Canonical Wnt Signaling Is Required for Pancreatic ... · Tumor and Stem Cell Biology Canonical Wnt...

Tumor and Stem Cell Biology

Canonical Wnt Signaling Is Required for PancreaticCarcinogenesis

Yaqing Zhang1, John P. Morris IV8, Wei Yan2,5, Heather K. Schofield6, Austin Gurney9, Diane M. Simeone1,3,7,Sarah E. Millar10,11, Timothy Hoey9, Matthias Hebrok8, and Marina Pasca di Magliano1,4,7

AbstractWnt ligand expression and activation of the Wnt/b-catenin pathway have been associated with pancreatic

ductal adenocarcinoma, but whether Wnt activity is required for the development of pancreatic cancer hasremained unclear. Here, we report the results of three different approaches to inhibit the Wnt/b-cateninpathway in a established transgenic mousemodel of pancreatic cancer. First, we found that b-catenin null cellswere incapable of undergoing acinar to ductal metaplasia, a process associated with development ofpremalignant pancreatic intraepithelial neoplasia lesions. Second, we addressed the specific role of ligand-mediated Wnt signaling through inducible expression of Dkk1, an endogenous secreted inhibitor of thecanonical Wnt pathway. Finally, we targeted the Wnt pathway with OMP-18R5, a therapeutic antibody thatinteracts withmultiple Frizzled receptors. Together, these approaches showed that ligand-mediated activationof the Wnt/b-catenin pathway is required to initiate pancreatic cancer. Moreover, they establish that Wntsignaling is also critical for progression of pancreatic cancer, a finding with potential therapeutic implications.Cancer Res; 73(15); 4909–22. �2013 AACR.

IntroductionPancreatic ductal adenocarcinoma (PDA), the most com-

mon form of pancreatic cancer, is one of the deadliest humanmalignancieswith an overall 5-year survival rate of less than 5%(1). There is a dire need for improving our understanding of thebiology of pancreatic cancer to develop effective therapeuticoptions.PDA is almost universally associated with the presence of a

mutant, constitutively active form of Kras–most frequently theglycine to aspartic acid substitution KrasG12D–a small GTPaseprotein that is involved in signal transduction downstream ofreceptor tyrosine kinases (for review see ref. 2). Progression toPDA seems to occur through ductal precursor lesions, mostcommonly pancreatic intraepithelial neoplasia (PanIN), thataccrue morphologic and molecular atypia. PDA is character-ized by genomic instability, and a large number of mutations

and chromosomal abnormalities are found in each individualtumor (3). Notwithstanding the complexity of their genome,most human PDAs display mutations in 12 core signalingpathways including Wnt and other embryonic signaling path-ways (3). Genomic studies support previous work, indicatinginappropriate activation of Hedgehog, Notch, and Wnt signal-ing in PDA (for review see ref. 4). Although the functional role ofHedgehog and Notch signaling has been studied in depth, lessis known about the function of Wnt signaling during diseaseprogression.

TheWnt signaling pathway is complex and includes canon-ical and noncanonical branches (for review see ref. 5). For thepurpose of this study, we focused our analysis to canonicalWnt signaling. The canonical pathway is activated by Wntfamily ligands binding to Frizzled/LRP receptor complexes.The cascade of events that ensues prevents b-catenin degra-dation within the cytoplasm and allows its stabilization andnuclear translocation. In the nucleus, b-catenin binds totranscriptional factors of the Tcf/Lef family to form a tran-scriptional activator (6). Both inhibitory components of thepathway, such as Axin 2, and transcriptional coactivators,such as Tcf1 and Lef1, are also target genes, contributing tothe intricate feedback mechanisms that control pathwayactivity.

Mice with Cre-dependent, conditional expression of onco-genic KrasG12D in the pancreas (hereby referred to as KCmice)develop PanIN lesions that progress to PDA, closely mimick-ing the human disease (7). Because these animals developtumors in a stepwise manner in the context of an intactmicroenvironment, they have emerged as a relevant modelfor basic discovery as well as for preclinical studies (for reviewsee ref. 8). KCmice also recapitulate the aberrant reactivation

Authors' Affiliations: Departments of 1Surgery, 2Pathology, 3Molecularand Integrative Physiology, and 4Cell and Developmental Biology; 5Michi-gan Center for Translational Pathology; 6Medical Scientist Training Pro-gram; 7Comprehensive Cancer Center, University of Michigan MedicalSchool, Ann Arbor, Michigan; 8Diabetes Center, University of California,San Francisco, San Francisco; 9OncoMed Pharmaceutical, Redwood City,California; Departments of 10Dermatology, and 11Cell and DevelopmentalBiology, University of Pennsylvania, Philadelphia, Pennsylvania

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Marina Pasca di Magliano, Department of Sur-gery, University of Michigan Medical School, 1500 E Medical Center Drive,Ann Arbor, MI 48109. Phone: 734-615-7424; Fax: 734-647-9654; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-12-4384

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 4909

on May 27, 2020. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst June 12, 2013; DOI: 10.1158/0008-5472.CAN-12-4384

http://cancerres.aacrjournals.org/

of embryonic signaling pathways, including Wnt, observed inthe human disease (9). Thus, they constitute a valid model tostudy the functional requirement for Wnt signaling duringpancreatic carcinogenesis.

Previous functional studies addressing Wnt signaling inPDA have been mainly based on enforced aberrant b-cateninactivity using a dominant-active, degradation-resistant allele.When combined with mutant Kras, stabilized b-catenin pre-vents the formation of PanINs and PDA and instead leads tothe development of tumors reminiscent of human intraductaltubular tumors (ITT; ref. 10), a rare form of pancreatic cancerunrelated to PDA. Thus, high-level activation of Wnt signalingis incompatible with PDA formation. On the other hand,accumulation of b-catenin and activation of Wnt target genesare observed in PanINs and PDA (4, 9, 11, 12), and b-cateninfunctionally supports maintenance of PDA cell proliferationand tumor-forming capacity in xenograft models (9, 13). Thisparadox compelled us to explore whether activation of theWnt/b-catenin pathway is required during pancreatic carci-nogenesis, and what role it plays during different stages ofdisease progression.

We used three complementary approaches to block Wnt/b-catenin activation in the KC mouse model. First, we genet-ically inactivated b-catenin, a key pathway component. Sec-ond, we used a tetracycline-regulated system to overexpressDkk1, a secreted protein that has been shown to specificallyand effectively inhibit canonical-Wnt pathway activation (14,15). Third, we took a therapeutically relevant approach byusing OMP-18R5, a monoclonal antibody that inhibits Wntsignaling by binding Frizzled (Fzd) receptors (16).

Materials and MethodsMouse strains

The mice were housed in specific pathogen-free facilities atthe Comprehensive Cancer Center, University of Michigan(Ann Arbor, MI), or at the University of California, San Fran-cisco (San Francisco, CA). All the studies were conducted incompliance with the respective Institutional Committees onUse and Care of Animals guidelines. Detailed strain informa-tion is included in the Supplementary Methods.

Cell linesPrimary mouse pancreatic cancer cell lines 65671, 4292, and

9805 and primary human pancreatic cancer cell lines UM2,1319, UM18, and UM19 were used in this study. The cell line65671 was derived from the tumor of a KPC mouse (8) in pureFVBN genetic background. In addition, an E2F-luciferasereporter allele was carried by that mouse, thus the tumor cellsare labeledwith luciferase. The cell linewas derived in 2008 andgenotyped for recombination of the conditional alleles toexclude contamination by nonepithelial cells; genotyping wasrepeated in 2012. The cell lines 4292 and 9805 were derivedfrom iKras�p53� tumors (17) and genotyped for recombina-tion of the conditional alleles. The cell lines UM2, 1319, UM18,and UM19 were derived from patients with confirmed diag-nosis of pancreatic adenocarcinoma (University of Michigan;ref. 18). The cell lineswere passaged for less than 6months and

genotyped for mutant Kras to confirm their epithelial naturein 2012.

Doxycycline treatmentDoxycycline was administered in the drinking water, at a

concentration of 0.2 g/L in a solution of 5% sucrose, andreplaced every 3 to 4 days.

Tamoxifen treatmentTamoxifen was administered in 5 consecutive daily intra-

peritoneal injections of 1 mg. Mice were sacrificed for in vitrostudies 2 weeks later.

Three-dimensional acinar cell cultureThe three-dimensional (3D) culture of Elastase-Cre; b-cate-

ninf/f pancreatic acinar cells was prepared in a collagenmatrix,as previously described (19); the culture media was supple-mented with 100 ng/mL TGF-a to induce acinar-ductal meta-plasia. KC or KDC mice were treated with doxycycline 3 daysbefore harvest for 3D culture (also see Supplementary Data).The percentage of duct-like structures in 5 wells for each groupwas counted at day 3 of 3D culture. A two-tailed unpaired t testwas used for statistical analysis.

OMP-18R5 treatmentMonoclonal antibody OMP-18R5 was provided by OncoMed

Pharmaceuticals. OMP-18R5 was isolated from theMorphoSysHuCAL GOLD library. KC mice were treated with OMP-18R5(10 mg/kg, twice/wk) or PBS by intraperitoneal injection for 2months before sacrificed for study. Primary mouse pancreaticcancer cell line 65671 and human pancreatic cancer cell lineUM2 were treated with OMP-18R5 at 10 mg/mL and 20 mg/mL,respectively, in culture.

Recombinant Dkk1 treatmentThe mouse pancreatic cancer cell line 65671 and the human

pancreatic cancer cell line UM2were treatedwith recombinantmouse or human Dkk-1, respectively (500 ng/mL).

Histopathologic analysisHistopathologic analysiswas conducted by a pathologist (W.

Yan) on deidentified slides. Five images (�20 objective) weretaken in standardized positions (as to cover the whole section)for each slide. A minimum of 50 total acinar or ductal clusterswere counted from at least 3 independent animals for eachgroup. Each cluster counted was classified as normal (nl),ADM, PanIN1A, 1B, 2, or 3 based on the classification consen-sus (20). The data were expressed as percentage of totalcounted clusters. Error bars represent SEM.

Immunohistochemistry and immunofluorescenceImmunohistochemistry and immunofluorescence were

conducted as previously described (21). A list of antibodies isincluded in the Supplementary Table S1. Images were takenwith an Olympus BX-51 microscope, Olympus DP71 digitalcamera, and DP Controller software. The immunofluorescentimages were acquired using an Olympus IX-71 confocal micro-scope and FluoView FV500/IX software.

Zhang et al.

Cancer Res; 73(15) August 1, 2013 Cancer Research4910




Proliferation analysisThe proliferation index was calculated as percentage of

Ki67-positive cells. Error bars represent SEM. A two-tailedunpaired t test was used for statistical analysis.

TUNEL stainingFor apoptosis detection, the ApopTag Red In Situ Apoptosis

Detection Kit (S7165; Millipore) was used in accordance withthe manufacturer's protocol.

Western blottingWestern blotting was conducted as previously described

(21). Detailed antibody information is included in Supplemen-tary Table S1.

Quantitative RT-PCRQuantitative real-time PCR (qRT-PCR) was conducted as

previously described (21). The primers are listed in Supple-mentary Table S2. Values were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as housekeeping geneexpression control, and expressed as ratio over E-cadherin/Cdh1 expression, to account for the different proportion ofepithelium across different samples. A two-tailed unpaired ttest was used for statistical analysis.Detailed protocols and standard procedures are described in

the Supplementary Methods.

Resultsb-Catenin–negative cells do not contribute toKras-driven PanIN lesionsTodetermine ifb-catenin is required for PanIN formation,we

generated Ptf1a-Cre; LSL-KrasG12D; b-cateninf/f (KBC) micewhere both alleles of b-catenin are floxed in the context ofmutant Kras (Supplementary Fig. S1A). Recombination of thefloxed b-catenin allele results in lack of functional protein (22).KBCmice survived at a slightly lower thanMendelian ratio, butthen reached adulthood in apparent normal health. Pancreatadissected from 2-months-old KC mice had frequent PanINlesions, surrounded by desmoplastic stroma (SupplementaryFig. S1C). The pancreatic tissue in KBC mice presented withareas of fatty replacement surrounding clusters of normal aciniand islets, acinar to ductal metaplasia (ADM), and PanINs(Supplementary Fig. S1C). b-Catenin is essential for acinardevelopment (23), and thus the presence of fatty replacementlikely reflects the loss of acinar cells during embryogenesis. Inwild-type (WT) pancreata, b-catenin was localized at the cellmembrane; accumulation of cytoplasmic and nuclearb-cateninwas observed in PanIN lesions in KC mice, as previouslyreported (Supplementary Fig. S1D; refs. 9, 13). KBC tissuescontained a mosaic of b-catenin–positive and -negative cells;of note,mosaicism in the recombinationof thisfloxedb-cateninallele has been described previously (23–25). Interestingly, ADMand PanIN lesions consisted uniquely of b-catenin–positivecells; b-catenin–negative cells persisted in the tissue as normalacinar or ductal cells (Supplementary Fig. S1D). Occasionally,b-catenin–negative cells were observed within dilated ducts.Invariably, they were small, had cuboidal morphology, andlacked intracellular mucin, thus having characteristics of nor-

mal duct cells, even when surrounded by PanIN cells (Supple-mentary Fig. S1B). Further analysis revealed that b-cateninknockout clusters in KBC pancreata neither expressed CK19and p-ERK1/2 nor displayed an increase in proliferation, unlikeb-catenin–positive clusters within the same KBC tissues and incontrolKCpancreata (Supplementary Fig. S1E). Thus,b-cateninnull cells did not seem to contribute to PanIN lesions.

Mutant Kras can promote PanIN formation by driving ADM,a process by which acinar cells dedifferentiate to a duct-likestate capable of giving rise to PanIN lesions (10, 26). To det-ermine if b-catenin–negative acini were able to undergo ductalreprogramming, we used a 3D culture system. We isolatedacinar clusters from adult, tamoxifen-treated Elastase-CreERT2;b-cateninf/þ or Elastase-CreERT2; b-cateninf/f animals to studycells with specific deletion of b-catenin in the adult acinarcompartment. The clusters were cultured in a collagen matrixand stimulated to undergo ADM with TGF-a treatment, aspreviously described (19). In control, Elastase-CreERT2; b-cate-ninf/þADMwaswidespread, with few remaining acinar clustersafter 5 days of TGF-a treatment (Supplementary Fig. S1F). Incontrast, cultures derived from Elastase-CreERT2; b-cateninf/f

mice displayed a mixture of ductal structures and persistentacinar clusters (Supplementary Fig. S1F). By coimmunofluor-escence, we observed that the ductal structures (expressing theductalmarker CK19)were nearly universally positive for b-cate-nin, whereas the remaining acinar clusters (identified by mor-phology and amylase expression) were predominantly formedby b-catenin null cells (Supplementary Fig. S1G). Quantificationof these results identified a striking change in cell fate betweenthe b-catenin–positive and -negative cell population: cellsexpressing b-catenin were overwhelmingly found in theCK19þ (ductal) population, whereas virtually all the b-cateninnull cells were CK19-negative (Supplementary Fig. S1G). There-fore, b-catenin is required for acinar cells to undergo ADM.Taken together, these data provide, to our knowledge, thefirst functional evidence of the cell-autonomous requirementfor intact b-catenin function during the onset of pancreaticcarcinogenesis.

Inhibition of ligand-dependent Wnt signaling inhibitsPanIN formation

Because b-catenin is not only a key component of Wntsignaling, but also a structural component of the cadherin-based adherens junctions (27), we sought a different approachto block Wnt signaling during the onset of pancreatic carci-nogenesis. Therefore, we used a different genetic approach totarget canonical-Wnt signaling. We generated Ptf1a-Cre;Rosa26rtTa/þ; TetO-Dkk1 mice (referred to as DKK1 mice;Supplementary Fig. S2A) that allow inducible, doxycycline-dependent expression of the secreted Wnt inhibitor Dkk1 inthe pancreas. Dkk1 specifically inhibits canonical Wnt signal-ing by binding to theKremen family of receptors, a process thatresults in endocytosis, and thus elimination of the obligatecanonical Wnt LRP5/6 coreceptors from the cell surface (28).First, we validated our system by administering doxycyclineto 1-month-old DKK1 animals to induce Dkk1 expression(Supplementary Fig. S2B). By activating Dkk1 in postweaningmice, we were able to bypass the potential developmental

Wnt/b-Catenin in Pancreatic Cancer

www.aacrjournals.org Cancer Res; 73(15) August 1, 2013 4911




effects of inhibiting Wnt signaling during pancreas develop-ment. Two months later, we observed widespread, but mosaicDkk1 expression by immunohistochemistry (SupplementaryFig. S2D). Histologic analysis showed normal pancreatic his-tology, indicating that Dkk1 expression did not affect homeo-stasis of the adult pancreas (Supplementary Fig. S2C).

To address the role of ligand-dependent Wnt signalingduring PDA initiation, KC mice were crossed withRosa26rtTa/rtTa (29) and TetO-Dkk1 (14) mice to generate KC;Rosa26rtTa/þ; TetO-Dkk1 quadruple transgenics, herebyreferred to as KDC (Fig. 1A). In absence of doxycycline, thedisease progression in KDC mice was histologically indistin-guishable from that of KC mice (Supplementary Fig. S2E andS2F). Then, we expressed Dkk1 in the adult pancreatic epithe-lium of KDCmice, starting at 1month of age (Fig. 1B) when KCand KDC mice only have rare PanINs (7, 30); thus, in thisexperiment, Dkk1 expression preceded PanIN formation.Doxycycline water was continuously administered to KC andKDC littermates until the pancreatic tissue was harvested. Onthe basis of the expected disease progression in KC mice, weharvested tissue after 1, 2, 5, and 8 months (Fig. 1B) with aminimumof 4 animals per genotype per time point. Onemonthfollowing doxycycline administration, the pancreatic paren-chyma in KC mice was almost completely replaced by PanINlesions, ranging from PanIN1A to PanIN2, with occasionalPanIN3 lesions. In contrast, in age-matched KDC mice, mostof the pancreas was composed of normal acini, with rarePanINs, mainly PanIN1A (Fig. 1C). Both KC and KDC miceshowed areas of ADM. The analysis of tissues 2 months afterdoxycycline administration showed more and higher gradePanINs inKCanimals, but stillmostly normal acini andADM inKDC littermates. At later time points (5 and 8 months afterdoxycycline administration), however, PanINs ranging fromPanIN1A to PanIN2, interspersed between normal acini andADMwere also observed in KDCmice (Fig. 1C). At these stages,KC littermates had a higher percentage of PanIN2 and PanIN3with almost no normal acini and much fewer ADM lesions.Histopathologic quantification confirmed that KDC mice hadfewer PanINs overall and lower grade lesions than KC litter-mates at all of the time points (Fig. 1D). Summarily, KDC micedisplayed inhibited PanIN development and progression.

The inhibitory effect on PanIN formation observed in KDCmice, however, diminished over time (Fig. 1C), as revealed bythe progressive increase in several PanIN markers (Supple-mentary Figs. S3A–S3C and S4Aand S4B). Of note, similarly lowlevels of apoptosis was observed in PanIN lesions in KC andKDC mice (Supplementary Fig. 4C).

We then determined the ability of Dkk1 to inhibit Wnt/b-catenin signaling in the pancreas harvested at the indicatedtimepoints (Fig. 2A). Expression ofDkk1was confirmed inKDC,but not KC mice by Western blot analysis (Fig. 2B). We thenmeasured the levels of totalb-catenin,which accumulates in thepresence of active Wnt signaling, and phospho-b-catenin, pres-ent when theWnt pathway is not active. In KC animals, PanINsshowed high levels of total b-catenin and very little to nophospho-b-catenin; in contrast, KDC tissues had lower levelsof total b-catenin but increased levels of phospho-b-catenin atall the time points (Fig. 2B and C). However, at 8 months,

expression of Dkk1 was lower in KDC mice, and correspond-ingly we observed less phospho-b-catenin. Finally, we analyzedthe expression ofWnt target genes and pathway components inKC and KDC mice, comparing cohorts of 3-month-old animalsthat had been on doxycycline for 2 months. As expected, Dkk1expression was higher in KDC pancreata; the Wnt target genesAxin2, Lef1, MMP7, and the Wnt pathway components Wnt2,Fzd1, and Fzd2 had lower expression in KDC tissues, consistentwith reduced Wnt signaling. The effect was not due to overallreduced transcription, as we observed no change in other Wntpathway components not regulated by canonical pathwayactivity, such as LRP5, LRP6, and Wnt3a, among others (Fig.2D and data not shown). In addition, the Hedgehog pathwayligands sonic hedgehog (Shh) and Indian hedgehog (Ihh) andthe Notch target gene Hes1 were also significantly reduced inKDC tissues compared with KC (Fig. 2D), consistent with thelower PanIN load in these animals. Thus, our data show thatDkk1 expression efficiently blocks activation of the Wnt sig-naling pathway and inhibits PanIN formation, therefore pro-viding evidence that activation of theWnt/b-catenin pathway isrequired to initiate pancreatic carcinogenesis. To determinewhether the inhibitionof theWntpathwaywasmaintained overtime, we analyzed the expression of the Wnt target gene Axin2(Fig. 2D). Initially, Axin2 expression was lower in KDC than inKC animals, but its expression later rebounded in KDCmice. Infact, Axin2 expression was significantly higher in KDC than KCmice at the 8 month on doxy time point, a finding that mightreflect the different cellular composition of the tissues in eachcohort at this time point and possibly activation of the pathwayin nonepithelial cell types, or positive selection for cells expres-sing higher levels of Wnt components.

To determine whether inhibition ofWnt signaling prolongedoverall survival, we aged a cohort of KC (n¼ 46) and KDC (n¼24) mice on doxycycline from 1 month of age. Although onlyabout one third of the KC mice developed frank adenocarci-noma, most animals were sacrificed because of declining bodycondition within 1 year of age and presented at autopsy withPanIN3 and severe exocrine insufficiency due to loss of thepancreas parenchyma. KDC mice had significantly longer sur-vival (Supplementary Fig. S5A). We observed invasive tumorswith liver metastases both in KC and KDC mice and lungmetastases only in the KC cohort. Tumors from KC and KDCmice were highly proliferative, expressed phospho-ERK1/2,Muc1, andMMP7, and displayed accumulation of desmoplasticstroma (Supplementary Fig. S5B and S5C). We also observedaccumulation of b-catenin, indicating active Wnt pathway.Furthermore, immunostaining for Dkk1 in the KDC tumorrevealed that its expression was lost, in contrast with thesurrounding tissue (Supplementary Fig. S5C). These findingssuggest that the invasive tumor either originated from cells thathadnever recombined the rtTa transgene, thusnever expressingDkk1, or from cells with inactivated Dkk1 expression possiblydue to transgene silencing or insufficient doxycycline exposure.

Treatment with the OMP-18R5 therapeutic monoclonalantibody recapitulates Dkk1 overexpression

Given the genetic evidence of the ability of Wnt inhibitionto prevent PanIN formation, as shown earlier, we sought to

Zhang et al.





complement our studies by using a therapeutically relevantapproach. Namely, we used the monoclonal antibody OMP-18R5 that binds to multiple Frizzled receptors and blocks their

activity (16). One-month-old KC mice were treated with OMP-18R5 (10 mg/kg, twice/wk) or isotype control by intraperito-neal injection for 2 months, and then sacrificed (Fig. 3A). As a

Figure 1. Inhibition of the Wnt/b-catenin signaling by Dkk1 prevents PanIN formation. A, genetic makeup of the KDC model. B, experimental design. C, H&Estaining of KC and KDC pancreata at 1, 2, 5, and 8months following induction of Dkk1 expression. Scale bar, 100 mm. D, pathologic analysis. Data representmean � SEM, n ¼ 4. The statistical difference was determined by two-sided Student t test.






Zhang et al.





biomarker of the OMP-18R5 activity, we used hair growth, asWnt/b-catenin signaling in hair follicle precursor cells servesas a crucial signal for the telogen–anagen transition in mice(31). Treatment with OMP-18R5 prevented hair growth in KCmice after depilation, indicating effective Wnt inhibition (Sup-plementary Fig. S6B). Two months after initiating the study,histologic analysis of the pancreas revealed extensive PanINs,including high-grade lesions, in control KC animals, whereasKC mice treated with OMP-18R5 had a prevalence of normalacini and ADM, and fewer lesions (Fig. 3B and quantificationin Fig. 3C). Importantly, we observed an increase in phospho-b-catenin, indicating effective inhibition of Wnt signaling(Fig. 3B). In dissected tissues, pathway inhibition was con-firmed by qRT-PCR for the Wnt target genes Axin2, Lef1, andMMP7 (Fig. 3D).Moreover, we detected lower phospho-ERK1/2in OMP-18R5–treated samples, and a significant decrease inKi67-positive cells (Fig. 3B). Thus, treatment with OMP-18R5recapitulated the effect of Dkk1 overexpression, and validatedthe role of ligand-driven Wnt signaling in PanIN formation.OMP-18R5 binds multiple Frizzled receptors and thus

potentially inhibits both canonical and noncanonical Wntsignaling. In addition, inhibition of canonical Wnt signalingmight lead to redirection of the signaling toward the nonca-nonical pathway through a feedback mechanism (32). There-fore, we examined if OMP-18R5 treatment also affected non-canonical Wnt signaling. We measured the expression of thecore component of the planar cell polarity (PCP) pathway,Vangl2 (33) and the core component of Wnt/Ca2þ pathwayCdc42 (34), as well as the noncanonical ligand Wnt5a andfound no significant change in KDC mice or KC mice treatedwith OMP18R5 compared with control (Fig. 3E).

Cross-talk of Wnt/b-catenin and MAPK/ERK signalingduring PanIN formationActivation of the mitogen-activated protein kinase/extra-

cellular signal–regulated kinase (MAPK/ERK) pathway accom-panies, and is necessary for, PanIN formation (35, 36); even inpresence of oncogenic Kras, activation of the MAPK/ERKpathway requires additional stimuli, such as inflammation orEGF signaling (4, 37–42). We measured phospho-ERK1/2 asreadout of the MAPK/ERK pathway activation in KC andKDC mice at the time points indicated in Fig. 4A. Phospho-ERK1/2 was highly expressed in PanINs in KC pancreata (Fig.4B and C). In contrast, KDC pancreata had lower levels ofphospho-ERK1/2 as measured by immunohistochemistry andWestern blot analysis. Intriguingly, even the levels of totalERK1/2 protein were slightly lower in KDC pancreata than inKC animals, possibly indicating additional levels of regulationbesides phosphorylation. However, when PanIN lesions wereobserved in KDC tissues, they had high expression of phospho-ERK1/2, comparable with that observed in KC tissues (Fig. 4C;see 5 and 8 months time points). To further investigatethis finding, we conducted coimmunofluorescence for CK19

(a ductal marker expressed in ADM and PanINs), Dkk1, andphospho-ERK1/2. In 2-month-old KDC animals, cells expres-sing Dkk1weremostly CK19-negative (Fig. 4D). In 9-month-oldKC mice, we frequently observed high-grade PanIN lesionswith high phospho-ERK1/2 expression. High-grade PanINlesions were rare in age-matched KDC animals, and thoselesions that were present were largely negative for Dkk1expression (Fig. 4D). In 9-month-old KDC mice, Dkk1 wasexpressed in ADM and low-grade PanINs, but like the 2-monthtime point, cells expressing Dkk1 were largely CK19-negative,and cells coexpressing Dkk1 and CK19 generally displayed lowor near undetectable levels of phospho-ERK1/2 expression.Taken together, our data indicate that Dkk1-expressing cellsinefficiently undergo Kras-driven ADM; moreover, even if theyundergo ADM they rarely upregulate phospho-ERK1/2 corre-lating with impaired PanIN specification and progression tohigher lesion grade. Thus, the increase in Wnt target geneexpression and phospho-ERK1/2–positive lesions in olderKDCmice might be explained by a progressive loss of Dkk1-expres-sing cells–possibly due to proliferation disadvantage.

We then investigated the Wnt/MAPK cross-talk in estab-lished pancreatic cancer cells. Primary mouse pancreatic can-cer cells (65671) were derived from tumor-bearing KPC mice(Ptf1a-Cre; LSL-KrasG12D; p53flox/þ) inbred in the FVBN1 strain;they are a pure epithelial and highly tumorigenic population.The 65671 cells expressed Wnt ligands and had active Wntsignaling. Treatment of 65671 cells in culturewith recombinantrDkk1 or OMP-18R5 led to inhibition of Wnt activity, asdetermined by decrease in expression of the Wnt target geneAxin2 (Fig. 4E). Interestingly, rDkk1 or OMP-18R5 treatmentalso lead to downregulation of theMAPK signaling pathway (asindicated by reducedphospho-ERK1/2 levels), but hadno effecton phospho-Akt (Fig. 4F). We then investigated noncanonicalWnt signaling in response to OMP-18R5 treatment in this panelof primary tumor cells by measuring the level of activation oftwo kinases, c-jun-NH2-kinase (JNK) and protein kinase C(PKC), which mediate the Wnt/PCP and Wnt/Ca2þ signalingbranches, respectively (for review see ref. 43). We observed nochanges in p-JNK levels, and correspondingly no change in theexpression level of Vangl2 (Supplementary Fig. S6C and S6E). Incontrast, p-PKC levels transiently decreased upon treatmentwith OMP-18R5 (Supplementary Fig. S6C), a finding that mightindicate changes in the Wnt/Ca2þ signaling pathway.

To determine whether the cross-talk between the Wnt andMAPK pathway exists in human tumors, we used the primaryhuman PDA cell lines UM2, 1319, UM18, and UM19 (18). Wntinhibition resulted in downregulation of p-ERK1/2 and p-Akt(Supplementary Fig. S6C). Thus, even in primary human tumorcells,Wnt signaling is a positive regulator of theMAPKpathway.Using RT-PCR of canonical Wnt-targets to verify pathwayinhibition, treatment with rhDKK1 only inhibited Axin2 expres-sion in 2 of 4 cell lines tested andOMP-18R5was effective in 3 of4 cell lines (Supplementary Fig. S6F). Similarly, among 3primary

Figure 2. Dkk1 expression inhibitsWnt/b-catenin signaling. A, experimental design. B,Western blot analysis for Dkk1,b-catenin, andphospho-b-catenin inKCand KDC pancreas; quantification of the fold change of phospho-b-catenin/b-catenin. Data represent mean � SEM, n ¼ 3. C, immunohistochemistry forb-catenin and phospho-b-catenin (inset). Scale bar, 20 mm. D, qRT-PCR analysis of Dkk1, Wnt/b-catenin target genes and components. Each dot representsone mouse. Data represent mean � SEM. The statistical difference was determined by two-sided Student t test.






Figure 3. Inhibition of the Wnt/b-catenin signaling by OMP-18R5 prevents PanIN formation. A, experimental design for OMP-18R5 treatment; n¼ 5 mice percohort. B, H&E staining and immunohistochemical analysis forb-catenin, phospho-b-catenin, phospho-ERK1/2, andKi67. Scale bar, 50mm.C, quantificationof the lesions. D, qRT-PCRanalysis ofWnt/b-catenin target genes. E, qRT-PCRanalysis of noncanonicalWnt signaling components. Eachdot represents onemouse. Data represent mean � SEM. The statistical difference was determined by two-sided Student t test. ns, not significant.

Zhang et al.





mouse pancreatic cancer cell lines, both agents inhibited Axin2expression in 2 lines, 65671 and 4292, whereas only OMP-18R5inhibited Axin2 expression in the third line 9805 (Supplemen-tary Fig. S6D). The different response of individual cell lines toDkk1 and OMP-18R5 might reflect biologic differences such asmutations in signaling components that affect the functionalityof each inhibitor differently, based on their different mechan-isms of action.

Wnt/b-catenin signaling regulates pancreatic cellproliferation and Kras-driven acinar cellreprogrammingThe adult pancreas has a very low frequency of Ki67-positive

cells, as most cells are not dividing (Supplementary Fig. S7C).Proliferation was similarly low in Dkk1-expressing pancreata(Supplementary Fig. S7C). In KC animals, significantlyincreased proliferation–both in the epithelial compartmentand in the surrounding stroma–accompanies PanIN formation(Supplementary Fig. S7B and S7C). In contrast, we observed adecrease in proliferation in KDC tissues, specifically in theepithelial compartment (Supplementary Fig. S7B and S7C),both in ductal structures (ADM and PanINs) and acinar cells(Supplementary Fig. S7D). Therefore, our data suggest thatactive Wnt signaling supports Kras-dependent epithelial pro-liferation during the early steps of pancreatic carcinogenesis.We then investigated the effect of Wnt signaling inhibition

on the onset of ADM. To follow a specific cell cluster over time,we used 3D cultures of acinar clusters in Matrigel. WT and KCacinar clusters spontaneously underwent ADM in Matrigel;however, the formation of duct-like structures was significant-ly faster and more penetrant in KC acini (Fig. 5A and B).Moreover, KC-derived ductal structures accumulated moreintraluminal debris and more intracellular mucin over time(as shown by PAS staining; Fig. 5C). Formation of ductalstructures in KDC-derived cultures was equivalent to WT,with slower and less penetrant duct formation compared withKC clusters (Fig. 5A and quantification in Fig. 5B). Moreover,mucin accumulation–an indication of progression form ADMto PanIN–was reduced in KDC, compared with KC clusters(Fig. 5C). Thus, Dkk1 expression inhibited the capacity of Krasto drive acinar to ductal reprogramming and enforce anaberrant mucinous program in metaplastic ducts. This set ofdata was consistent with our in vivo observation that bothKras-driven ADM formation and progression to PanINs areinhibited in KDC pancreata.

Inhibition of Wnt signaling in established PanINsThe inducible Dkk1 expression in KDC mice gave us the

opportunity to investigate the effect of Wnt inhibition inestablished PanINs. In this set of experiments, doxycyclinewas administered to 2-month-old KC andKDC animals (n¼ 4),when a discrete number of PanINs lesions can be observed inthe tissues (Supplementary Fig. S2E and S2F). The animalswere euthanized 6 weeks later and the pancreas was dissected(see scheme in Fig. 5D), and pancreas size was measured as aproxy for disease progression as previously described (44). KDCmice had a 2-fold smaller pancreas compared with KC litter-mates, reflecting reduced pancreas growth over time (Fig. 5E

and G). Histologic analysis of the tissue did not reveal anysignificant morphologic difference between the two cohorts[Fig. 5G; hematoxylin and eosin (H&E) staining]; however, Ki67staining indicated a greater than 2-fold, in the proliferationindex of KDC tissues compared with KC (Fig. 5G, and quan-tification in Fig. 5F) both in the epithelial compartment and inthe surrounding stroma. Moreover, the number of apoptoticcells in the stroma of KDC animals [measured by cleavedcaspase-3 and by terminal deoxynucleotidyl transferase–medi-ated dUTP nick end labeling (TUNEL)] was higher in KDCtissues (Supplementary Fig. S8), suggesting that Wnt signalingmight exert a paracrine, prosurvival effect on the tumorstroma. Therefore, Wnt signaling inactivation blocks prolifer-ation of established PanIN lesions and may contribute tomaintenance of desmoplasia during PDA development.

DiscussionWnt signaling regulates numerous aspects of pancreatic

biology. During pancreas development, both inappropriateactivation and inactivation of the pathway lead to profounddefects of organ formation, including pancreatic agenesis (23,45, 46). Moreover, b-catenin, the critical transcriptional node incanonical Wnt signaling, is essential for exocrine regenerationfollowing damage such as induction of acute pancreatitis inmice (10, 47). Wnt signaling activity is gradually increasedduring pancreatic carcinogenesis (9, 10), but the levels of act-ivation are within physiologic limits even in advanced tumors.In fact, it was previously shown that forced high-level activationof the pathway prevents Kras-driven specification of acinar-derived PanINs (10) and diverts pancreatic cells undergoingKras-driven transformation from the PanIN/PDA progressiontoward an unusual tumor type, reminiscent of rare human ITT(48). Pancreatic cancer cells express several Wnt ligands (9,49, 50), possibly indicating autocrine activation of the pathway.Moreover, mutations in Wnt signaling components have beenidentified in the majority of human pancreatic tumors (3), aswell as in amouse screen formutations synergizingwithKras indriving pancreatic carcinogenesis (51). The exact biologic sig-nificance of each of themutations identified is in need of furtherexploration. In addition to canonical Wnt signaling, noncanon-ical signaling may also play an important role in this disease(52–56). The relative importance of the different modalities ofWnt signaling requires further investigation.

Here, we show that inhibition of Wnt signaling significantlydelayed PanIN formation. Taken together, our results suggestthat there is aminimal threshold of ligand-mediated, canonicalWnt signaling required for PanIN formation (Fig. 6). In con-trast, high levels of Wnt activation prevent specification of thePanIN-PDA lineage (Fig. 6; refs. 10, 48). In the future, geneticmodels that permit controllable, graded increases in b-cateninactivity, or inducible expression of Wnt ligands will elucidatethe threshold at which canonical Wnt signaling is no longercompatible with Kras-driven PanIN development.

An essential step for the initiation of pancreatic carcino-genesis is activation of the MAPK/ERK signaling pathway. InKC mice, activation of MAPK signaling is a key step for PanINformation (35). Mutant Kras and Wnt signaling pathwaycooperate to drive tumorigenesis in other tissues, for example






Figure 4. Cross-talk of Wnt andMAPK signaling pathways. A, experimental design; n¼ 3–5mice per time point. B, Western blot analysis for total ERK1/2 andphospho-ERK1/2; quantification of the fold change of phospho-ERK1/2/total ERK1/2. Data represent mean � SEM, n ¼ 3. C, immunohistochemistry for

Zhang et al.





in lung and colon adenocarcinoma (57, 58). Epistatic interac-tions between MAPK and Wnt signaling have been recentlydescribed both in melanoma and in colon cancer (59, 60).Intriguingly, these interactions are clearly tissue-specific: inmelanomaWnt signaling inhibits MAPK signaling (60), where-

as in colon cancer Wnt signaling stabilizes Ras and thuspromotes MAPK activity (59). Here, we show that Wnt signal-ing in the pancreas promotesMAPK signaling andhas a tumor-promoting effect. Thus, the identity of b-catenin targets thatantagonize or promote Kras-dependent transformation may

Figure 5. Wnt signaling is requiredfor Kras-driven acinar-ductalmetaplasia andPanIN proliferation.A, transmitted light images, H&Estaining, and amylaseimmunofluorescent staining of KCand KDC pancreatic cell clusters in3D culture. Arrows, ductstructures; asterisks, acinarclusters. Scale bar, 20 mm. B,quantification of duct-likestructures at day 3. Data representmean� SEM; n¼ 5. The statisticaldifference was determined by two-sided Student t test. C, PASstaining. Arrows, PAS-positivecells. Scale bar, 20 mm. D,experimental design; n¼ 4mice. E,pancreas size (mean�SEM; n¼4).The statistical difference wasdetermined by two-sided Student ttest. F, proliferation index inepithelial and stromalcompartments. Data representmean� SEM; n¼ 4. The statisticaldifference was determined by two-sided Student t test. G, grossmorphology of KC and KDCpancreata (scale bar, 2 mm);H&E and Ki67 staining (scale bar,50 mm).

phospho-ERK1/2. Scale bar, 50 mm. D, coimmunofluorescence for Dkk1 (green), CK19 (red), pERK1/2 (magenta), and 40,6-diamidino-2-phenylindole (DAPI;blue). Arrows, Dkk1-positive cells. Scale bar, 25 mm. Primary pancreatic cancer cells 65671 were treated with rDkk1. E, qRT-PCR for Wnt/b-catenin targetgenes Axin2 and Lef1. Data represent mean� SEM; n¼ 3. The statistical difference was determined by two-sided Student t test. F, Western blot analysis forphospho-ERK1/2, total ERK, phospho-Akt, total Akt, and GAPDH. �, P < 0.05.






be tissue specific, and thusmay provide unique cancer-specifictherapeutic targets.

Studying the role of Wnt/b-catenin during the initiation ofPDA highlighted the potential role of targeting Wnt/b-cateninsignaling in disease initiation. This finding has implications forthe potential use of Wnt inhibitors in patients with pancreaticcancer, at a time when several Wnt inhibitors, including OMP-18R5 have entered the clinical pipeline for treatment of avariety of cancers, including PDA (for review see refs. 61, 62).OMP-18R5 has been found to be active in combination with avariety of chemothapeutic agents in patient-derived xenografts,including gemcitabine inPDAxenografts (16). Our data indicatethat the mechanism of action might involve a reduction in theactivation of the MAPK/ERK signaling, and inhibition of pro-liferation. Further studies might elucidate whether interactionsbetween the tumor cells and the surrounding stroma are alsoregulated through Wnt signaling.

Disclosure of Potential Conflicts of InterestM.Hebrok is a consultant/advisory boardmember ofMerck.M.P. diMagliano

has a commercial research grant from OncoMed. No potential conflicts ofinterest were disclosed by the other authors.

DisclaimerThe study sponsors had no role in the design of this study or in the collection,

analysis, and interpretation of the data.

Authors' ContributionsConception and design: Y. Zhang, D.M. Simeone, M.P. di MaglianoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y. Zhang, J.P. Morris IV, W. Yan, H.K. Schofield, D.M.Simeone, S.E. Millar, M. HebrokAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Zhang, J.P. Morris IV, W. Yan, H.K. Schofield, D.M.Simeone, T. Hoey, M. Hebrok, M.P. di MaglianoWriting, review, and/or revision of themanuscript: Y. Zhang, J.P. Morris IV,H.K. Schofield, A. Gurney, D.M. Simeone, T. Hoey, M.P. di MaglianoAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): A. GurneyStudy supervision: M.P. di Magliano

AcknowledgmentsThe authors thank Monique Verhagen and Anj Dlugosz for assistance with

the hair growth assay. The Ptf1a-Cre mouse was a generous gift from Dr. ChrisWright (Vanderbilt University, Nashville, TN).

Grant SupportThis project was supported by the University of Michigan Biological Scholar

Program and NCI-1R01CA151588-01 and a grant fromOncoMed Pharmaceutical(M.P. diMagliano and D.M. Simeone). Work inMatthias Hebrok's laboratory wassupported by a NIH grant (CA112537).

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 to indicate thisfact.

Received December 11, 2012; revised May 3, 2013; accepted May 6, 2013;published OnlineFirst June 12, 2013.

References1. Katz MH, Wang H, Fleming JB, Sun CC, Hwang RF, Wolff RA,

et al. Long-term survival after multidisciplinary management ofresected pancreatic adenocarcinoma. Ann Surg Oncol 2009;16:836–47.

2. Pylayeva-Gupta Y, Grabocka E, Bar-Sagi D. RAS oncogenes: weavinga tumorigenic web. Nat Rev Cancer 2011;11:761–74.

3. Jones S, Zhang X, Parsons DW, Lin JC, Leary RJ, Angenendt P, et al.Core signaling pathways in human pancreatic cancers revealed byglobal genomic analyses. Science 2008;321:1801–6.

4. Morris JP, Wang SC, Hebrok M. KRAS, Hedgehog, Wnt and thetwisted developmental biology of pancreatic ductal adenocarcinoma.Nat Rev Cancer 2010;10:683–95.

ADM PanIN

No Wnt/ββ-cateninWnt/β-catenin functional MAPK/ERK activation

High Wntsignaling

Low Wntsignaling

Intraductaltubular tumors

Kra

s-m

uta

nt

pan

crea

s

Morris and colleagues, 2010

No Wntsignaling

High MAPK signaling

Normal acini

Low MAPK signaling

Figure 6. Wnt/b-catenin signaling inKras-driven pancreaticcarcinogenesis. A physiologic levelofWnt signaling (lowWnt signaling,magenta arrow) is maintainedduring the formation of pancreaticcancer precursor lesions. If Wntsignaling levels are altered, eitherby overexpression (red arrow) or byinactivation (blue arrow),pancreatic carcinogenesis isprevented.

Zhang et al.





5. Angers S, Moon RT. Proximal events in Wnt signal transduction. NatRev Mol Cell Biol 2009;10:468–77.

6. Gordon MD, Nusse R. Wnt signaling: multiple pathways, multiplereceptors, and multiple transcription factors. J Biol Chem 2006;281:22429–33.

7. Hingorani SR, Petricoin EF, Maitra A, Rajapakse V, King C, JacobetzMA, et al. Preinvasive and invasive ductal pancreatic cancer and itsearly detection in the mouse. Cancer Cell 2003;4:437–50.

8. Olive KP, Tuveson DA. The use of targeted mouse models for pre-clinical testing of novel cancer therapeutics. Clin Cancer Res 2006;12:5277–87.

9. Pasca di Magliano M, Biankin AV, Heiser PW, Cano DA, Gutierrez PJ,Deramaudt T, et al. Common activation of canonical Wnt signaling inpancreatic adenocarcinoma. PLoS ONE 2007;2:e1155.

10. Morris JP, Cano DA, Sekine S, Wang SC, Hebrok M. Beta-cateninblocks Kras-dependent reprogramming of acini into pancreatic cancerprecursor lesions in mice. J Clin Invest 2010;120:508–20.

11. Al-Aynati MM, Radulovich N, Riddell RH, TsaoMS. Epithelial-cadherinand beta-catenin expression changes in pancreatic intraepithelialneoplasia. Clin Cancer Res 2004;10:1235–40.

12. Zeng G, Germinaro M, Micsenyi A, Monga NK, Bell A, Sood A, et al.Aberrant Wnt/beta-catenin signaling in pancreatic adenocarcinoma.Neoplasia 2006;8:279–89.

13. Wang L, Heidt DG, Lee CJ, Yang H, Logsdon CD, Zhang L, et al.Oncogenic function of ATDC in pancreatic cancer through Wnt path-way activation and beta-catenin stabilization. Cancer Cell 2009;15:207–19.

14. Chu EY, Hens J, Andl T, Kairo A, Yamaguchi TP, Brisken C, et al.Canonical WNT signaling promotes mammary placode developmentand is essential for initiation of mammary gland morphogenesis.Development 2004;131:4819–29.

15. Yang SH, Andl T, Grachtchouk V, Wang A, Liu J, Syu LJ, et al.Pathological responses to oncogenic Hedgehog signaling in skin aredependent on canonical Wnt/beta3-catenin signaling. Nat Genet2008;40:1130–5.

16. Gurney A, Axelrod F, BondCJ, Cain J, Chartier C, Donigan L, et al. Wntpathway inhibition via the targeting of Frizzled receptors results indecreased growth and tumorigenicity of human tumors. Proc NatlAcad Sci U S A 2012;109:11717–22.

17. Collins MA, Brisset JC, Zhang Y, Bednar F, Pierre J, Heist KA, et al.Metastatic pancreatic cancer is dependent on oncogenic Kras inmice.PLoS ONE 2012;7:e49707.

18. Li C, Heidt DG, Dalerba P, Burant CF, Zhang L, Adsay V, et al.Identification of pancreatic cancer stem cells. Cancer Res 2007;67:1030–7.

19. Miyamoto Y, Maitra A, Ghosh B, Zechner U, Argani P, Iacobuzio-Donahue CA, et al. Notch mediates TGF alpha-induced changes inepithelial differentiation during pancreatic tumorigenesis. Cancer Cell2003;3:565–76.

20. Hruban RH, Adsay NV, Albores-Saavedra J, Anver MR, Biankin AV,Boivin GP, et al. Pathology of genetically engineered mousemodels ofpancreatic exocrine cancer: consensus report and recommendations.Cancer Res 2006;66:95–106.

21. Collins MA, Bednar F, Zhang Y, Brisset JC, Galban S, Galban CJ, et al.Oncogenic Kras is required for both the initiation and maintenance ofpancreatic cancer in mice. J Clin Invest 2012;122:639–53.

22. Brault V, Moore R, Kutsch S, Ishibashi M, Rowitch DH, McMahon AP,et al. Inactivation of the beta-catenin gene by Wnt1-Cre-mediateddeletion results in dramatic brain malformation and failure of cranio-facial development. Development 2001;128:1253–64.

23. Murtaugh LC, Law AC, Dor Y, Melton DA. Beta-catenin is essential forpancreatic acinar but not islet development. Development 2005;132:4663–74.

24. Dessimoz J, Bonnard C, Huelsken J, Grapin-Botton A. Pancreas-specific deletion of beta-catenin reveals Wnt-dependent and Wnt-independent functions during development. Curr Biol 2005;15:1677–83.

25. Papadopoulou S, Edlund H. Attenuated wnt signaling perturbspancreatic growth but not pancreatic function. Diabetes 2005;54:2844–51.

26. Habbe N, Shi G, Meguid RA, Fendrich V, Esni F, Chen H, et al.Spontaneous induction of murine pancreatic intraepithelial neoplasia(mPanIN) by acinar cell targeting of oncogenic Kras in adult mice. ProcNatl Acad Sci U S A 2008;105:18913–8.

27. Valenta T, Hausmann G, Basler K. The many faces and functions ofbeta-catenin. EMBO J 2012;31:2714–36.

28. MaoB,WuW, DavidsonG,Marhold J, Li M,Mechler BM, et al. Kremenproteins are Dickkopf receptors that regulate Wnt/beta-catenin sig-nalling. Nature 2002;417:664–7.

29. Belteki G, Haigh J, Kabacs N, Haigh K, Sison K, Costantini F, et al.Conditional and inducible transgene expression in mice through thecombinatorial use of Cre-mediated recombination and tetracyclineinduction. Nucleic Acids Res 2005;33:e51.

30. Aguirre AJ, Bardeesy N, Sinha M, Lopez L, Tuveson DA, Horner J,et al. Activated Kras and Ink4a/Arf deficiency cooperate to producemetastatic pancreatic ductal adenocarcinoma. Genes Dev 2003;17:3112–26.

31. VanMaterD, Kolligs FT,DlugoszAA, FearonER. Transient activationofbeta -catenin signaling in cutaneous keratinocytes is sufficient totrigger the active growth phase of the hair cycle in mice. Genes Dev2003;17:1219–24.

32. Grumolato L, Liu G, Mong P, Mudbhary R, Biswas R, Arroyave R, et al.Canonical and noncanonical Wnts use a common mechanism toactivate completely unrelated coreceptors. Genes Dev 2010;24:2517–30.

33. Montcouquiol M, Rachel RA, Lanford PJ, Copeland NG, Jenkins NA,Kelley MW. Identification of Vangl2 and Scrb1 as planar polarity genesin mammals. Nature 2003;423:173–7.

34. Axelrod JD, Miller JR, Shulman JM, Moon RT, Perrimon N. Differentialrecruitment of Dishevelled provides signaling specificity in the planarcell polarity and Wingless signaling pathways. Genes Dev 1998;12:2610–22.

35. Collisson EA, Trejo CL, Silva JM, Gu S, Korkola JE, Heiser LM, et al. Acentral role for RAF!MEK!ERKsignaling in the genesis of pancreaticductal adenocarcinoma. Cancer Discov 2012;2:685–93.

36. Qu C, DiRenzo D, Konieczny SF. Identifying molecular signatures forthe earliest KRAS-induced transformation events in pancreatic ade-nocarcinoma development. Pancreas 2011;40:1349.

37. Ardito CM, Gruner BM, Takeuchi KK, Lubeseder-Martellato C, Teich-mann N, Mazur PK, et al. EGF receptor is required for KRAS-inducedpancreatic tumorigenesis. Cancer Cell 2012;22:304–17.

38. Daniluk J, Liu Y, Deng D, Chu J, Huang H, Gaiser S, et al. An NF-kappaB pathway-mediated positive feedback loop amplifies Rasactivity to pathological levels in mice. J Clin Invest 2012;122:1519–28.

39. Guerra C, ColladoM, Navas C, Schuhmacher AJ, Hernandez-Porras I,Canamero M, et al. Pancreatitis-induced inflammation contributes topancreatic cancer by inhibiting oncogene-induced senescence. Can-cer Cell 2011;19:728–39.

40. Guerra C, Schuhmacher AJ, Canamero M, Grippo PJ, Verdaguer L,Perez-Gallego L, et al. Chronic pancreatitis is essential for induction ofpancreatic ductal adenocarcinoma byK-Ras oncogenes in adultmice.Cancer Cell 2007;11:291–302.

41. Navas C, Hernandez-Porras I, Schuhmacher AJ, Sibilia M, Guerra C,Barbacid M. EGF receptor signaling is essential for k-ras oncogene-driven pancreatic ductal adenocarcinoma. Cancer Cell 2012;22:318–30.

42. Carriere C, Young AL, Gunn JR, Longnecker DS, Korc M. Acutepancreatitis accelerates initiation and progression to pancreatic can-cer in mice expressing oncogenic Kras in the nestin cell lineage. PLoSONE 2011;6:e27725.

43. SemenovMV,HabasR,MacdonaldBT,HeX.SnapShot: noncanonicalWnt signaling pathways. Cell 2007;131:1378.

44. Nolan-Stevaux O, Lau J, Truitt ML, Chu GC, Hebrok M, Fernandez-Zapico ME, et al. GLI1 is regulated through Smoothened-independentmechanisms in neoplastic pancreatic ducts and mediates PDAC cellsurvival and transformation. Genes Dev 2009;23:24–36.

45. Heiser PW, Lau J, Taketo MM, Herrera PL, Hebrok M. Stabilizationof beta-catenin impacts pancreas growth. Development 2006;133:2023–32.






46. Wells JM,Esni F, BoivinGP, AronowBJ,StuartW,CombsC, et al.Wnt/beta-catenin signaling is required for development of the exocrinepancreas. BMC Dev Biol 2007;7:4.

47. Keefe MD, Wang H, De La OJ, Khan A, Firpo MA, Murtaugh LC.b-Catenin is selectively required for the expansion and regenerationof mature pancreatic acinar cells in mice. Dis Model Mech 2012;5:503–14.

48. Heiser PW, Cano DA, Landsman L, Kim GE, Kench JG, Klimstra DS,et al. Stabilization of beta-catenin induces pancreas tumor formation.Gastroenterology 2008;135:1288–300.

49. ArensmanMD, Kovochich AN, Kulikauskas RM, Lay AR, Yang PT, Li X,et al. WNT7B mediates autocrine Wnt/beta-catenin signaling andanchorage-independent growth in pancreatic adenocarcinoma.Oncogene. 2013 Feb 18. [Epub ahead of print].

50. Nawroth R, van Zante A, Cervantes S, McManus M, Hebrok M, RosenSD. Extracellular sulfatases, elements of the wnt signaling pathway,positively regulate growth and tumorigenicity of human pancreaticcancer cells. PLoS ONE 2007;2:e392.

51. Mann KM, Ward JM, Yew CC, Kovochich A, Dawson DW, Black MA,et al. SleepingBeautymutagenesis reveals cooperatingmutations andpathways in pancreatic adenocarcinoma. Proc Natl Acad Sci U S A2012;109:5934–41.

52. Yu M, Ting DT, Stott SL, Wittner BS, Ozsolak F, Paul S, et al. RNAsequencing of pancreatic circulating tumour cells implicates WNTsignalling in metastasis. Nature 2012;487:510–3.

53. Crnogorac-Jurcevic T, Efthimiou E, Capelli P, Blaveri E, Baron A, TerrisB, et al. Gene expression profiles of pancreatic cancer and stromaldesmoplasia. Oncogene 2001;20:7437–46.

54. Griesmann H, Ripka S, Pralle M, Ellenrieder V, Baumgart S, BuchholzM, et al. WNT5A-NFAT signaling mediates resistance to apoptosis inpancreatic cancer. Neoplasia 2013;15:11–22.

55. Pilarsky C, Ammerpohl O, Sipos B, Dahl E, Hartmann A, WellmannA, et al. Activation of Wnt signalling in stroma from pancreaticcancer identified by gene expression profiling. J Cell Mol Med2008;12:2823–35.

56. Ripka S, Konig A, Buchholz M, Wagner M, Sipos B, Kloppel G, et al.WNT5A—target of CUTL1 and potentmodulator of tumor cell migrationand invasion in pancreatic cancer. Carcinogenesis 2007;28:1178–87.

57. Pacheco-Pinedo EC, Durham AC, Stewart KM, Goss AM, Lu MM,Demayo FJ, et al. Wnt/beta-catenin signaling accelerates mouse lungtumorigenesis by imposing an embryonic distal progenitor phenotypeon lung epithelium. J Clin Invest 2011;121:1935–45.

58. Janssen KP, Alberici P, Fsihi H, Gaspar C, Breukel C, Franken P, et al.APC and oncogenic KRAS are synergistic in enhancing Wnt signalingin intestinal tumor formation and progression. Gastroenterology2006;131:1096–109.

59. Jeong WJ, Yoon J, Park JC, Lee SH, Kaduwal S, Kim H, et al. Rasstabilization through aberrant activation of Wnt/beta-catenin signalingpromotes intestinal tumorigenesis. Sci Signal 2012;5:ra30.

60. Biechele TL, Kulikauskas RM, Toroni RA, Lucero OM, Swift RD, JamesRG, et al. Wnt/beta-catenin signaling and AXIN1 regulate apoptosistriggered by inhibition of the mutant kinase BRAFV600E in humanmelanoma. Sci Signal 2012;5:ra3.

61. Li H, Pamukcu R, Thompson WJ. beta-Catenin signaling: therapeuticstrategies in oncology. Cancer Biol Ther 2002;1:621–5.

62. PolakisP.DruggingWnt signalling in cancer. EMBOJ2012;31:2737–46.


Zhang et al.




2013;73:4909-4922. Published OnlineFirst June 12, 2013.Cancer Res Yaqing Zhang, John P. Morris IV, Wei Yan, et al. Canonical Wnt Signaling Is Required for Pancreatic Carcinogenesis

Updated version

10.1158/0008-5472.CAN-12-4384doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/08/02/0008-5472.CAN-12-4384.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/15/4909.full#ref-list-1

This article cites 61 articles, 23 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/15/4909.full#related-urls

This article has been cited by 15 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]

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

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/73/15/4909To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-12-4384

http://cancerres.aacrjournals.org/content/suppl/2013/08/02/0008-5472.CAN-12-4384.DC1

http://cancerres.aacrjournals.org/content/73/15/4909.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/15/4909.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/15/4909


Canonical Wnt Signaling Is Required for Pancreatic ... · Tumor and Stem Cell Biology Canonical Wnt...

Documents

Transcript of Canonical Wnt Signaling Is Required for Pancreatic ... · Tumor and Stem Cell Biology Canonical Wnt...