 25 Amol/L gefitinib for 120 min beforeincubating with or without PEG for 24 h at 37jC.Proliferation was assayed using WST-1 reagent. As shownin Fig. 3A, in control (DMSO) treated cells, PEG causeda significant decrease in rate of proliferation; however,

Figure 3. Causal relationship between PEG induced EGFR down-regulation and its antiproliferative activity. HT29 cells were eitherpretreated with well-known EGFR kinase inhibitor, gefitinib, or transfectedwith EGFR shRNA transfections as discussed in Materials and Methods.Serum-starved cells in 96-well plates were pretreated with DMSO or25 Amol/L gefitinib for 120 min before PEG treatment for 24 h and thenassayed for proliferation using standard WST-1 assay. A, whereas inDMSO-treated cells PEG caused more than 50% decrease in proliferationrate, this effect was abolished in cells pretreated with EGFR inhibitorgefitinib. B, shRNA that caused f50% inhibition in EGFR expression asmeasured by Western blotting, blunted the PEG effect on PCNA (C), amarker of cellular proliferation.





in gefitinib-treated cells, the antiproliferative effect of PEGwas lost. These studies strongly suggest that the antipro-liferative effects of PEG may indeed be mediatedby its effect on the EGFR down-regulation.

EGFR Gene Knockdown in HT-29 Cells Blunts theAntiproliferativeActivity of PEGTo further establish a direct causal relationship between

EGFR down-regulation and antiproliferative activity ofPEG, we determined if specific EGFR knockdown couldblunt the effect of PEG on the expression of PCNA in HT-29cells. The expression of PCNA is known to be stimulated byEGFR (27) and is widely used as a marker of cellproliferation (30, 31). As shown in Fig. 3B, transfection ofEGFR shRNA into HT-29 cells caused 50% reduction in theexpression of EGFR compared with vector-transfected cellsas control. Whereas in control cells PEG (10% for 24 h)caused 60% inhibition in the PCNA expression (Fig. 3C), inEGFR knockdown cells it produced only 25% inhibition.Taken together, these studies provide evidence that EGFRdown-regulation may be an important mediator of theantiproliferative effects of PEG in colon cancer cells.

EGFR, an Upstream Effecter of PEG-Induced Inhibi-tion ofTranscriptional Repressor SnailOur laboratory has shown previously that Snail is

activated in colon cancer (32) and PEG treatment inhibitsit (12). This down-regulation of Snail leads to increasedplasma membrane E-cadherin, which in turn limitsh-catenin transcriptional activity resulting in cyclin D1inhibition and hence decreased proliferation (12). BecauseEGFR is also known to induce Snail (21) and increase cyclinD1 (29), we tested if this membrane receptor may bemodulated by PEG and mediate its antiproliferative effects

downstream of the Snail/h-catenin pathway. As depictedin Fig. 4A (representative immunohistochemical images ofcolonic sections from AOM-treated rats with or withoutPEG-3350 supplementation), PEG clearly reduced theexpression of Snail as well as that of EGFR. To show acausal relationship between PEG-induced EGFR and Snailexpression, we used the EGFR shRNA knockdownapproach to study the effect of PEG on both EGFR andSnail 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 cytometricanalysis to measure EGFR and Snail levels as describedin ‘‘Materials and Methods.’’ As shown in Fig. 4B, PEGsignificantly decreased both EGFR and Snail levels (P <0.001) in the control cells. However, the effect of PEGwas blunted for both EGFR and Snail in shRNA-EGFR-transfected cells. These studies provide strong evidencethat the effect of PEG on Snail and related antiproliferativesignaling may be initiated from its upstream modulationof membrane EGFR.

DiscussionWe extend our previous finding that PEG suppressesepithelial proliferation via modulation of Snail/h-cateninsignaling by identifying upstream regulators of thisprocess. In the present study, we show, for the first time,an essential role of EGFR down-regulation in the PEG-mediated antiproliferative and hence chemopreventiveeffects. The present study provides evidence that PEGdown-regulates membrane EGFR possibly via lysosomaldegradation and this event may control the Snail/h-catenin

Figure 4. A causal relationship between PEG-mediated reduced expression of EGFR and Snail. Because EGFR is known to induce Snail, which we havereported previously to be inhibited by PEG, we tested if this receptor may be the upstream membrane target for PEG to interact and initiate antiproliferativesignaling via Snail. 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 show a relationship between EGFR and Snail expression, we transfected HT-29cells 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 with10% PEG-3350 for 24 h. Cells were subjected to flow cytometric analysis to measure EGFR and Snail levels as described in Materials and Methods. B, PEGcaused 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 andSnail in shRNA-EGFR cells.





pathway. The significance of EGFR in PEG chemopreven-tion is further highlighted by our results indicating thatnot only is EGFR down-regulated by PEG in the pre-malignant mucosa, but by blocking EGFR in cell culturethere is a marked decrease in the antiproliferative efficacyof PEG.Previous studies have suggested that PEG is a remark-

ably potent chemopreventive agent with effects seenthroughout the spectrum of carcinogenesis. Specifically,PEG has been shown to cause regression of establishedlesions such as aberrant crypt foci (33) and also inhibit theearliest stages of colon carcinogenesis including at thepredysplastic mucosa (12). It is well established that there isan increased epithelial proliferation in the histologicallynormal mucosa of the AOM-treated rats before develop-ment of neoplasia. Importantly, in humans, analysis of theproliferation in the histologically normal mucosa canpredict which patients harbor neoplasia elsewhere in thecolon (34). The proliferation rate has been used as abiomarker in several chemopreventive studies.Proliferation in the colonic mucosa is governed by a

numerous genetic and epigenetic events. Arguably, themost important is dysregulation of the h-catenin signalingcascade, which, by transactivation of Tcf-1/Lef-1, inducestranscription of myriad of genes responsible for prolifera-tion (cyclin D1, c-myc, etc.). h-Catenin signaling can beaugmented through increased protein stability (via muta-tions in adenomatous polyposis coli or CTNNB1) or itsaltered cellular distribution (35, 36). The latter occursthrough loss of E-cadherin, a membrane protein that avidlybinds h-catenin to the cell membrane and hence away fromthe nucleus (37). The importance of E-cadherin is under-scored by the observation that its loss triggers increasedtumor initiation in the adenomatous polyposis coli-drivenmurine model of intestinal tumorigenesis (38). Moreover,several reports have indicated that E-cadherin inductionmay be important mechanism for chemoprevention withagents such as cyclooxygenase-2 inhibitors, nonsteroidalanti-inflammatory drugs, and ursodeoxycholic acid (18,24). Thus, corroborating our earlier findings that PEGinduced E-cadherin was biologically plausible as a meansof chemoprevention (12).We have shown recently that the induction of E-cadherin

by PEG is related to down-regulation of the transcriptionalrepressor Snail (39). We and others have shown that Snailwas overexpressed in colorectal cancers (32, 40). Impor-tantly, we showed that using antisense oligonucleotides toSnail induced E-cadherin expression in the uninvolvedMIN mouse mucosa lead to decreased proliferation andtumorigenesis (20). Based on those studies, we presented aparadigm that, in the uninvolved mucosa, PEG leadsto#Snail!"E-cadherin!#h-catenin transcriptional activi-ty!#cell proliferation!#tumorigenesis (Fig. 5).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 moreprobable that PEG is controlling Snail levels via membranereceptor events rather than directly accessing the nucleus.

The candidate cell surface target of PEGwould need to meetseveral criterion such as (a) should be able to regulate Snailexpression, (b) should be overexpressed early in coloncarcinogenesis (that is when PEG begins to show effective-ness), and (c) its inhibition must lead to a profound decreasein tumorigenesis (which is consonant with the effect ofPEG). In this context, EGFR seems to meet all these requisiteconditions. For instance, Lu et al. published a seminal reportlinking EGFR to Snail regulation in breast cancer cells (21).Our data from the EGFR knockdowns confirmed this byindicating an increase in basal Snail protein. Moreover,Mann et al. showed that inhibition of EGFR signalingblocked Snail and hence Snail-induced degradation of theprocarcinogenic prostaglandin E2 (22). Similarly, reducedradiotherapy-induced apoptosis in tumors and cancer celllines expressing high levels of EGFR has been recentlylinked to altered Snail/slug expression (41).Secondly, emerging evidence indicates that increased

expression of EGFR is not only relevant in establishedcolorectal cancers but also plays an important role early onat the premalignant stage of colon carcinogenesis (42). Forinstance, increased expression of EGFR has been reportedto be involved in the development of large human aberrantcrypt foci (27) and formation of microadenoma in an animalmodel of colon cancer (28). Our present study clearly showsEGFR up-regulation in the histologically normal mucosaof the AOM-treated rat at a predysplastic time point as amarker of field carcinogenesis. The relevance to humandisease is supported by the observation that EGFRactivation in the histologically normal-appearing rectalmucosa is a marker of neoplasia in the proximal colon (43).Thus, EGFR clearly meets the relevance/timing criteria as amolecular target of PEG.Thirdly, biological importance of EGFR in colorectal

cancer is underscored by the demonstration that directtargeting of EGFR, achieved with either monoclonal anti-bodies or pharmacologic inhibitors (44, 45), is a effectivestrategy in treating colorectal cancer. With regards toprevention of neoplasia by the EGFR inhibitors, EKB-569

Figure 5. Proposed model for the role of EGFR in the chemopreventiveactivity of PEG. EGFR down-regulation leads to decrease in Snailexpression with a consequent decrease in h-catenin signaling andconsequently suppression of proliferation. PEG may down-regulate EGFRby lysosomal degradation.





caused a profound reduction in the polyp number in theAPC(Min) mouse model (45). Similarly, EGFR inhibitorgefitinib reduced tumorigenesis in the AOM-treated ratmodel by 69% with a concomitant reduction in aberrantcrypt foci (46). It is therefore clear that EGFR is a plausiblecandidate PEG target during colorectal cancer chemo-prevention. Our data show that PEG treatment decreasesEGFR expression in vivo (AOM-treated rat) and in vitro(HT-29 cells). Furthermore, we show that EGFR is critical inPEG chemopreventive effect by showing that a modestdecrease of EGFR using shRNA approach in HT-29markedly blunted the responsiveness of these cells toPEG as assessed by either the antiproliferative effect or theability to decrease Snail.The mechanism of EGFR down-regulation by PEG

remains largely unexplored. Our studies rule out regula-tion at the transcriptional level but suggest post-transla-tional modulation. EGFR protein levels are regulatedthrough a complex process of endocytosis, targeting thevesicles for either destruction via the lysosomes or recycledback to the membrane. The lysosomal inhibitor experi-ments suggest that PEG induces endocytosis and hencedegradation as has been reported previously for sulindac(47). The mechanism by which PEG induces EGFRendocytosis is not clear. EGFR may involve clathrin-dependent pathway that includes early and late endosomes(48), or clathrin-independent pathways that involve mem-brane invaginations (caveolae; ref. 49). The latter pathwayhas been implicated in the proto-oncogene decorin regula-tion of EGFR (49). In addition, mechanisms involvingnoncaveolar lipid rafts (50) and formation of transient,circular dorsal ruffles or ‘‘waves’’ are some other recentlyrecognized processes that have been shown to internalizeEGFR. More studies are planned to elucidate the mecha-nisms through which PEG treatment results in internaliza-tion and degradation of EGFR.In summary, we have described for the first time that

down-regulation EGFR is an important mechanism for thechemopreventive activity of PEG. EGFR down-regulationleads to decrease in Snail expression with a consequentdecrease in h-catenin signaling (as we described previous-ly) and hence suppression of proliferation (Fig. 5). Thisstudy provides fundamental insights into the mechanism ofaction of PEG and will bolster further clinical studies. Inthat context, we are embarking on a phase II trial of PEG inhuman colon carcinogenesis. Although it is possible thatthe required chemopreventive dose may lead to increasedbowel movement, this would still allow PEG to be used inthe 20% to 30% of the population with chronic constipation(a putative colorectal cancer risk factor). Thus, given thesafety and efficacy data, PEG has the potential of being anoutstanding chemopreventive agent. Moreover, this novelmechanism of action may allow development of inexpen-sive, nontoxic topical agents that could target EGFR-dependent malignancies.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Acknowledgments

We thank Drs. Hamid Band and Sri Kumar Raja for the generous gift of theanti-EGFR 528 antibody and the technical help with flow cytometricanalysis of EGFR.

References

1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008. CA Cancer JClin 2008;58:71–96.

2. Chan AT, Giovannucci EL, Schernhammer ES, et al. A prospectivestudy of aspirin use and the risk for colorectal adenoma. Ann Intern Med2004;140:157–66.

3. Sandler RS, Halabi S, Baron JA, et al. A randomized trial of aspirin toprevent colorectal adenomas in patients with previous colorectal cancer.N Engl J Med 2003;348:883–90.

4. Dube C, Rostom A, Lewin G, et al. The use of aspirin for primaryprevention of colorectal cancer: a systematic review prepared for the U.S.Preventive Services Task Force. Ann Intern Med 2007;146:365–75.

5. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention ofcolorectal adenomatous polyps. N Engl J Med 2006;355:885–95.

6. Baron JA, Sandler RS, Bresalier RS, et al. A randomized trial ofrofecoxib for the chemoprevention of colorectal adenomas. Gastroentero-logy 2006;131:1674–82.

7. Van Guelpen B, Hultdin J, Johansson I, et al. Low folate levels mayprotect against colorectal cancer. Gut 2006;55:1461–6.

8. Cho E, Smith-Warner SA, Spiegelman D, et al. Dairy foods, calcium,and colorectal cancer: a pooled analysis of 10 cohort studies. J NatlCancer Inst 2004;96:1015–22.

9. Corpet DE, Parnaud G. Polyethylene-glycol, a potent suppressor ofazoxymethane-induced colonic aberrant crypt foci in rats. Carcinogenesis1999;20:915–8.

10. Corpet DE, Parnaud G, Delverdier M, Peiffer G, Tache S. Consistentand fast inhibition of colon carcinogenesis by polyethylene glycol in miceand rats given various carcinogens. Cancer Res 2000;60:3160–4.

11. Corpet DE, Tache S. Most effective colon cancer chemopreventiveagents in rats: a systematic review of aberrant crypt foci and tumor data,ranked by potency. Nutr Cancer 2002;43:1–21.

12. Roy HK, Kunte DP, Koetsier JL, et al. Chemoprevention of coloncarcinogenesis by polyethylene glycol: suppression of epithelial prolifera-tion via modulation of snail/h-catenin signaling. Mol Cancer Ther 2006;5:2060–9.

13. Roy HK, Gulizia J, DiBaise JK, et al. Polyethylene glycol inhibitsintestinal neoplasia and induces epithelial apoptosis in Apc(min) mice.Cancer Lett 2004;215:35–42.

14. Dorval E, Jankowski JM, Barbieux JP, et al. Polyethylene glycol andprevalence of colorectal adenomas. Gastroenterol Clin Biol 2006;30:1196–9.

15. Cleveland MV, Flavin DP, Ruben RA, Epstein RM, Clark GE. Newpolyethylene glycol laxative for treatment of constipation in adults: arandomized, double-blind, placebo-controlled study. South Med J 2001;94:478–81.

16. Wong NA, Pignatelli M. h-Catenin—a linchpin in colorectal carcino-genesis? Am J Pathol 2002;160:389–401.

17. Hao X, Sun Y, Yang CS, et al. Inhibition of intestinal tumorigenesis inApc(min/+) mice by green tea polyphenols (polyphenon E) and individualcatechins. Nutr Cancer 2007;59:62–9.

18. Roy HK, Karolski WJ, Wali RK, Ratashak A, Hart J, Smyrk TC. Thenonsteroidal anti-inflammatory drug, nabumetone, differentially inhibits h-catenin signaling in the MIN mouse and azoxymethane-treated rat modelsof colon carcinogenesis. Cancer Lett 2005;217:161–9.

19. Parnaud G, Corpet DE, Gamet-Payrastre L. Cytostatic effect ofpolyethylene glycol on human colonic adenocarcinoma cells. Int J Cancer2001;92:63–9.

20. Roy HK, Iversen P, Hart J, et al. Down-regulation of SNAIL suppressesMIN mouse tumorigenesis: modulation of apoptosis, proliferation, andfractal dimension. Mol Cancer Ther 2004;3:1159–65.

21. Lu Z, Ghosh S, Wang Z, Hunter T. Downregulation of caveolin-1function by EGF leads to the loss of E-cadherin, increased transcriptionalactivity of h-catenin, and enhanced tumor cell invasion. Cancer Cell 2003;4:499–515.

22. Mann JR, Backlund MG, Buchanan FG, et al. Repression of





prostaglandin dehydrogenase by epidermal growth factor and snailincreases prostaglandin E2 and promotes cancer progression. Cancer Res2006;66:6649–56.

23. O’Dwyer PJ, Benson AB III. Epidermal growth factor receptor-targeted therapy in colorectal cancer. Semin Oncol 2002;29:10–7.

24. Wali RK, Khare S, Tretiakova M, et al. Ursodeoxycholic acid andF(6)-D(3) inhibit aberrant crypt proliferation in the rat azoxymethanemodel of colon cancer: roles of cyclin D1 and E-cadherin. CancerEpidemiol Biomarkers Prev 2002;11:1653–62.

25. Kunte DP, Wali RK, Koetsier JL, et al. Down-regulation of thetumor suppressor gene C-terminal Src kinase: an early event duringpremalignant colonic epithelial hyperproliferation. FEBS Lett 2005;579:3497–502.

26. Yanez-Mo M, Lara-Pezzi E, Selgas R, et al. Peritoneal dialysis andepithelial-to-mesenchymal transition of mesothelial cells. N Engl J Med2003;348:403–13.

27. Cohen G, Mustafi R, Chumsangsri A, et al. Epidermal growth factorreceptor signaling is up-regulated in human colonic aberrant crypt foci.Cancer Res 2006;66:5656–64.

28. Fichera A, Little N, Jagadeeswaran S, et al. Epidermal growth factorreceptor signaling is required for microadenoma formation in the mouseazoxymethane model of colonic carcinogenesis. Cancer Res 2007;67:827–35.

29. Alwan HA, van Zoelen EJ, van Leeuwen JE. Ligand-inducedlysosomal epidermal growth factor receptor (EGFR) degradation ispreceded by proteasome-dependent EGFR de-ubiquitination. J Biol Chem2003;278:35781–90.

30. Einspahr J, Alberts D, Xie T, et al. Comparison of proliferating cellnuclear antigen versus the more standard measures of rectal mucosalproliferation rates in subjects with a history of colorectal cancer andnormal age-matched controls. Cancer Epidemiol Biomarkers Prev 1995;4:359–66.

31. Kubben FJ, Peeters-Haesevoets A, Engels LG, et al. Proliferating cellnuclear antigen (PCNA): a new marker to study human colonic cellproliferation. Gut 1994;35:530–5.

32. Roy HK, Smyrk TC, Koetsier J, Victor TA, Wali RK. The transcriptionalrepressor SNAIL is overexpressed in human colon cancer. Dig Dis Sci2005;50:42–6.

33. Parnaud G, Tache S, Peiffer G, Corpet DE. Polyethylene-glycolsuppresses colon cancer and causes dose-dependent regression ofazoxymethane-induced aberrant crypt foci in rats. Cancer Res 1999;59:5143–7.

34. Akedo I, Ishikawa H, Ioka T, et al. Evaluation of epithelial cellproliferation rate in normal-appearing colonic mucosa as a high-risk markerfor colorectal cancer. Cancer Epidemiol Biomarkers Prev 2001;10:925–30.

35. Clements WM, Lowy AM, Groden J. Adenomatous polyposis coli/h-catenin interaction and downstream targets: altered gene expression ingastrointestinal tumors. Clin Colorectal Cancer 2003;3:113–20.

36. Clements WM, Wang J, Sarnaik A, et al. h-Catenin mutation is afrequent cause of Wnt pathway activation in gastric cancer. Cancer Res2002;62:3503–6.

37. Morin PJ. h-Catenin signaling and cancer. Bioessays 1999;21:1021–30.

38. Smits R, Ruiz P, Diaz-Cano S, et al. E-cadherin and adenomatouspolyposis coli mutations are synergistic in intestinal tumor initiation inmice. Gastroenterology 2000;119:1045–53.

39. Martin TA, Goyal A, Watkins G, Jiang WG. Expression of thetranscription factors snail, slug, and twist and their clinical significance inhuman breast cancer. Ann Surg Oncol 2005;12:488–96.

40. Pena C, Garcia JM, Garcia V, et al. The expression levels of thetranscriptional regulators p300 and CtBP modulate the correlationsbetween SNAIL, ZEB1, E-cadherin and vitamin D receptor in human coloncarcinomas. Int J Cancer 2006;119:2098–104.

41. Come C, Arnoux V, Bibeau F, Savagner P. Roles of the transcriptionfactors snail and slug duringmammarymorphogenesis and breast carcinomaprogression. J Mammary Gland Biol Neoplasia 2004;9:183–93.

42. Salomon DS, Brandt R, Ciardiello F, Normanno N. Epidermal growthfactor-related peptides and their receptors in human malignancies. Crit RevOncol Hematol 1995;19:183–232.

43. Malecka-Panas E, Kordek R, Biernat W, Tureaud J, Liberski PP,Majumdar AP. Differential activation of total and EGF receptor (EGF-R)tyrosine kinase (tyr-k) in the rectal mucosa in patients with adenomatouspolyps, ulcerative colitis and colon cancer. Hepatogastroenterology 1997;44:435–40.

44. Chua YJ, Cunningham D. Recent data with anti-epidermal growthfactor receptor antibodies and irinotecan in colon cancer. Clin ColorectalCancer 2005;5 Suppl 2:S81–8.

45. Torrance CJ, Jackson PE, Montgomery E, et al. Combinatorialchemoprevention of intestinal neoplasia. Nat Med 2000;6:1024–8.

46. Dougherty U, Sehdev A, Cerda S, et al. Epidermal growth factorreceptor controls flat dysplastic aberrant crypt foci development and coloncancer progression in the rat azoxymethane model. Clin Cancer Res 2008;14:2253–62.

47. Pangburn HA, Kraus H, Ahnen DJ, Rice PL. Sulindac metabolitesinhibit epidermal growth factor receptor activation and expression.J Carcinog 2005;4:16.

48. Martinez-Arca S, Bech-Serra JJ, Hurtado-Kuttner M, Borroto A,Arribas J. Recycling of cell surface pro-transforming growth factor-hregulates epidermal growth factor receptor activation. J Biol Chem 2005;280:36970–7.

49. Zhu JX, Goldoni S, Bix G, et al. Decorin evokes protractedinternalization and degradation of the epidermal growth factor receptorvia caveolar endocytosis. J Biol Chem 2005;280:32468–79.

50. Roepstorff K, Thomsen P, Sandvig K, van Deurs B. Sequestration ofepidermal growth factor receptors in non-caveolar lipid rafts inhibits ligandbinding. J Biol Chem 2002;277:18954–60.





2008;7:3103-3111. Mol Cancer Ther Ramesh K. Wali, Dhananjay P. Kunte, Jennifer L. Koetsier, et al. and Snailchemoprevention: roles of epidermal growth factor receptor Polyethylene glycol-mediated colorectal cancer

Updated version

http://mct.aacrjournals.org/content/7/9/3103

Access the most recent version of this article at:

Cited articles

http://mct.aacrjournals.org/content/7/9/3103.full.html#ref-list-1

This article cites 50 articles, 20 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/7/9/3103.full.html#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications


http://mct.aacrjournals.org/content/7/9/3103

http://mct.aacrjournals.org/content/7/9/3103.full.html#ref-list-1

http://mct.aacrjournals.org/content/7/9/3103.full.html#related-urls

http://mct.aacrjournals.org/cgi/alerts

mailto:[email protected]

mailto:[email protected]


Polyethylene glycol-mediated colorectal cancer chemoprevention: roles of epidermal growth factor receptor and Snail

Documents