merlin/nf2 suppresses pancreatic tumor growth and metastasis … · indicated that merlin...

Molecular and Cellular Pathobiology

Merlin/NF2 Suppresses Pancreatic Tumor Growthand Metastasis by Attenuating the FOXM1-Mediated Wnt/b-Catenin SignalingMing Quan1,2, Jiujie Cui1,2, Tian Xia3, Zhiliang Jia2, Dacheng Xie1, Daoyan Wei2,Suyun Huang4, Qian Huang1, Shaojiang Zheng5, and Keping Xie2

Abstract

Merlin, theprotein encodedby theNF2 gene, is amemberof theband 4.1 family of cytoskeleton-associated proteins and functionsas a tumor suppressor formany types of cancer.However, the rolesand mechanism of Merlin expression in pancreatic cancer haveremained unclear. In this study, we sought to determine theimpact of Merlin expression on pancreatic cancer developmentand progression using human tissue specimens, cell lines, andanimal models. Decreased expression of Merlin was pronouncedin human pancreatic tumors and cancer cell lines. Functionalanalysis revealed that restored expression of Merlin inhibitedpancreatic tumor growth and metastasis in vitro and in vivo.

Furthermore, Merlin suppressed the expression of Wnt/b-cateninsignaling downstream genes and the nuclear expression of b-cate-nin protein, and overexpression of Forkhead box M1 (FOXM1)attenuated the suppressive effect of Merlin on Wnt/b-cateninsignaling. Mechanistically, Merlin decreased the stability ofFOXM1protein, which plays critical roles in nuclear translocationofb-catenin.Collectively, thesefindingsdemonstrated thatMerlincritically regulated pancreatic cancer pathogenesis by suppressingFOXM1/b-catenin signaling, suggesting that targeting novel Mer-lin/FOXM1/b-catenin signaling is an effective therapeutic strategyfor pancreatic cancer. Cancer Res; 75(22); 4778–89. �2015 AACR.

IntroductionPancreatic ductal adenocarcinoma (PDAC) has a dismal prog-

nosis and is the seventh leading cause of cancer-related deaths (1).Its incidence is increasing annually, and in the United States,researchers estimated that 46,420 new PDAC cases would bediagnosed and 39,590 patients would die of it in 2014 (2).Despite improvements in early diagnosis and treatment, morethan 80% of PDAC cases are diagnosed at late stages, and the

5-year survival rate for PDAC is less than 5% (3). Thus, identi-fication of the molecular mechanisms underlying PDAC devel-opment and progression is urgently needed.

Merlin, the protein encoded by the NF2 gene, is a member ofthe band 4.1 families of cytoskeleton-associated proteins, whichlink the integral membrane proteins with the actin cytoskeleton(4, 5). Investigators first identifiedMerlin as being associatedwithneurofibromatosis type 2 and that it functions as a tumor sup-pressor (6). A number of studies demonstrated that Merlin is aversatile tumor suppressor that can inhibit cancer cell prolifera-tion and motility by modulating a wide range of signaling path-ways (7). Furthermore, mutation of the NF2 gene and loss ofMerlin protein occur in many different types of cancer, suggestinga general tumor-suppressive role for Merlin (8–13). Evidently,Merlin shuttles between the cell cortex and the nucleus and thatboth cortical Merlin and nuclear Merlin have been implicated intumor suppression via interaction with angiomotin and CRL4-DCAF1, respectively (11, 13). In a study of ezrin, anothermemberof the band 4.1 protein superfamily, overexpression of Merlininhibited SW1990 PDAC cell proliferation, migration, and adhe-sion (14). However, the roles and mechanism of Merlin expres-sion in PDAC development and progression have remainedunclear.

Wnt/b-catenin is one of themost important signaling pathwaysin PDACdevelopment and progression (15). Briefly, this pathwayis initiated by binding of Wnt ligands to receptors of the Frizzledfamily and coreceptors on the cell surface. With ligand binding,the cytoplasmic degradation complex, which consists of Axin,adenomatous polyposis coli, glycogen synthase kinase-3b, andcasein kinase 1, is inhibited, leading to nuclear localization ofb-catenin. In the nucleus, b-catenin binds to T-cell factor/lym-phoid enhancer factor to activate Wnt downstream target genes,

1Department of Oncology, Shanghai Jiaotong University AffiliatedFirst People's Hospital, Shanghai, PR China. 2Department of Gastro-enterology, Hepatology and Nutrition, The University of Texas MDAnderson Cancer Center, Houston, Texas. 3Department of Gastroen-terology, Shanghai Changhai Hospital, Second Military MedicalUniversity, Shanghai, PR China. 4Department of Neurosurgery, TheUniversity of Texas MD Anderson Cancer Center, Houston, Texas.5Pathology Department of Affiliated Hospital, Hainan Provincial KeyLaboratory of Carcinogenesis and Intervention, Hainan Medical Col-lege, Haikou, Hainan, PR China.

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

M. Quan, J. Cui, and T. Xia contributed equally to this article.

Corresponding Authors: Keping Xie, Department of Gastroenterology, Hepa-tology & Nutrition, Unit 1466, The University of Texas MD Anderson CancerCenter, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: 713-794-5073;Fax: 713-745-3654; E-mail: [email protected]; Qian Huang, Departmentof Oncology, Shanghai Jiaotong University Affiliated First People's Hospital,Shanghai, PR China. Phone: 86-21-63240090-3121; E-mail:[email protected]; and Shaojiang Zheng, Department of Pathology,Hainan Medical College Affiliated Hospital, Haikou, PR China. Phone: 86-898-66723333; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-1952

�2015 American Association for Cancer Research.

CancerResearch

Cancer Res; 75(22) November 15, 20154778

on October 16, 2020. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 19, 2015; DOI: 10.1158/0008-5472.CAN-14-1952

http://cancerres.aacrjournals.org/

including cyclinD1,matrixmetalloproteinase 2 (MMP2),MMP9,and VEGF. Although studies of mouse models demonstrated thatactivation of Wnt/b-catenin signaling was insufficient to initiatePDAC, this pathway played key roles in PDAC stem cell mainte-nance, progression, and drug resistance (15, 16). Recent studiesindicated that Merlin suppressed the Wnt/b-catenin signalingpathway (9, 17, 18); however, the molecular mechanism of thissuppression and the relevance of the interaction between Merlinand b-catenin remain to be elucidated.

Forkhead boxM1 (FOXM1) is a transcription factor in the FOXsuperfamily characterized by a conserved winged helix DNA-binding domain (19). FOXM1 is a key regulator of cell-cycleprogression, and series of studies suggested that it plays criticalroles in PDAC growth, angiogenesis, invasion, and metastasis(20, 21). Recently, we found that FOXM1 bound directly tob-catenin and had a key role in mediation of nuclear accumula-tion of b-catenin and downstream target gene expression (22, 23).

In the present study, we sought to determine the roles andmechanism ofMerlin expression in PDAC growth andmetastasis.We discovered that expression of Merlin was suppressed in PDACcells and that restored expression of Merlin inhibited PDAC cellgrowth and metastasis in vitro and in vivo. Further studies dem-onstrated that re-expression of Merlin decreased the nucleartranslocation of b-catenin by inducing instability of FOXM1protein.

Materials and MethodsCell culture and treatment

Human embryonic kidney 293 (HEK293) cells and thehuman PDAC cell lines PANC-1, MiaPaCa-2, AsPC-1, BxPC-3, CaPan-1, CaPan-2, and PA-TU-8902 were purchased fromthe ATCC. The PDAC cell line MDA Panc-28 was a gift from Dr.Paul J. Chiao (The University of Texas MD Anderson CancerCenter, Houston, TX). The human PDAC cell line FG wasobtained from Michael P. Vezeridis (Brown Medical School,Providence, RI; ref. 24). The human metastatic PDAC cell lineCOLO357 and its fast-growing liver-metastatic variant in nudemice, L3.7, were described previously (20). All of these cell lineswere maintained in plastic flasks as adherent monolayers inEagle's Minimal Essential Medium supplemented with 10%FBS, sodium pyruvate, nonessential amino acids, L-glutamine,and a vitamin solution (Flow Laboratories). Immortalizedhuman pancreatic ductal epithelial cells (provided by Dr. M.S. Tsao, Ontario Cancer Institute, Toronto, Ontario, Canada)were maintained in a keratinocyte serum-free medium supple-mented with EGF and bovine pituitary extract (Invitrogen). Thecell lines obtained directly from the ATCC were subjected to cellline characterization or authentication via short tandem repeatprofiling and passaged in our laboratory for fewer than 6 monthsafter receipt. The following drugs were used in DNA methylationexperiments: 5-aza-20-deoxycytidine (5-aza), MG132, and cyclo-heximide (Sigma-Aldrich). PDAC cell lines were treated with5-aza at 5 mmol/L for 72 hours, and total RNAs were extractedfrom the cells and analyzed using quantitative real-time reversetranscription (RT) PCR.

Human PDAC specimens and immunohistochemical analysisTissuemicroarray (TMA) construction and immunohistochem-

ical analysis were conducted with anti-Merlin and anti-FOXM1antibodies (Santa Cruz Biotechnology) as described previously

(25). The use of human specimens was approved by the relevantinstitutional review boards. The staining results were scored bytwo investigators blinded to the clinical data as described previ-ously (26).

Plasmids and siRNAsThe plasmids pcDNA3.1-FOXM1B (pFOXM1), Flag-tagged

FOXM1B (Flag-FOXM1), and a control vector were describedpreviously (22, 27). The plasmid pcDNA3.0-Flag-tagged Merlinwas the full-length Merlin isoform I and was obtained fromAddgene. Full-length Merlin was cloned into the pcDNA3.0(pMerlin) and pcDNA3.0-hemagglutinin (HA; HA-Merlin) vec-tors as a BamHI-EcoRI fragment. Mutated Merlin-S518A wasgenerated via overlapping PCR and inserted into pcDNA3.0(pM-S518A) and pcDNA3.0-HA (HA-M-S518A) vectors (ampli-fied via two-step PCR using the following primers: 50-atgcggatc-catggccggggccatcgcttcccg-30 and 50-ctttttctttctctatctccatggcaagccgct-tcatgtcagtatctttg-30 plus 50-caaagatactgacatgaagcggcttgccatggagata-gagaaagaaaaag-30 and 50-atgcgatatcctagagctcttcaaagaaggccact-30

for the first reaction and 50-atgcggatccatggccggggccatcgcttcccg-30

and 50-atgcgatatcctagagctcttcaaagaaggccact-30 for the second reac-tion). MutatedMerlin-S518Dwas obtained using the samemeth-od (primers: 50-atgcggatccatggccggggccatcgcttcccg-30 and 50-ctt-tttctttctctatctccatgtcaagccgcttcatgtcagtatctttg-30 plus 50-caaagatact-gacatgaagcggcttgacatggagatagagaaagaaaaag-30 and 50-atgcgatat-cctagagctcttcaaagaaggccact-30 for the first reaction and 50-atgcggatccatggccggggccatcgcttcccg-30 and 50-atgcgatatcctagagctctt-caaagaaggccact-30 for the second reaction). An HA-ubiquitinvector described previously was used as a control (28). Full-lengthMerlin was cloned into the retroviral expression vector pBABE-puro as a BamHI-EcoRI fragment (pBABE-Merlin). Retroviruseswere produced by transfecting packaging cells (HEK293) with athree-plasmid system consisting of the empty pBABEpuro vectoror pBABE-Merlin, the packaging plasmid pUMVC, and the enve-lope plasmid pCMV-VSV-G. pUMVC and pCMV-VSV-G wereobtained from Addgene. Retroviruses were frozen at �80�C forlong-term storage. Each amplified DNA fragment was verified bysequencing the inserts and flanking regions of the plasmids.siRNA#1 targeting Merlin (siMerlin), which was obtained fromCell Signaling Technology, was a pool of siRNAs for the NF2 gene(sense strand: 50-ccuguaaauuuguucuuua-30; antisense strand: 50-uaaagaacaaauuuacagg-30). siRNA#2 targeting Merlin wasdesigned on the basis of the sequences specific to human NF2gene cDNA (sense strand: 50-ggacaagaagguacuggaucaugau-30;antisense strand: 50-aucaugauccaguaccuucuugucc-30), which wasreported previously (29). siRNA targeting p21-activated kinase 1(PAK1) was designed on the basis of the sequences specific tohuman PAK1 gene cDNA (sense strand: 50-cucacacccgcuaggauu-cuucuuu-30; antisense strand: 50-aaagaagaauccuagcgggugugag-30),which was reported previously (30). siRNA targeting FOXM1(siFOXM1) was described previously (25).

Quantitative real-time RT-PCRQuantitative real-time RT-PCR analysis of the expression of

the Merlin, FOXM1, and b-catenin genes was performed usingtotal RNA and the SYBR Green reagent with an ABI Prism7000HT sequence detection system (Applied Biosystems;ref. 28). The sequences of the PCR primers were as follows:Merlin, 50-cggtgtccttgatcgtgtactg-30 (forward) and 50-tcaattgcga-gatgaagtggaa-30 (reverse); FOXM1, 50-acgtccccaagccaggctc-30

(forward) and 50-ctactgtagctcaggaataa-30 (reverse); b-catenin,

Regulation of Pancreatic Cancer Progression by Merlin

www.aacrjournals.org Cancer Res; 75(22) November 15, 2015 4779




50-tgttcgtgcacatcaggatacc-30 (forward) and 50-acatcccgagctag-gatgtgaag-30 (reverse); and glyceraldehyde-3-phosphate dehy-drogenase (GAPDH), 50-acagtccatgccatcactgcc-30 (forward) and50-gcctgcttcaccaccttcttg-30 (reverse).

Methylation-specific PCRMethylation-specific PCRwas performed using gDNA from the

tumor cells, which was modified with bisulfate using an EZ DNAmethylation kit according to the manufacturer's instructions. Fordetection of unmethylated DNA, the forward primer was 50-agttattttaaaggaggtgggatgg-30, and the reverse primer was 50-aaa-caaaacccctaaacaacaa-30. For detection of methylated DNA, theforward primer was 50-gagttattttaaaggaggcgggac-30, and thereverse primer was 50-gaaacccctaaacgacaacgac-30. The primerswere designed using the MethPrimer software program asdescribed previously (31). The negative control was water. Thepositive control was gDNA that was methylated using the CpGmethylase M.Sss I (New England Biolabs). The PCR conditionsconsisted of an initial 5 minutes at 95�C; 40 cycles of 94�C for 25seconds, 58�C for 25 seconds, and72�C for 30 seconds; and afinalextension step at 72�C for 10 minutes. The PCR products wereconfirmed using 2% agarose gel electrophoresis and ethidiumbromide staining.

Gene transfectionFor retroviral transduction, PDAC cells were plated into 6-well

plate 24 hours prior to viral infection (�50% confluent) andwereinfected with a mixture of retroviruses and hexadimethrine bro-mide (Polybrene; 5 mg/mL), and stable populations of the cellswere obtained via selection with 2 mg/mL puromycin. For tran-sient transfection, cells were transfected with plasmids or siRNAfor 48 hours before functional assays using Lipofectamine LTXand Lipofectamine 2000 CD (Invitrogen), respectively. PDACcells treated with the transfection reagents alone were includedas mock controls.

Western blot analysisTotal cell lysates, cytoplasmic, and nuclear protein fractions

were extracted as described previously (20). Standard Westernblotting was carried out using primary anti-Merlin, anti-FOXM1,anti-b-catenin, anti-cyclinD1, anti-c-Myc, anti-mPAR, anti-MMP9,and anti-VEGF (rabbit; Santa Cruz Biotechnology); and anti-MMP2 (rabbit; Cell Signaling Technology) antibodies. Equalprotein sample loading was monitored using an anti-GAPDHantibody for total cell protein lysates (rabbit; Santa Cruz Biotech-nology), an anti-a-tubulin antibody for cytoplasmic fractions(mouse; Oncogene), or an anti-histone H1 antibody for nuclearfractions (mouse; Santa Cruz Biotechnology). Secondary antibo-dies were anti-mouse IgG or anti-rabbit IgG (Santa Cruz Biotech-nology). The bands were quantified using the Quantity OneAnalysis Software Program (version 4.6; Bio-Rad).

Pulse-chase analysisPulse-chase experiments were performed as reported previ-

ously (32). Briefly, BxPC-3 cells were cotransfected withpFOXM1B and a control vector or pMerlin, and PANC-1 cellswere cotransfected with pFOXM1 and siMerlin or a controlvector 24 hours before being subjected to treatment withcycloheximide at a final concentration of 100 mg/mL for indi-cated times. Protein extracts were prepared and analyzed viaWestern blotting for FOXM1 expression.

Immunoprecipitation and ubiquitination assaysHEK293 cells were cotransfected with HA-ubiquitin, Flag-

FOXM1,HA-Merlin, orHA-M-S518A/D as indicated. Thirty hoursafter transfection, the cells were treated with 10 mmol/LMG132, aproteasome inhibitor for 6 hours, and then harvested in immu-noprecipitated lysis buffer (Thermo Scientific). Cell lysates wereincubated with prewashed Flag affinity beads (Sigma) overnightat 4�C. Immunoprecipitates of the lysates were washed four timeswith the lysis buffer, boiled for 5 minutes in a protein loadingbuffer, and analyzed using Western blotting. UbiquitinatedFOXM1 was detected using an HA-tagged antibody (rabbit; CellSignaling Technology).

Colony formation assayTwo hundred cells from each group as indicated were plated in

24-well plates and allowed to grow for 14days in culturemedium;themediumwas changed twice a week. Cells were then fixed with4% paraformaldehyde and stained with 0.1% crystal violet solu-tion for 10 minutes. Colonies (>20 cells) were counted using aninverted microscope at �40 magnification. The results werecalculated as the percentage of proper control. All experimentswere performed in triplicate and repeated twice.

Tumor cell migration/invasion assayBoth cell scratch-wound assay and modified Boyden chambers

assay were performed to determine the migrating and invadingability of PDAC cells with altered Merlin expression as describedpreviously (33).

Animal experimentsFemale pathogen-free athymic nudemice were purchased from

the National Cancer Institute. The animals were maintained infacilities approved by the Association for Assessment and Accred-itation of Laboratory Animal Care International in accordancewith the current regulations and standards of theU.S. Departmentof Agriculture andDepartment of Health andHuman Services. Toanalyze tumorigenicity, pancreatic cancer cells (1�106) in 0.1mLof Hank balanced salt solution were injected subcutaneously intothe right scapular region of nude mice. The tumor-bearing micewere killed when they became moribund or on day 35 afterinoculation and their tumors were removed and weighed.

Statistical analysisThe significance of the patient specimen data was determined

using the Pearson correlation coefficient. The two-tailed c2 test orFisher exact test was used to determine the significance of thedifferences among the covariates. All the in vitro and in vivoexperiments were repeated 2 to 4 times or as indicated, and thedata from independent experiments were shown as mean � SEMor as indicated otherwise, and the significance of the data wasdetermined using the Student t test (two-tailed) or one-wayANOVA. In all of the tests, P < 0.05 was considered statisticallysignificant. The SPSS software program (version 13.0; IBM Cor-poration) was used for statistical analysis.

ResultsReduced expression of Merlin directly associated withpathologic features in PDAC

To determine the role of Merlin expression in PDAC patho-genesis, we first investigated Merlin protein expression in the 154primary PDAC specimens, 34 tumor-adjacent tissue specimens,

Quan et al.

Cancer Res; 75(22) November 15, 2015 Cancer Research4780




and 22 normal pancreatic tissue specimens in the TMA describedpreviously (25). The clinicopathologic characteristics of theTMA specimens are listed in Supplementary Table S1. Weobserved Merlin-positive staining in the cytoplasm and nuclei(Supplementary Fig. S1) and that the expression of Merlin in thetumor-adjacent and normal pancreatic tissue specimens wasmuch higher than that in the tumors (Fig. 1A and B). Merlinexpression was negatively correlated with disease pT classification(P < 0.001), regional lymph node metastasis (P ¼ 0.021),TNM stage (P < 0.001), and tumor differentiation (P < 0.001;

(Supplementary Tables S2 and S3; Fig. 1C and D). Consistently,Merlin protein expression was markedly lower in most of thepancreatic cancer cell lines than that in human pancreatic ductalepithelial and HEK293 cells (Fig. 1E and F). These data indicatedthat Merlin may be a tumor suppressor for PDAC and thatdecreased expression ofMerlinmay promote PDAC developmentand progression.

The frequently reduced expression of Merlin in PDAC cellssuggested a higher rate of NF2 gene inactivation than predictedfrom mutation analysis, because only about 10% of the PDAC

Figure 1.Expression of Merlin and its association with clinicopathologic features of PDAC. TMA PDAC specimenswere immunostained with a specific anti-Merlin antibody. A,representative images of Merlin expression in PDAC specimens. Adjacent normal pancreatic tissue specimens are also shown. B, markedly lower Merlin expressionin tumors (TT) than in adjacent normal tissue (TN) and normal tissue (NN) and no difference in Merlin expression between adjacent normal and normaltissue specimens. C and D, negative association of the expression of Merlin with TNM stage (top) and tumor differentiation (bottom). E, Western blot analysis of theexpression of Merlin in PDAC cell lines. F, quantitative results of the Western blot analysis obtained via densitometric analysis, standardized according toGAPDH expression, and expressed as the percentage of HPDE.






cases had heterozygous losses or inactivating mutations at theNF2 locus in The Cancer Genome Atlas dataset. Previous studiesdemonstrated that hypermethylation of the CpG islands in thepromoter region of the NF2 gene played critical roles indecreased expression of Merlin (34, 35). Indeed, the promoterregion of NF2 contains typical CpG islands (Supplementary Fig.S2A). Methylation-specific PCR revealed that AsPC-1, CaPan-1,and PA-TU-8902 cells exhibited hypermethylation in the pro-moter region ofNF2 (Supplementary Fig. S2B). Furthermore, weincubated 8 PDAC cell lines, which had relatively low expres-sion levels of Merlin, in a medium alone or containing thedemethylating agent 5-aza for 72 hours and analyzed the NF2mRNA expression in them using quantitative PCR. As shown inSupplementary Fig. S2C, exposure to 5-azacytidine producedhigher NF2 mRNA expression in AsPC-1 and PA-TU-8902 cellsthan in their control cells. Therefore, promoter hypermethyla-tion may contribute to reduced Merlin expression in a subset ofPDAC cases.

Merlin suppressed PDAC cell growth, migration, and invasionin vitro and tumorigenicity and metastasis in vivo

To assess the impact of Merlin expression on PDAC biology,we infected AsPC-1, BxPC-3, and FG cells, which have very lowor intermediate levels of endogenous Merlin expression, withretroviruses containing Merlin (AsPC-1/BxPC-3/FG-pBABE-Merlin) and empty retroviral expression vectors used as con-trols (AsPC-1/BxPC-3/FG-pBABEpuro). After selection of theinfected cells with puromycin, we found the pooled drug-resistant cells to have high levels of Merlin expression and thatthe restored Merlin expression was still a little lower than theendogenous level of Merlin expression in HEK293 cells (Sup-plementary Fig. S3A). As shown in Fig. 2A and B, restoredexpression of Merlin significantly suppressed the growth of andcolony formation by PDAC cells (P < 0.05).

Conversely, we decreased Merlin expression using specificsiRNA against Merlin and found that both siRNA#1 and siRNA#2knocked down Merlin expression effectively (SupplementaryFig. S3B). In scratch-wound assay, increased Merlin expressiondecreased the flattening and spreading of BxPC-3 cells, whereasknockdown of Merlin expression promoted the flattening andspreading of PANC-1 cells (P < 0.05). Similarly, increased Merlinexpression suppressed themigration and invasion of BxPC-3 cells,whereas knockdown of Merlin expression attenuated the migra-tion and invasion of PANC-1 cells (P < 0.05; SupplementaryFig. S4A and S4B).

Next, we determined the effect of altered Merlin expression onPDAC growth and metastasis in vivo. We found that increasedexpression of Merlin markedly suppressed tumor growth andexperimental liver metastasis (P < 0.05; Fig. 3A and B). Collec-tively, these data clearly demonstrated that Merlin suppressed thegrowth, migration, and invasion of PDAC cells in vitro and PDACtumorigenicy andmetastasis in vivo, supporting thatMerlinmightfunction as a tumor suppressor for PDAC.

Merlin regulated Wnt/b-catenin pathway in PDAC cellsThe Wnt/b-catenin pathway is one of the most important

signaling pathways in PDAC development and progression(9, 15, 17, 18). To further identify the molecular mechanismsunderlying the tumor-suppressive function of Merlin in PDACcases, we initially focused on the impact of Merlin on Wnt/b-catenin signaling. We found that increased expression ofMerlinin BxPC-3 cells drastically suppressed the activity of the TOPFlashreporter, a tool used to evaluate the transcriptional activity ofb-catenin, and that knockdownof expressionofMerlin inPANC-1cells increased its activity (Fig. 4A). Previous studies demonstratedthe existence of Merlin in both phosphorylation and dephos-phorylation and that the requirement of dephosphorylation forthe tumor-suppressive activity ofMerlin (10, 36). Interconversion

Figure 2.Merlin suppressedPDAC cell growth invitro. AsPC-1 and BxPC-3 cells wereinfected with retroviruses and stablyoverexpressedMerlin. Cell growthwasassessed via cell counting at theindicated time points (A), colonyformation assay was performed in24-well plates, and numbers ofcolonies were counted 14 days afterretroviral infection (B). Datarepresent mean � SEM of threeindependent experiments; � , P valuesare indicated, as compared betweenpBABE-Merlin and Mock orpBABEpuro control.

Quan et al.





between the phosphorylation and dephosphorylation forms ofMerlin is regulated by PAK,which phosphorylatesMerlin at Ser518

and maintains the protein in the inactive form (10, 37). Toinvestigate the effect of Merlin phosphorylation on b-cateninactivity, we generated a Merlin-S518A expression vector, whichcannot be phosphorylated by PAK, and a Merlin-S518D expres-

sion vector, whichmimics the phosphorylated form of Merlin. Asshown in Supplementary Fig. S6A, Merlin-S518A decreased TOP-Flash reporter activity similar to that of wild-type Merlin, whereasMerlin-S518D did not significantly decrease the TOPFlash report-er activity (4, 10).

Further studies demonstrated that increased expression ofMerlin downregulated the expression of the Wnt/b-catenintarget genes cyclin D1, c-Myc, MMP2, MMP9, VEGF, and mPAR,which are the key genes related to cell proliferation, invasion,and metastasis, whereas knockdown of expression of Merlin ledto elevated expression of these genes (Fig. 4B). Consistentwith published studies, increased Merlin expression did notsignificantly reduce b-catenin protein or mRNA expression(Fig. 4B and Supplementary Fig. S5; refs. 15, 18). Interestingly,we observed that increased Merlin expression markedly down-regulated FOXM1 protein expression and that knockdown ofMerlin expression upregulated FOXM1 expression but did notaffect FOXM1 mRNA expression (Fig. 4B and SupplementaryFigs. S3B and S5). Given that the interaction between FOXM1and b-catenin and its critical roles in PDAC growth, angiogen-esis, invasion, and metastasis (18–21), our data suggest that thesuppressive role of Merlin involves in the FOXM1/Wnt/b-cate-nin signaling axis.

Merlin negatively regulated FOXM1 expression in PDAC cellsTo confirm that Merlin negatively regulate FOXM1 expression

in PDAC cells, we further examined the effects of altered expres-sion of Merlin on FOXM1 expression in various PDAC cell lines.Clearly, increased Merlin expression reduced FOXM1 proteinexpression, whereas knockdown of Merlin expression led toupregulation of FOXM1 expression (Supplementary Fig. S6B).Merlin-S518A decreased FOXM1 expression similar to that ofwild-type Merlin whereas Merlin-S518D did not (Fig. 6A), con-firming the role of Merlin phosphorylation in repressing theexpression of FOXM1. Consistently increased expression of Mer-lin markedly decreased the expression of FOXM1 in xenografttumors of BxPC-3 and FG cells (Fig. 5A). Furthermore, expressionof FOXM1wasnegatively associatedwith expressionofMerlin at astatistically significant level (r¼�0.486, P < 0.001; Fig. 5B and C)in the same cohort of human PDAC TMA specimens (Fig. 1).Collectively, these experimental and clinical data strongly sup-ported the notion that Merlin negatively regulated the expressionof FOXM1 and that phosphorylation of Merlin by PAK abolishedthe suppressive effect of Merlin on FOXM1.

Merlin promoted ubiquitination and degradation of FOXM1Because Merlin did not change FOXM1 mRNA expression, we

determined whether Merlin affected FOXM1 protein stability. Weobserved that MG132, a typical proteasome inhibitor, attenuatedthedownregulating effect ofMerlin onFOXM1protein expression(Fig. 6B). To further confirm that Merlin decreased the stability ofFOXM1protein, we investigated the effect ofMerlin expression onthe FOXM1 protein half-lives using pulse-chase analyses. Asshown in Fig. 6C and D, increased Merlin expression shortenedthe half-life of FOXM1 from 3.5 to 1.5 hours in BxPC-3 cells,whereas knockdownofMerlin expression extended the half-life ofFOXM1 from 4 to 8 hours in PANC-1 cells. Next, in proteinubiquitination experiment, we found that ubiquitination ofFOXM1 was increased in the cells with increased Merlin orMerlin-S518A expression but not in that with Merlin-S518D(Fig. 6E). Therefore, Merlin is a negative regulator of FOXM1

Figure 3.Merlin inhibited PDAC growth and metastasis in vivo. A, BxPC-3 and FG cellswith stableMerlin overexpressionwere injected subcutaneously into the rightscapular region in nudemice (n¼ 5). The resulting tumorswere removed fromthemice andweighted.Data representedmean�SDof 5mice per group. Thisexperiment was repeated once with similar results. B, FG cells with stableMerlin overexpression were injected intravenously into the ileocolic vein innude mice (n ¼ 5). Representative gross livers with metastasis andhematoxylin and eosin–stained liver metastasis sections obtained from themice are shown. The numbers in the insets represent the median (range)numbers of liver metastases in each group. This experiment was repeatedonce with similar results. � , P values are indicated, as compared betweenpBABE-Merlin and Mock control or pBABEpuro control.






expression, by enhancing FOXM1ubiquitination and subsequentdegradation.

FOXM1mediated the negative regulatory effect ofMerlin on theWnt/b-catenin signaling

To determine the role of FOXM1 in mediation of the suppres-sive effect of Merlin on Wnt/b-catenin signaling, we reanalyzedthe effect ofMerlin on TOPFlash reporter activitywith FOXM1as arescue. In BxPC-3 cells, restored expression of Merlin suppressedthe activity of the TOPFlash reporter, whereas overexpression ofFOXM1 attenuated this suppressive effect (Fig. 7A). Consistently,in PANC-1 cells, knockdown of FOXM1 and Merlin expression

markedly suppressed the TOPFlash reporter activity, which waselevated by knockdown of Merlin expression (Fig. 7B). Further-more, Wnt3a ligand treatment increased the regulatory effect ofMerlin and Merlin plus FOXM1 on TOPFlash reporter activity(Fig. 7A and B).

Moreover, subcellular protein fractionation study confirmedthe presence of Merlin in the nucleus in positive lines; increasedMerlin expression reduced both cytoplasmic and nuclear FOXM1protein expression and drastically reduced nuclear b-cateninexpression (Fig. 7C), whereas knockdown of Merlin did theopposite (Fig. 7D). Consistent with the results of TOPFlashreporter activity assay, FOXM1 overexpression attenuated thesuppressive effect of Merlin on nuclear translocation of b-catenin,and knockdown of FOXM1 reduced the nuclear expression ofb-catenin, which was partially reversed by downregulation ofMerlin expression (Fig. 7C and D). These data demonstrated thatFOXM1 plays a critical role in mediation of the suppressive effectof Merlin expression on the Wnt/b-catenin pathway.

In addition to being phosphorylated by PAK, Merlin is aninhibitor of PAK activity, and knockdown of Merlin expressionleads to increased PAK activation because of the presence of anegative feedback loop (37). Previous studies also demonstratedthat Merlin inactivated Wnt/b-catenin signaling via the Rac/PAKpathway (18). We therefore examined the role of PAK1 in medi-ation of the suppressive function of Merlin on Wnt/b-cateninsignaling. The results demonstrated that knockdown of Merlinexpression in PANC-1 cells resulted in increased b-catenin tran-scriptional activity, whereas knockdown of PAK1 and Merlinexpression rescued increased TOPFlash activity (SupplementaryFig. S6C). Collectively, these data demonstrated that both FOXM1and PAK1 played important roles inmediation of the suppressiveeffect of Merlin expression on Wnt/b-catenin signaling in PDACcells.

DiscussionIn the present study, we obtained five lines of experimental and

clinical evidence supporting the tumor-suppressive role of Merlinby aberrant activation of FOXM1/b-catenin signaling in PDACdevelopment and progression. First, the expression of Merlin wasmarkedly downregulated in PDAC and negatively associated withdisease pT classification, regional lymph node metastasis, TNMstage, and tumor differentiation; Merlin in the nucleus in vivo andthat loss of nuclear staining predicted higher grade as well as lossof total staining. Second, restored expression of Merlin inhibitedPDAC cell growth, invasion,migration, andmetastasis in vitro andin vivo. Third, expression of Merlin was markedly downregulatedin PDAC cells, and Merlin negatively regulated the expression ofFOXM1 by promoting ubiquitination and degradation ofFOXM1. In addition, the expression of Merlin was inverselycorrelated with that of FOXM1 in PDAC specimens. Fourth,increased Merlin expression reduced TOPFlash reporter activityand nuclear b-catenin andWnt/b-catenin downstream target geneexpression. Fifth, FOXM1 attenuated the suppressive effect ofMerlin on the activity of the TOPFlash reporter and nucleartranslocation of b-catenin. Collectively, these findings demon-strated novel Merlin/FOXM1/b-catenin signaling in PDAC cells,which contributed to pancreatic tumorigenesis.

NF2 gene is first identified in 1993 and classified as a tumorsuppressor for two reasons. First, an inactivatingmutation of NF2led to loss of the Merlin protein and may occur in schwannoma,

Figure 4.Merlin suppressed Wnt/b-catenin signaling in PDAC cells. A, TOPFlashreporter activity assay. BxPC-3 cells were cotransfected with 0.2 mg ofTOPFlash reporter and 0, 0.3, or 0.6 mg of pMerlin (pMer) or a control vector(pCtr). PANC-1 cells were cotransfected with 0.2 mg of TOPFlash reporter and0, 25, or 50 nmol/L siMerlin (siMer) or control siRNA (siCtr). Promoteractivity (mean � SD) was examined using a dual luciferase assay kit. Datarepresentedmean� SEM of three independent experiments. � , P values wereindicated, as compared between pMer and pCtr control or between siMerlinand control siRNA. B, BxPC-3 and AsPC-1 cells were transfected with pMerlinor a control vector for 48 hours, and PANC-1 cells were transfected withsiMerlin or a control siRNA for 48 hours. Total cell proteins were extractedand expression levels of Merlin and others were measured using Westernblotting. This experiment was repeated once with similar results.

Quan et al.





meningioma, hamartoma, and ependymoma cases sporadicallyor as part of the autosomal dominant syndrome neurofibroma-tosis type 2 (7). Second, expression of Merlin was lost in manytypes of cancer, and modulated several signaling pathways thatplayed critical roles in cancer cell proliferation, motility, andapoptosis (13). In PDAC cells, Merlin expression was elevatedduring cell-cycle progression and mitotic arrest; furthermore,overexpression of Merlin inhibited SW1990 cell proliferation,migration, and adhesion in vitro (14, 38). Thesefindings indicateda tumor-suppressive role for Merlin expression in PDAC. In thepresent study, we systematically analyzed the Merlin expressionpatterns in PDAC cell lines and tumor specimens. The resultsrevealed that expression of Merlin was downregulated in most ofthe cell lines. In addition, statistical analysis of the relationshipsamongMerlin expression and clinicohistopathologic PDAC para-meters demonstrated drastically decreased Merlin expression inPDAC and that the relationships were closely correlated with themalignant features of PDAC. In functional studies, our in vitro andin vivo experimental results revealed that Merlin overexpression

suppressed PDAC cell growth and metastasis, which was consis-tent with the role ofMerlin in other types of cancer (9, 11, 39, 40).Taken together, these clinical and experimental findings stronglyand clearly supported Merlin as a tumor suppressor in PDAC.Moreover, reduced expressionofMerlin innervous system tumorspredominantly resulted from mutations, LOH, and hypermethy-lationofCpG islands in the promoter regionof theNF2 gene (35).Also, in breast cancer, osteopontin-initiated, Akt-mediated phos-phorylation of Merlin contributed to loss expression of Merlin(12, 39, 41–43). Because only 10% of PDAC harbored hetero-zygous losses or mutations of theNF2 gene, we analyzed the roleof hypermethylation in Merlin expression in PDAC cell lines. Weobserved that DNA hypermethylation contributed to reducedexpression of Merlin in a subset of PDAC cases. However, histonemethylation may contribute little to the reduced expression ofMerlin in pancreatic cancer (Supplementary Fig. S7). Our resultsare consistent with the findings from prior reports (44, 45).

Mechanisms underlying the suppressive function of Merlin areunclear. The Wnt/b-catenin pathway is widely dysregulated in

Figure 5.Merlin suppressed FOXM1 expression in vivo. A, subcutaneous tumor specimens were stained immunohistochemically with specific anti-Merlin and anti-FOXM1antibodies. Shown are representative photographs of the expression of Merlin and FOXM1 in BxPC-3 and FG xenograft tumors. B, Merlin and FOXM1 proteinexpression TMA tissue sections were represented from the cohort described in Fig. 1. C, negative correlation of Merlin expression with FOXM1 expression wasassessed using Pearson correlation coefficient analysis (n¼ 154; r¼�0.486; P < 0.001). Some of the dots on the graph represent more than one specimen and thenumbers of overlapping scores were then labeled.






cancer, and it is one of the pivotal drivers of pancreatic tumor-igenesis (15, 46). Stabilization and nuclear translocation ofb-catenin are the core events in persistent activation of thispathway. In colorectal cancer cases, high levels of nuclear b-cate-nin result frommutations in Wnt/b-catenin–associated signalingcascades (47). However, in PDAC, mutations in the Wnt/b-cate-nin pathway are seldom detected, and b-catenin does not containa nuclear localization sequence. Thus, cross-talkwith other factorsand epigenetic modulations play key roles in activation of Wnt/

b-catenin signaling (15). In previous studies, we demonstratedthat FOXM1 bound directly to b-catenin and facilitated nucleartranslocation of it (3, 22). Authors also recently reported thatMerlin inhibited the Wnt/b-catenin pathway and suppresseddownstream target gene expression (9, 17, 18, 48). Mechanisti-cally, Merlin regulated the formation of adherence junctions bydirectly binding with b-catenin and the E/N-cadherin complex atthe plasmamembrane. In addition, Merlin decreased the numberof Frizzled-1 receptors and reduced Wnt signaling. In the present

Figure 6.Merlin promoted ubiquitination anddegradation of FOXM1 in PDAC cells. A, BxPC-3 andAsPC-1 cellswere transfectedwith pMerlin, pMerlin-S518A, pMerlin-S518D, ora control vector. Total proteinwas harvested 48 hours after transfection tomeasureMerlin and FOXM1 expression usingWestern blotting. B, BxPC-3 andAsPC-1 cellstransfected with pMerlin or a control vector were treated with or without MG132. Merlin and FOXM1 protein expression was determined using Westernblotting. C, BxPC-3 cells cotransfected with pFOXM1 and pMerlin or pControl were treated with or without cycloheximide (CHX) for the indicated times. Total celllysates were used to measure FOXM1 using Western blotting. Semiquantification was performed using GAPDH as a loading control and relative FOXM1expression levels at time 0 was set as 100%. D, PANC-1 cells cotransfected with pFOXM1 and siMerlin or siControl were treated with or without cycloheximidefor the indicated times. The FOXM1 was determined as described above. E, HEK293 cells cotransfected with Flag-FOXM1, HA-ubiquitin, and HA-Merlin/HA-Merlin-S518A/HA-Merlin-S518D/Control were lysed in IP lysis buffer. Overexpressed Flag-FOXM1 protein was captured using Flag affinity beads andimmunoblot analysis with an HA-tag antibody. All experiments were performed two times with similar results.

Quan et al.





study, we observed for the first time that Merlin suppressedTOPFlash reporter activity and Wnt downstream target geneexpression in PDAC cells. Furthermore, Merlin inhibited thenuclear translocation of b-catenin and activation of Wnt/b-cate-nin signaling via downregulation of FOXM1protein expression inPDAC cells, as overexpression of FOXM1 attenuated the suppres-

sive effect of Merlin on Wnt/b-catenin signaling. This novelfinding regarding Merlin/FOXM1/b-catenin signaling is criticalto identifying themolecular mechanisms underlying PDAC path-ogenesis given that the Wnt/b-catenin pathway is one of the keydrivers of PDAC initiation and progression and that expression ofMerlin is generally downregulated in PDAC cells.

Figure 7.Merlin/FOXM1 signaling regulatesthe Wnt/b-catenin pathway. A,TOPFlash reporter wascotransfected into BxPC-3 cells witha control vector, pMerlin, or pMerlinplus pFOXM1 with or without Wnt3atreatment. The resulting promoteractivity (mean � SD of triplicates;� , Student t test) was analyzed usinga dual luciferase assay. Datarepresent mean � SEM of threeindependent experiments. � , P valuesare indicated, as compared betweengroups indicated. B, TOPFlashreporter was cotransfected intoPANC-1 cells with control siRNA,siMerlin, or siMerlin plus siFOXM1withor without Wnt3a treatment, and theresulting promoter activity wasanalyzed. C and D, a control vector,pMerlin, or pMerlin plus pFOXM1 wastransfected into BxPC-3 cells (C) andcontrol siRNA, siMerlin, or siMerlinplus siFOXM1 was transfected intoPANC-1 cells (D) for 48 hours.Cytosolic and nuclear proteins wereextracted, and the levels of Merlin,FOXM1, and b-catenin proteinexpression were measured usingWestern blotting. These experimentswere repeated once with similarresults. E, model of Merlininactivating FOXM1/b-cateninsignaling in PDAC cells. Loss ofMerlinstabilizes FOXM1 protein andsubsequently enhances the Wnt/b-catenin pathway, resulting inincreased expression of Wntdownstream target genes andpromoting PDAC development andprogression.






FOXM1 is a key regulator of tumor growth, angiogenesis,metastasis, and metabolism via transduction of upstream sig-nals to downstream effectors (21, 25). FOXM1 is overexpressedin PDAC cells. However, the upstream regulators of FOXM1expression are still unknown. A study of malignant mesothe-lioma demonstrated that the Hippo pathway transcriptionalcoactivator Yes-associated protein induced cell proliferation bytranscriptionally upregulating the expression of FOXM1 (49).Furthermore, previous studies demonstrated that Merlin was anupstream regulator of the Hippo pathway and inhibited theactivity of Yes-associated protein (13). However, in the presentstudy, we found that Merlin did not regulate the expression ofFOXM1 at the transcriptional level via Hippo signaling, giventhat altered expression of Merlin did not change the FOXM1mRNA expression. Instead, we found that Merlin regulated theexpression of FOXM1 by promoting ubiquitination and deg-radation of FOXM1. Recently, authors reported that Merlintranslocated to the nucleus and bound directly to the E3ubiquitin ligase CRL4DCAF1. This binding inhibited the activityof CRL4DCAF1 and resulted in suppression of oncogenic geneexpression and, finally, inhibition of tumorigenesis (11, 13).Also, Merlin may indirectly regulate modification of FOXM1protein, which decreases the stability of FOXM1. However, thedetailed mechanisms underlying the regulation of Merlin ofFOXM1 protein stability warrant further elucidation.

In summary, the present study confirmed that Merlin is animportant tumor suppressor in PDAC and identified novel Mer-lin/FOXM1/b-catenin signaling in PDAC pathogenesis. Given theimportance of FOXM1 expression and the Wnt/b-catenin signal-ing pathway in PDAC development and progression (50), ourfindings not only provide further understanding of the molecularmechanisms underlying FOXM1/b-catenin signaling activation

but also identify aberrantMerlin/FOXM1/b-catenin signaling as apromising new therapeutic target for PDAC.

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

Authors' ContributionsConception and design: M. Quan, J. Cui, K. XieDevelopment of methodology: J. Cui, T. Xia, Z. Jia, D. Wei, S. ZhengAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):M. Quan, J. Cui, T. Xia, Z. Jia, D. Xie, D. Wei, S. ZhengAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis):M.Quan, J. Cui, T. Xia, Z. Jia,D. Xie,D.Wei, S.Huang,S. Zheng, K. XieWriting, review, and/or revision of themanuscript:M.Quan, J. Cui, S. Huang,Q. Huang, K. XieAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): S. Zheng, K. XieStudy supervision: K. Xie

AcknowledgmentsThe authors thank Don Norwood for editorial assistance and Xuemei Wang,

Associate Director ofQuantitative Research, for assistance in statistical analyses.

Grant SupportThis work was supported by grants R01-CA129956, R01-CA148954,

R01-CA152309, and R01-CA172233 from the National Cancer Institute, NIH(K. Xie); grant 81460461 (S. Zheng); and grant 81120108017 from theNationalNatural Science Foundation of China (Q. Huang).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 8, 2014; revised July 27, 2015; accepted August 20, 2015;published OnlineFirst October 19, 2015.

References1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer

statistics. CA Cancer J Clin 2011;61:69–90.2. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin

2014;64:9–29.3. Hidalgo M. Pancreatic cancer. N Engl J Med 2010;362:1605–17.4. Cooper J, Giancotti FG. Molecular insights into NF2/Merlin tumor sup-

pressor function. FEBS Lett 2014;588:2743–52.5. Li W, Cooper J, Karajannis MA, Giancotti FG. Merlin: a tumour suppressor

with functions at the cell cortex and in the nucleus. EMBO Rep 2012;13:204–15.

6. Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP, Parry DM,et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for theneurofibromatosis 2 tumor suppressor. Cell 1993;72:791–800.

7. Pecina-Slaus N. Merlin, the NF2 gene product. Pathol Oncol Res 2013;19:365–73.

8. Morris ZS, McClatchey AI. Aberrant epithelial morphology and persistentepidermal growth factor receptor signaling in a mouse model of renalcarcinoma. Proc Natl Acad Sci U S A 2009;106:9767–72.

9. Lau YK, Murray LB, Houshmandi SS, Xu Y, Gutmann DH, Yu Q.Merlin is a potent inhibitor of glioma growth. Cancer Res 2008;68:5733–42.

10. Sher I, Hanemann CO, Karplus PA, Bretscher A. The tumor suppressormerlin controls growth in its open state, and phosphorylation converts itto a less-active more-closed state. Dev Cell 2012;22:703–5.

11. LiW,YouL,Cooper J, SchiavonG, Pepe-CaprioA, ZhouL, et al.Merlin/NF2suppresses tumorigenesis by inhibiting the E3 ubiquitin ligase CRL4(DCAF1) in the nucleus. Cell 2010;140:477–90.

12. Morrow KA, Das S, Metge BJ, Ye K, Mulekar MS, Tucker JA, et al. Loss oftumor suppressor Merlin in advanced breast cancer is due to post-trans-lational regulation. J Biol Chem 2011;286:40376–85.

13. Cooper J, LiW,YouL, SchiavonG,Pepe-CaprioA, ZhouL, et al.Merlin/NF2functions upstream of the nuclear E3 ubiquitin ligase CRL4DCAF1 tosuppress oncogenic gene expression. Sci Signal 2011;4:pt6.

14. Zhou J, Feng Y, Tao K, Su Z, Yu X, Zheng J, et al. The expression andphosphorylation of ezrin and merlin in human pancreatic cancer. Int JOncol 2014;44:2059–67.

15. Morris JPt, Wang SC, Hebrok M. KRAS, Hedgehog, Wnt and the twisteddevelopmental biology of pancreatic ductal adenocarcinoma. Nat RevCancer 2010;10:683–95.

16. Cui J, Jiang W, Wang S, Wang L, Xie K. Role of Wnt/beta-cateninsignaling in drug resistance of pancreatic cancer. Curr Pharm Des 2012;18:2464–71.

17. Bosco EE, Nakai Y, Hennigan RF, Ratner N, Zheng Y. NF2-deficient cellsdepend on the Rac1-canonical Wnt signaling pathway to promote the lossof contact inhibition of proliferation. Oncogene 2010;29:2540–9.

18. Zhou L, Ercolano E, Ammoun S, SchmidMC, BarczykMA, Hanemann CO.Merlin-deficient human tumors show loss of contact inhibition andactivation of Wnt/beta-catenin signaling linked to the PDGFR/Src andRac/PAK pathways. Neoplasia 2011;13:1101–12.

19. Clark KL,Halay ED, Lai E, Burley SK. Co-crystal structure of theHNF-3/forkhead DNA-recognition motif resembles histone H5. Nature 1993;364:412–20.

20. Huang C, Qiu Z, Wang L, Peng Z, Jia Z, Logsdon CD, et al. A novel FoxM1-caveolin signaling pathway promotes pancreatic cancer invasion andmetastasis. Cancer Res 2012;72:655–65.

21. Huang C, Du J, Xie K. FOXM1 and its oncogenic signaling in pancreaticcancer pathogenesis. Biochim Biophys Acta 2014;1845:104–16.

22. Zhang N, Wei P, Gong A, Chiu WT, Lee HT, Colman H, et al. FoxM1promotes beta-catenin nuclear localization and controls Wnt target-geneexpression and glioma tumorigenesis. Cancer Cell 2011;20:427–42.


Quan et al.




23. Li L, Li Z, Kong X, Xie D, Jia Z, Jiang W, et al. Downregulation ofMicroRNA 494 via loss of SMAD4 increases FOXM1 and betacateninsignaling in pancreatic ductal adenocarcinoma cells. Gastroenterology2014;147:485–97.

24. Vezeridis MP, Tzanakakis GN, Meitner PA, Doremus CM, Tibbetts LM,Calabresi P. In vivo selection of a highly metastatic cell line from a humanpancreatic carcinoma in the nude mouse. Cancer 1992;69:2060–3.

25. Cui J, Shi M, Xie D, Wei D, Jia Z, Zheng S, et al. FOXM1 promotes thewarburg effect and pancreatic cancer progression via transactivation ofLDHA expression. Clin Cancer Res 2014;20:2595–606.

26. Li Z, Jia Z, Gao Y, Xie D,Wei D, Cui J, et al. Activation of vitaminD receptorsignaling downregulates the expression of nuclear FOXM1 proteinand suppresses pancreatic cancer cell stemness. Clin Cancer Res 2015;21:844–53.

27. Liu M, Dai B, Kang SH, Ban K, Huang FJ, Lang FF, et al. FoxM1B isoverexpressed in human glioblastomas and critically regulates the tumor-igenicity of glioma cells. Cancer Res 2006;66:3593–602.

28. Xue J, Lin X, Chiu WT, Chen YH, Yu G, Liu M, et al. Sustained activation ofSMAD3/SMAD4 by FOXM1 promotes TGF-beta-dependent cancer metas-tasis. J Clin Invest 2014;124:564–79.

29. Lee JY, Moon HJ, Lee WK, Chun HJ, Han CW, Jeon YW, et al. Merlinfacilitates ubiquitination and degradation of transactivation-responsiveRNA-binding protein. Oncogene 2006;25:1143–52.

30. Zhu G,Wang Y, Huang B, Liang J, Ding Y, Xu A, et al. A Rac1/PAK1 cascadecontrols beta-catenin activation in colon cancer cells. Oncogene 2012;31:1001–12.

31. Li LC, Dahiya R. MethPrimer: designing primers for methylation PCRs.Bioinformatics 2002;18:1427–31.

32. Hwang YC, Yang CH, Lin CH, Ch'ang HJ, Chang VH, Yu WC. Destabili-zation of KLF10, a tumor suppressor, relies on thr93 phosphorylation andisomerase association. Biochim Biophys Acta 2013;1833:3035–45.

33. Khoo CP, Micklem K, Watt SM. A comparison of methods for quantifyingangiogenesis in the Matrigel assay in vitro. Tissue Eng Part C Methods2011;17:895–906.

34. Kino T, Takeshima H, Nakao M, Nishi T, Yamamoto K, Kimura T, et al.Identification of the cis-acting region in the NF2 gene promoter as apotential target for mutation and methylation-dependent silencing inschwannoma. Genes Cells 2001;6:441–54.

35. Gonzalez-Gomez P, Bello MJ, Alonso ME, Lomas J, Arjona D, Campos JM,et al. CpG island methylation in sporadic and neurofibromatis type 2-associated schwannomas. Clin Cancer Res 2003;9:5601–6.

36. Sherman L, Xu HM, Geist RT, Saporito-Irwin S, Howells N, Ponta H, et al.Interdomain bindingmediates tumor growth suppression by theNF2 geneproduct. Oncogene 1997;15:2505–9.

37. Kissil JL,Wilker EW, JohnsonKC, EckmanMS, YaffeMB, Jacks T.Merlin, theproduct of the Nf2 tumor suppressor gene, is an inhibitor of the p21-activated kinase, Pak1. Mol Cell 2003;12:841–9.

38. Kokkinakis DM, Liu X, Neuner RD. Modulation of cell cycle and geneexpression in pancreatic tumor cell lines by methionine deprivation(methionine stress): implications to the therapy of pancreatic adenocar-cinoma. Mol Cancer Ther 2005;4:1338–48.

39. James MF, Han S, Polizzano C, Plotkin SR, Manning BD, Stemmer-Rachamimov AO, et al. NF2/merlin is a novel negative regulator of mTORcomplex 1, and activation of mTORC1 is associated withmeningioma andschwannoma growth. Mol Cell Biol 2009;29:4250–61.

40. Zhan Y, Modi N, Stewart AM, Hieronimus RI, Liu J, Gutmann DH, et al.Regulation of mixed lineage kinase 3 is required for Neurofibromatosis-2-mediated growth suppression in human cancer. Oncogene 2011;30:781–9.

41. Nunes F, Shen Y, Niida Y, Beauchamp R, Stemmer-Rachamimov AO,Ramesh V, et al. Inactivation patterns of NF2 and DAL-1/4.1B (EPB41L3)in sporadic meningioma. Cancer Genet Cytogenet 2005;162:135–9.

42. Wallace AJ,WatsonCJ, Oward E, EvansDG, Elles RG.Mutation scanning oftheNF2 gene: an improved service based onmeta-PCR/sequencing, dosageanalysis, and loss of heterozygosity analysis. Genet Test 2004;8:368–80.

43. Lasota J, Wasag B, Dansonka-Mieszkowska A, Karcz D, Millward CL, Rys J,et al. Evaluation of NF2 and NF1 tumor suppressor genes in distinctivegastrointestinal nerve sheath tumors traditionally diagnosed as benignschwannomas: s study of 20 cases. Lab Invest 2003;83:1361–71.

44. Bracken AP, Dietrich N, Pasini D, Hansen KH, Helin K. Genome-widemapping of Polycomb target genes unravels their roles in cell fate transi-tions. Genes Dev 2006;20:1123–36.

45. Tiwari N, Tiwari VK, Waldmeier L, Balwierz PJ, Arnold P, Pachkov M, et al.Sox4 is a master regulator of epithelial-mesenchymal transition by con-trolling Ezh2 expression and epigenetic reprogramming. Cancer Cell2013;23:768–83.

46. Reya T, Clevers H. Wnt signalling in stem cells and cancer. Nature 2005;434:843–50.

47. White BD, Chien AJ, Dawson DW. Dysregulation of Wnt/beta-cateninsignaling in gastrointestinal cancers. Gastroenterology 2012;142:219–32.

48. Lallemand D, Curto M, Saotome I, Giovannini M, McClatchey AI. NF2deficiency promotes tumorigenesis and metastasis by destabilizing adhe-rens junctions. Genes Dev 2003;17:1090–100.

49. Mizuno T, Murakami H, Fujii M, Ishiguro F, Tanaka I, Kondo Y, et al. YAPinduces malignant mesothelioma cell proliferation by upregulating tran-scription of cell cycle-promoting genes. Oncogene 2012;31:5117–22.

50. Xie D, Xie K. Pancreatic cancer stromal biology and therapy. Genes Dis2015;2:133–143.






2015;75:4778-4789. Published OnlineFirst October 19, 2015.Cancer Res Ming Quan, Jiujie Cui, Tian Xia, et al.

-Catenin Signalingβby Attenuating the FOXM1-Mediated Wnt/Merlin/NF2 Suppresses Pancreatic Tumor Growth and Metastasis

Updated version

10.1158/0008-5472.CAN-14-1952doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/10/17/0008-5472.CAN-14-1952.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/22/4778.full#ref-list-1

This article cites 50 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/75/22/4778.full#related-urls

This article has been cited by 1 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/75/22/4778To 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-14-1952

http://cancerres.aacrjournals.org/content/suppl/2015/10/17/0008-5472.CAN-14-1952.DC1

http://cancerres.aacrjournals.org/content/75/22/4778.full#ref-list-1

http://cancerres.aacrjournals.org/content/75/22/4778.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/22/4778


merlin/nf2 suppresses pancreatic tumor growth and metastasis … · indicated that merlin...

Documents