oncogenic kras and p53 loss drive gastric tumorigenesis ......tumor and stem cell biology oncogenic...

Tumor and Stem Cell Biology

Oncogenic KRAS and p53 Loss Drive GastricTumorigenesis in Mice That Can Be Attenuated byE-Cadherin ExpressionJacob E. Till1, Changhwan Yoon2, Bang-Jin Kim1, Kerry Roby1, Prince Addai1,Evan Jonokuchi1, Laura H. Tang3, Sam S.Yoon2, and Sandra Ryeom1

Abstract

Gastric adenocarcinoma is the third leading cause of cancer-related death worldwide, but no models exist to readily investi-gate distant metastases that are mainly responsible for mortalityin this disease. Here we report the development of a geneticallyengineered mouse model of gastric adenocarcinoma tumorigen-esis based on KrasG12D expression plus inactivation of E-cadherin(Cdh1) and p53 in the gastric parietal cell lineage. Intestinal anddiffuse gastric tumors arise rapidly in this model that displays amedian survival of 76 days. Tumors occur throughout the stom-

ach,withmetastases documented in lymphnodes, lung, and liver.Mice otherwise identical but retaining one wild-type Cdh1 alleleexhibited longer survival with only 20% penetrance of invasivetumors and no apparent lung or liver metastases. Notably,increased RAS activity and downstream MAPK signaling wasobserved in stomachs only when E-cadherin was absent. Thismodel offers a valuable tool to investigate gastric adenocarcinomasubtypes where RAS/MAPK pathway activation and E-cadherinattenuation are common. Cancer Res; 77(19); 5349–59. �2017 AACR.

IntroductionEach year, there are one million cases of gastric cancer world-

wide and over 700,000 deaths making it the fifth most commoncancer and the third leading cause of cancer-related deaths (1).Gastric adenocarcinomas comprise themajority of gastric cancers.Except in a few countries where there is endoscopic screening forgastric adenocarcinoma, the majority of patients present withadvanced or metastatic disease (2). Overall survival for patientswith metastatic disease is 3–5 months with best supportive care.The response rate to multiagent chemotherapy is 50% or greater,but nearly all patients develop chemotherapy resistance, andmedian survival is extended only to 9–11 months (3).

In 1965, Lauren described two distinct histologic types ofgastric adenocarcinomas: intestinal and diffuse (4). Morerecently, a third category referred to as "mixed-type" has beenadded to the Lauren classification to include neoplasms withboth intestinal- and diffuse-type characteristics (5). The intes-tinal type exhibits components of glandular, solid, or intestinalarchitecture as well as tubular structures. The diffuse typedemonstrates single cells or poorly cohesive cells infiltratingthe gastric wall, and progressive disease can ultimately lead tolinitis plastica (a.k.a. leather bottle stomach). The tumor sup-pressor p53 (TP53) is altered in about 50% of gastric adeno-

carcinomas, with these alterations more common in intestinalthan diffuse tumors (6, 7). The diffuse type is often character-ized by loss of E-cadherin, encoded by the CDH1 gene, due tomutation or hypermethylation. A germline mutation in CDH1leads to a lifetime risk of hereditary diffuse gastric cancer of over80% (8).

Current animal models of gastric adenocarcinoma have longlatency periods and do not adequately recapitulate advanceddisease because of the lack of disseminated metastatic disease orthe presence of metastasis only to regional nodes (9). To furtherour understanding of the cellular and molecular mechanismsunderlying gastric adenocarcinoma progression and to developnew treatments for gastric adenocarcinoma, there is a need for abetter mouse model that more closely phenocopies advancedhuman gastric adenocarcinoma.

A previously published mouse model of gastric adenocarci-noma was generated by conditional deletion of E-cadherin(Cdh1) and p53 (Trp53) driven by ATP4B-Cre, expressed exclu-sively in the gastric parietal cell lineage, including pre-parietalcells. Invasive gastric adenocarcinoma was detected in 25% ofthese mice at 9 months with 40% of mice developing metastasisto regional lymph nodes at one year (10). Molecular charac-terization of gastric adenocarcinoma patient samples by TCGAnetwork identified other mutations in these tumors including60% of patient samples harboring alterations in the RTK/Rassignaling pathway. The EGFR family (EGFR, ERBB2, ERBB3,and ERBB4) was amplified or mutated in 44% of these casesand KRAS amplified or mutated in 6% or 11% of cases,respectively, with the vast majority of KRASmutations in codon12 or 13 (6, 7). On the basis of these data, we investigated thecontribution of oncogenic Kras to the progression of gastricadenocarcinoma. We generated a mouse model of gastricadenocarcinoma driven by gastric parietal cell lineage–specific(Atp-4b-Cre) expression of oncogenic KRAS (LSL-KrasG12D)combined with loss of E-cadherin (Cdh1fl/fl), loss of p53(Trp53fl/fl), and expression of a yellow fluorescent protein (YFP)

1Department of Cancer Biology, Perelman School ofMedicine at theUniversity ofPennsylvania, Philadelphia, Pennsylvania. 2Department of Surgery, MemorialSloan Kettering Cancer Center, New York, NewYork. 3Department of Pathology,Memorial Sloan Kettering Cancer Center, New York, New York.

Corresponding Author: Sandra Ryeom, Department of Cancer Biology, Perel-man School of Medicine at the University of Pennsylvania, 421 Curie Boulevard,Philadelphia, PA 19104. Phone: 215-573-5857; Fax: 215-573-2014; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-17-0061

�2017 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 5349

on May 15, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 31, 2017; DOI: 10.1158/0008-5472.CAN-17-0061

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-17-0061&domain=pdf&date_stamp=2017-9-18

http://cancerres.aacrjournals.org/

reporter allele (Rosa26LSL-YFP) and refer to this model as tripleconditional (Tcon) mice.

Expression of oncogenic Kras and loss of Trp53 in mice issufficient for rapid tumorigenesis in most tissues as has beenshown by mouse models of lung cancer (11), pancreatic ductaladenocarcinoma (12), squamous cell carcinoma (13), acutemye-loid leukemia (14), and undifferentiated pleomorphic sarcoma(15, 16). To determine whether oncogenic Kras and Trp53 lossis sufficient for gastric tumorigenesis, we also generated micewith oncogenic Kras and loss of Trp53 but included one wild-type allele of E-cadherin and refer to this model as doubleconditional E-cadherin heterozygous mice (Dcon-E-cad het).Here we show that 100% of our Tcon mice develop mixed-typegastric tumors with a median survival of 76 days. Both intestinaland diffuse gastric adenocarcinoma tumors are observed through-out the stomach as well as YFPþ metastases to nodes, lung, andliver. In contrast, mice with KrasG12D, p53 loss and a single copy ofCdh1 (Dcon-E-cad het) only 20% of mice developed invasivegastric adenocarcinoma with median survival of these tumor-bearing mice increased to 123 days with no gross lung or livermetastases. Our data show increased KRAS activity and MAPKsignaling in stomach tumors isolated from Tcon mice as com-pared with stomachs isolated from Dcon-Ecad het mice.

Materials and MethodsAnimals

The University of Pennsylvania Animal Care and Use Commit-tee approved all animal studies. Tcon mice were generated bycross breeding B6.FVB-Tg-Cdh1fl/f (N1) (from The Jackson Labo-ratory, #005319; ref. 17), B6.129-Tg-LSL-KrasG12D (18); Trp53fl/fl

(N10) (from Dr. Tyler Jacks, MIT, Boston, MA; ref. 19), andB6.129-Tg(Rosa26LSL-YFP) (from Dr. Ben Stanger, Universityof Pennsylvania School of Medicine, Philadelphia, PA; ref. 20)conditional mutants with B6.129-Tg-Atp4b-Cre (21) transgenicmice (from the Gordon Laboratory, Washington University in St.Louis,MO). For survival analysis, endpoint was reached at naturaldeath or humane euthanasia at loss of 20% maximum bodyweight. To account for the mixed background, ten independentbreeding pairswere utilized, each contributing two sibling pairs ofTcon and Dcon-Ecad het (Atp4b-Cre;Cdh1;Trp53;Rosa26LSL-YFP)mice. Ten of the Tcon mice from this cohort were analyzed formetastases. For analysis of gastric cancer progression, 5 Tcon andTcon-Cre negative (Cdh1fl/fl; LSL-KrasG12D;Trp53fl/fl; Rosa26LSL-YFP)matched control mice were sacrificed at 3, 6, and 9 weeks. Toaccount for mixed background, five independent sets of Tcon andTcon-Cre–negative control mice were used from five independentbreeding pairs. Treatment of Tcon mice with a MEK inhibitor(PD0325901, APExBIO) was initiated in 4-week-old mice. Thedrug was administered ad libitum in the mouse chow (Purina5010) at 7 mg/kg (incorporation by Research Diets Inc). ForDcon-Ecad het mice, endpoint was reached at natural death orhumane euthanasia at loss of 20%maximum body weight. Someanimals required humane euthanasia for subcutaneous tumorburden > 2 cm3 and were censored in the survival analysis. Toaccount for the mixed background, at least 10 independentbreeding pairs were utilized in generating the cohorts.

Histology and immunofluorescenceTissues were dissected from indicated mice, fixed in 10%

formalin in PBS, dehydrated, embedded in paraffin, sectioned at5 mm and stained with hematoxylin and eosin. For immunoflu-

orescence, sectionswere dewaxed, rehydrated, and antigen retriev-al was performed in a 2100Retriever (AptumBiologics Ltd.) usingAntigen Unmasking Solution (Vector Labs), and permeabilizedwith 0.2% Triton-X in PBS. Sections were incubated with non-mouse primary antibodies including anti-GFP (#AB13970;1:1,000; Applied Biosystems), E-cadherin (BD610181; 1:1,000;BD Biosciences), b-catenin (CST8480; 1:100; Cell Signaling Tech-nology) in 10% goat serum and 1%BSA in PBS overnight.When amouse-derived primary antibodywas used, sectionswerewashed,blocked with streptavidin/biotin Blocking Kit (Vector Labs), andhybridized tomouseprimary antibodyusing theMouseonMouse(M.O.M.) Basic Kit (Vector Labs). Sections were washed thenincubated with secondary antibodies including: anti-ChickenAlexa Fluor 488 (A-11039), Alexa Fluor 555-conjugated Strepta-vidin (S-21381), anti-Rabbit Alexa Fluor 647 (A-21244) in 1%BSA in PBS for 2 hours, andmounted with Fluor-Gel II with DAPI(Electron Microscopy Sciences). Bright-field and epifluorescenceimages were acquired on a Leica DMI6000B microscope andprocessed using the LAS-AF software or a Zeiss Axio Imager M2microscope and processed using the ZEN2 software. Confocalimages were acquired on a Leica TCS SP5 laser scanning confocalmicroscope and processed using the LAS-AF software.

Cell line derivation and tumor formationMultiple tumors were harvested from independent mice (n ¼

4–6) that had reached the humane endpoint of 20% maximumbody weight loss and minced under sterile conditions. Tissuepieces were digested in collagenase II (2.5 mg/mL, Worthington)and DNase (0.5 mg/mL, Worthington) in HBSS (Gibco) at 37�Cfor 45 minutes. Excess FBS (Hyclone) was added to quenchenzymes and digested tissue was pelleted by centrifugation. Pelletwas resuspended andplated innormal growthmedium: advancedDMEM (Cellgro) þ 10% FBS þ penicillin/streptomycin (Lonza)þ L-glutamine (Gibco). Adherent cells were washed with PBS(Sigma) and medium replaced 24 hours later. Multiple gastriccancer cell lines from independent mice were subcultured forseveral passages prior to use. Gastric cancer cell lines were gen-erated from gastric tumors harvested from 4 different Tcon trans-genic mice and gastric cell lines generated from the stomachs ofthree different Dcon-E-cadherin heterozygous mice. Cell lineswere generated from gastric tumors between November 8,2014 and May 13, 2015 and were used for experiments betweenpassages 3 and 8. Cell lines were tested for Mycoplasma via PCRusing the Universal Mycoplasma Detection Kit (ATCC) after eachcell line was established in culture. To generate flank tumors,gastric cancer cell lines were harvested with 0.25% Trypsin (Invi-trogen), quenched with excess medium, washed with PBS, andinjected subcutaneously in the flank at 5� 106 cells per 0.1 cc PBSper mouse. Tumors were measured every 2–3 days using digitalcalipers (Thermo Fisher Scientific) and tumor volume calculatedas (0.5) � (width2) � (length). For our experimental modelof metastasis, one gastric cancer cell line was inoculated subcu-taneously into the flanks of syngeneic wild-type mice and flanktumors were resected under sterile conditions when tumorvolumes were approximately 500 mm3. Mice were euthanizedand lung tissue analyzed formetastasis threeweeks post-resection.

RNA isolationGlandular stomachs were dissected from indicated mice,

homogenized using a gentleMACS Dissociator with M tubes(Miltenyi Biotec) in QIAzol Lysis Reagent (Qiagen), and RNA

Till et al.

Cancer Res; 77(19) October 1, 2017 Cancer Research5350




isolation performed using the RNeasy Microarray Tissue Mini Kitwith optional DNase treatment (Qiagen) according to the man-ufacturers' protocols. Quantification of total RNA was performedon a NanoDrop (Thermo Fisher Scientific).

Quantitative real-time PCRComplementary DNA (cDNA) was synthesized using the High

Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scien-tific) and qPCRwas performed on a ViiA 7 Real-Time PCR System(Thermo Fisher Scientific) using SYBR Green PCR Master Mix(Thermo Fisher Scientific) according to the manufacturers'instructions. Amplifications were performed as technical quad-ruplicates and biological sextuplicates. Values were normalized toHPRT expression.Mouse gene-specific primers were selected fromPrimer Bank (22–24) and include the following: Hprt (50-TCA-GTCAACGGGGGACATAAA, 50-GGGGCTGTACTGCTTAACCAG,142 bp, ID#7305155a1); Myc (50-ATGCCCCTCAACGTGAACT-TC, 50-CGCAACATAGGATGGAGAGCA, 228 bp, ID#27545183-a1); Ccnd1 (50-GCGTACCCTGACACCAATCTC, 50-CTCCTCTT-CGCACTTCTGCTC, 183 bp, ID#6680868a1); MMP7 (50-CTGC-CACTGTCCCAGGAAG, 50-GGGAGAGTTTTCCAGTCATGG, 175bp, ID#6754716a1); Axin2 (50-TGACTCTCCTTCCAGATCCCA,50-TGCCCACACTAGGCTGACA, 105 bp, ID#31982733a1); Lef1(50-TGTTTATCCCATCACGGGTGG, 50-CATGGAAGTGTCGCCT-GACAG, 67 bp, ID#27735019a1); CD44v1 (50-CACCATTGC-CTCAACTGTGC, 50-TTGTGGGCTCCTGAGTCTGA, 116 bp,ID#6491804a1).

Microarray target preparation and hybridizationThe University of Pennsylvania Molecular Profiling Facility

provided microarray services, including quality control tests ofthe total RNA samples by Agilent Bioanalyzer and Nanodropspectrophotometry. All protocols were conducted as described inthe Affymetrix WT Plus Reagent Kit Manual and the AffymetrixGeneChip Expression Analysis Technical Manual. Briefly, 250 ngof total RNA was converted to first-strand cDNA using reversetranscriptase primed by poly(T) and random oligomers thatincorporated the T7 promoter sequence. Second-strand cDNAsynthesis was followed by in vitro transcription with T7 RNApolymerase for linear amplification of each transcript, and theresulting cRNA was converted to cDNA, fragmented, assessed byBioanalyzer, and biotinylated by terminal transferase end label-ing. Labeled cDNA (5.5 mg) were added to Affymetrix hybridiza-tion cocktails, heated at 99�C for 5minutes and hybridized for 16hours at 45�C to Mouse Transcriptome 1.0 ST GeneChips (Affy-metrix Inc.) using the GeneChip Hybridization oven 645. Themicroarrays were then washed at low (6� SSPE) and high (100mmol/L MES, 0.1 mol/L NaCl) stringency and stained withstreptavidin–phycoerythrin. Fluorescence was amplified by add-ing biotinylated anti-streptavidin and an additional aliquot ofstreptavidin–phycoerythrin stain. A GeneChip 3000 7G scannerwas used to collect fluorescence signal. Affymetrix CommandConsole and Expression Console were used to quantitate expres-sion levels for targeted genes; default values provided by Affyme-trix were applied to all analysis parameters. Microarry data havebeen deposited in the ArrayExpress database at EMBL-EBI (www.ebi.ac.uk/arrayexpress) under accession number E-MTAB-5833.

Principal component analysis and hierarchical clusteringPrincipal component analysis was performed using Expression

Console Software (Affymetrix). The Penn Genomic Analysis Core

Bioinformatics group performed unsupervised hierarchical clus-tering of Tcon and Dcon-Ecad het stomach tissue expression datausing Partek Genomics Suite (Partek Inc).

Gene set overlap analysisThe Penn Genomic Analysis Core Bioinformatics group deter-

mined differentially expressed genes using the Significance Analy-sis of Microarrays algorithm (25) with a false discovery rateq-value cutoff of 5% and a fold-change cutoff of 2. Differentiallyexpressed mouse genes were manually mapped to their humanhomologs using the HomoloGene database (http://www.ncbi.nlm.nih.gov/homologene) from the National Center for Biotech-nology Information (26). Gene set overlaps were computed forgene lists using the Investigate Gene Sets function (http://software.broadinstitute.org/gsea/msigdb/annotate.jsp) on the GeneSet Enrichment Analysis page of the Broad Institute (27).

Western blot analysisGlandular stomachs were dissected from indicated mice and

homogenized using a gentleMACS Dissociator with M tubes(Miltenyi Biotec) in radioimmunoprecipitation assay buffer(RIPA, 50 mmol/L Tris HCl pH 8, 150 mmol/L NaCl, 1% Tri-ton-X, 0.5%sodiumdeoxycholate, and0.1%SDS)withCompleteProtease Inhibitor Cocktail Tablets (Roche 11697498001) andPhosSTOP (Roche 04906845001). Homogenates were incubatedat 4�C for 30 minutes, centrifuged at 16,000 � g at 4�C for30 minutes, supernatants were collected, and quantified usingPierce BCA Protein Assay Kit (Thermo Fisher Scientific 23225).SDS-PAGEwas performedusing 20mgprotein per sample at 150Von12%gelswith transfer to nitrocellulosemembranes performedat 110 V for 2 hours using the Bio-RadMini Protean Tetra Cell andMini Trans-Blot systems and reagents according to the manufac-turers' instructions. Blots were blocked in 5% non-fat dry milk(LabScientific M-0841) or BSA (Sigma A7906) in TBS (Bio-Rad161-0771) and incubatedwith primary antibody at 4�Covernightdiluted in blocking buffer including: MEK1/2 (CST4694, 1:1,000;Cell Signaling Technology) and p-MEK1/2 (S217/221; CST9154,1:1,000; Cell Signaling Technology).

After washing with TBSþ0.1%Tween 20 (TBS-T, Bio-Rad 170-6531), blots were incubated with secondary antibody includ-ing: anti-rabbit IgG, HRP-linked (CST7074; Cell SignalingTechnology) and anti-mouse IgG, HRP-linked (CST7076; CellSignaling Technology) for 2 hours at room temperature inblocking buffer and then washed with TBS-T. Bands werevisualized using self-prepared enhanced chemiluminescencereagent (100 mmol/L Tris pH 8.6, 0.2 mmol/L p-coumaricacid, 1.25 mmol/L luminol, 2.6 mmol/L; ref. 28).

RAS activity assayActivated Ras affinity precipitation assays were performed

according to the manufacturer's protocol. Briefly, 250 mg of celllysates were incubated with 10 mL of Raf-1 RBD agarose beads(#14-278, Upstate, Millipore) overnight at 4�C. After extensivewashing of the agarose beads three times with washing buffer(RIPA buffer, 1 mmol/L Na3VO4, 10 mg/mL leupeptin, 10 mg/mLaprotinin, and 25 mmol/L NaF). SDS-PAGE of 12% was used toseparate immunoprecipitation reactions and Western blottingwas performed with a Ras antibody. Antibodies used include:E-cadherin (BD610181#14472), Ras (Cell Signaling Technology,#3965), pERK1/2 (Cell Signaling Technology, #9101), ERK2(Santa Cruz Biotechnology, sc-154) and b-actin (Sigma, A5441).

E-Cadherin Attenuates KrasG12D-Driven Gastric Tumorigenesis

www.aacrjournals.org Cancer Res; 77(19) October 1, 2017 5351



www.ebi.ac.uk/arrayexpress

www.ebi.ac.uk/arrayexpress

http://www.ncbi.nlm.nih.gov/homologene

http://www.ncbi.nlm.nih.gov/homologene

http://software.broadinstitute.org/gsea/msigdb/annotate.jsp

http://software.broadinstitute.org/gsea/msigdb/annotate.jsp


Statistical analysisStatistical analyses (other than those previously described for

microarray analysis) were performed in Prism 5 (GraphPadSoftware). Survival analysis P values were calculated using thelog-rank (Mantel–Cox) test. All other P values were calculatedusing the Student t test (unpaired, two-tailed).

ResultsRapid development of mixed-type gastric adenocarcinoma

To investigate the contribution of oncogenic Kras combinedwith Cdh1 and Trp53 loss toward gastric tumorigenesis, wematedAtp4b-Cre transgenicmicewithCdh1fl/fl, Trp53fl/fl and LSL-KrasG12D

mice. Mice were also mated onto a Rosa26LSL-YFP background to

Figure 1.

Characterization of gastric adenocarcinoma in triple conditional mice (Atp4b-Cre;Rosa26LSL-YFP;Cdh1fl/fl,Trp53fl/fl;KrasLSLG12D/þ or Tcon). A, Survival curvedemonstrating 76.5-daymedian survival (n¼ 20) of Tconmice (red) comparedwith 322-daymedian survival (n¼ 19) of double conditionalmice (Atp4b-Cre;Cdh1fl/fl,Trp53fl/fl or KrasWT mice; blue). B, Graph representing the weights of individual Tcon and KrasWT mice in survival analysis over time. C, Representative grossimage of awild-type stomach and a Tcon stomach (67 and 77 days of age, respectively, scale bar: 1 cm).D,Representative images of intestinal-type and diffuse-typelesions found in Tcon mice. E, Representative H&E and immunofluorescence (blue, DAPI; yellow, YFP) images of Tcon and Cre- control stomachs harvestedfrom mice at 3, 6, and 9 weeks of age (scale bar, 1 mm).

Till et al.





confirm YFPþ metastases with these mice referred to as tripleconditional (Tcon) mice. Characterization of Tcon gastric cancermice revealed a rapid course of disease, resulting in a mediansurvival of 76.5days (range64–91days, SD7.2) as comparedwitha 322-day median survival of control mice (referred to as KrasWT)withwild-typeKras, Cdh1 loss and Trp53 loss in the gastric parietalcell lineage (range 272–386 days, SD 32.8, Fig. 1A). Diseaseprogression correlates with weight loss as Tcon mice begin tolose weight around 60 days of age while KrasWT control mice withwild-type Kras, Cdh1 loss and Trp53 loss in the gastric parietal celllineage only begin to loseweight around 300days of age (Fig. 1B).At necropsy, Tcon stomachs were dramatically enlarged withnormal stomach replaced by tumor displaying linitis plastica (a.k.a. leather bottle stomach) with thickened, rigid, and whitenedgastric walls (Fig. 1C; ref. 14). Histologically, the gastric adeno-carcinomas that developed in these mice were mixed type tumorscontaining both diffuse and intestinal histology (Fig. 1D). Weexamined the development of gastric adenocarcinoma over timeby analyzing stomachs of Tconmice andTconCre- controls at 3, 6,and 9 weeks of age. As expected, Cre- control stomachs displayednormal architecture at all timepoints (Fig. 1E). Tconmice had few

or no identifiable gastric parietal cells by 3 weeks of age anddemonstrated both high-grade dysplastic lesions and/or intramu-cosal carcinomas in 100%ofmice. By six weeks of age, there was adramatic increase in these lesions with invasive carcinomas in40% mice while by 9 weeks of age 100% of mice had invasivecarcinomas.

Widespread metastasis in mice with gastric adenocarcinomaTen mice were sacrificed at 64–83 days of age and analyzed

for gastric adenocarcinoma progression and metastasis (Fig. 2).Upon gross examination of the abdomen, the stomachs wereenlarged and YFPþ nodules were apparent within the perigastricfat in all mice examined, indicating their identity as perigastriclymph node metastases. Gross examination of the thoraciccavity identified lesions in the mediastinal lymph nodes adja-cent to the trachea in 50% of mice examined that were all YFPþ

confirming that they were mediastinal lymph node metastases(Fig. 2A–C). YFPþ lung metastases were identified in all micewhile analysis of YFP expression in livers from Tcon miceidentified micrometastatic lesions in 20% of mice (Fig. 2C).We isolated six independent cell lines from six different Tcon

Figure 2.

Tcon (Atp4b-Cre;Rosa26LSL-YFP;Cdh1fl/fl,Trp53fl/fl;KrasLSLG12D/þ) mice have extensive metastases.A, Representative gross and fluorescence images of theabdomen and thoracic cavity of a Tcon mouse.B and C, Hematoxylin and eosin and immunofluorescence(yellow, YFP; blue, DAPI) images of metastatic lesions foundin Tcon mice. C, Left, growth curve of subcutaneous flanktumors derived from a Tcon gastric tumor cell line (n ¼ 5).D, Left, data are presented as the mean tumor volume� SD.Right, hematoxylin and eosin image of lung metastases (redarrows) after a Tcon gastric cancer cell line was injectedsubcutaneously in the flank, followed by resection(n ¼ 3; right). Scale bars are all 1 mm except for the liverthat is 0.5 mm.






mice and one metastatic lesion that were expanded in culture.All cell lines tested formed flank tumors in syngeneic wild-typeC57BL/6 mice. Two of the gastric cancer–derived cell linesmetastasized to the lung in an experimental model of sponta-neous metastases (Fig. 2D).

Increased survival of mice with one copy of wild-typeE-cadherin

AsKrasG12D and p53 loss are sufficient to drive tumorigenesis inmost organs, we investigated whether E-cadherin status affectedthe growthandprogressionof gastric adenocarcinoma inour Tconmice. We generated Tcon mice with one wild-type allele of Cdh1and refer to them as Dcon-Ecad heterozygous mice (Atp-4b-Cre;Cdh1fl/þ;LSL-KrasG12D;Trp53fl/fl;Rosa26LSL-YFP). Thepresenceofonewild-type Cdh1 allele dramatically increased survival (Fig. 3A, P <0.0001) in comparison with Tcon mice with complete loss of E-cadherin.Of37Dcon-Ecadheterozygousmice,only7micediedofgastric tumors by 123 days. In comparison, a cohort of 20 Tconmice all died by 91 days of age. The majority of Dcon-Ecadheterozygous mice did not undergo the significant weight lossthat was observed in gastric cancer–bearing Tcon mice (Fig. 3B).

Loss of wild-type E-cadherin expression leads to gastricadenocarcinoma

Although stomachs isolated from Dcon-Ecad heterozygousmice between 67 and 123 days old are enlarged and hyperplastic,the majority do not harbor invasive tumors. Upon histologicanalysis, only 2 of 11 Dcon-Ecad heterozygous mice had detect-

able stomach tumors. Histologically, these tumors were mixed-type gastric tumors; however, these tumors were more focal innature, unlike the widespread gastric tumors identified in Tconmice (Fig. 3C). These data implicate the requirement of a sec-ondary event for tumorigenesis to occur in the gastric parietal celllineage cells of origin.

One explanation for the occasional stomach tumors observedinDcon-Ecad heterozygousmicewith one copy ofwild-typeCdh1could be due to the eventual loss of E-cadherin expression. Thus,we examined E-cadherin expression by immunofluorescent stain-ing in stomachs from Dcon-Ecad heterozygous mice with tumorsor with hyperplastic lesions at 108 days of age as compared withstomach tumors harvested from Tcon mice at 68 days of age (Fig.3D.) As expected, E-cadherin protein was absent in recombinedcells (YFPþ) in Tcon stomach tumors but present in adjacentnormal (YFP-negative) tissue. E-cadherin expression was detectedin sections from Dcon-Ecad heterozygous stomachs that lackedinvasive tumors (referred to as benign) and in normal tissue oftumor bearing-Dcon-Ecad heterozygous mice. However, in stom-ach tumors from Dcon-Ecad heterozygous mice, E-cadherinexpression was absent or decreased in YFPþ-tumor cells (Fig.3D) implicating loss of E-cadherin as a critical barrier for gastrictumorigenesis evenwithoncogenicKRASexpression andp53 loss.

Increased Ras activity and upregulation of Kras signalingin the absence of E-cadherin

To identify differences in gene expression between stomachsisolated from Tcon mice and Dcon-Ecad heterozygous mice, a

Figure 3.

Gastric cancer phenotype in double conditionalmicewith one copy of E-cadherin (Atp4b-Cre;Rosa26LSL-YFP;Cdh1fl/þ,Trp53fl/fl;KrasLSLG12D/þor Dcon-Ecad hetmice).A, Survival curve demonstrating 123-day median survival (n ¼ 27) of Dcon-Ecad het mice (green) compared with 76-day median survival (n ¼ 20) of Tcon (red)mice.B,Graph representing theweights of individual mice in survival analysis over time. C,Representative gross image of a Tcon stomach (80 days of age; scale bar,1 cm) as compared with stomachs from Dcon-E-cad het mice without a tumor (77 days of age, benign) or a stomach with a tumor (108 days of age, tumor).Arrows, tumors. D, Representative immunofluorescence images of YFP (green) and E-cadherin (white) expression with DAPI (blue) in stomachs isolated frommicewith the indicated genotypes.

Till et al.





cDNA microarray on glandular stomach tissue from 4 individualmice of each genotype was performed. Principal componentsanalysis revealed two clusters: one that contained all 4 of theTcon stomachs and another that contained 3 of the 4 Dcon-Ecadheterozygous mice. However, 1 of the 4 Dcon-Ecad heterozygousstomachs did not fall distinctly within either of these groups butwas closer to the Tcon stomachs (Fig. 4A). Unsupervised hierar-

chical clustering analysis indicates that the single outlier Dcon-Ecad heterozygous stomach sample clustered more closely to theTcon stomach cluster rather than to the other 3 Dcon-Ecadheterozygous stomachs (Fig. 4B). The frequency of invasive gastricadenocarcinoma observed in Dcon-Ecad heterozygous miceoccurred at a ratio of 1 in 5 mice. As it was likely that this singlemouse harbored a gastric adenocarcinoma, microarray data from

Figure 4.

Increased KRAS signaling in gastric tumors from Tcon mice. Principal component analysis (PCA; A) and unsupervised hierarchical clustering analysis of microarraydata (B) from Tcon (Cdh1fl/fl) and Dcon-Ecad heterozygous (het) stomachs (Cdh1fl/þ). C,Western blot analysis of RAS-GTP pulldown, total RAS, phopho-ERK, totalERK, andb-actin in gastric cancer cell lines isolated fromTcon (Cdh1fl/fl) gastric tumors and fromDcon-Ecad het (Cdh1fl/þ) tumors. Each lane is an independent gastriccancer cell line. Densitometric quantification of RAS-GTP relative to total RAS is shown on the right. D, Western blot analysis of phospho-MEK and total MEKexpression in stomach lysates from Tcon and Dcon-Ecad het (Cdh1fl/þ) mice. Densitometric quantification of phospho-MEK levels relative to total MEK is shownon the right. E, Survival curve of Tconmice treated with the MEK inhibitor PD0325901 or vehicle beginning at 4 weeks of age (indicated by arrow) demonstrating anincrease in median survival of 96 days in the presence of the MEK inhibitor over vehicle (P ¼ 0.001, n ¼ 8).






this outlier was removed from analysis. To determine differen-tially regulated pathways between the two genotypes, we per-formed gene set overlap analysis (27) on genes upregulated in theTcon stomachs versus Dcon-Ecad heterozygous stomachs and apublicly available dataset of transcription factor target gene sets.This analysis revealed that Receptor TyrosineKinase/KRAS–drivengene setswere upregulated in stomachs harvested fromTconmice.When comparing our gene set to the collection of cancer hallmarkgene sets, the second most significant hit was for "Genes upre-gulated by KRAS activation." Furthermore, significant overlapwasfound between our gene set and 9 EGFR/ERBB2/Kras regulatedoncogenic signature gene sets (Table 1).

To confirm increased RAS signaling in gastric tumors fromTcon mice, we measured RAS activity in gastric tumor cell linesgenerated from gastric tumors from Tcon and Dcon-Ecad het-erozygous mice. Utilizing a pull-down assay with the RAS-binding domain of RAF, which specifically binds the activeGTP-bound form of RAS, we found higher RAS activity in gastriccancer cell lines derived from Tcon mice as compared withgastric cancer cell lines fromDcon-Ecad heterozygous mice (Fig.4C). Furthermore, Tcon-derived gastric cancer cell lines alsodemonstrated increased expression of phospho-ERK as com-pared with gastric cancer cell lines from Dcon-Ecad heterozy-gous cell lines (Fig. 4C). Next, we probed for MEK activation instomach tissue harvested from Tcon and Dcon-Ecad heterozy-gous mice. Western blot analysis of phospho-MEK expression instomach tumors demonstrated upregulation of phospho-MEKrelative to total MEK in tumors from Tcon mice as comparedwith those from Dcon-Ecad heterozygous mice (Fig. 4D). Todetermine the contribution of increased MAPK signaling in theabsence of E-cadherin towards gastric tumorigenesis, we treatedTconmice starting at 4 weeks of age with the clinically approvedMEK inhibitor, PD324901. Tcon mice had an increase in medi-an survival of 18.5 days over treatment with vehicle alone (P ¼0.001, n ¼ 8; Fig. 4E). Collectively, these data implicate thatoncogenic KRAS promotes gastric tumorigenesis and metastasisin our Tcon model. Given that both Tcon and Dcon-Ecadheterozygous stomachs express oncogenic KRAS, these datasuggest that E-cadherin expression in gastric parietal cells mayattenuate the activity of oncogenic KRAS in our model.

Upregulation of b-catenin targets in the absence of E-cadherinWhile E-cadherin has been identified as the genetic hit in

Hereditary Diffuse Gastric Cancer, the exact mechanism by whichloss of E-cadherin expression leads to gastric adenocarcinoma isstill not known. One mechanism that has been proposed to

underlie E-cadherin's tumor suppressive activity is sequestrationof b-catenin at the adherens junction in the cadherin–catenincomplex at the cell membrane. While loss of E-cadherin allowsb-catenin to localize to the cytosol, this does not necessarilyindicate b-catenin nuclear translocation and downstream trans-activation of classical WNT pathway target genes. To test whetherthe lack of E-cadherin altered the subcellular localization and/oractivation of b-catenin, we coimmunostained stomachs fromTcon andDcon-Ecad heterozygousmicewith andwithout tumorsfor E-cadherin, b-catenin, and YFP expression (Fig. 5A). Tconstomachs had no E-cadherin expression in YFPþ tumor cells asexpected and also lacked b-catenin localization at the membranein YFPþ tumor cells but showed diffuse cytsolic b-catenin expres-sion (Fig. 5A). In contrast, nontumor YFP� cells in Tcon stomachsserve as an internal control demonstrating both E-cadherin andb-catenin expression at the cell membrane (Fig. 5A). Dcon-Ecadheterozygous stomachs without obvious tumors but expressingYFPþ cells throughout (referred to as benign) retained membra-nous expression of E-cadherin and b-catenin as indicated byimmunostaining (Fig. 5A). However, Dcon-Ecad heterozygousstomachs with YFPþ tumor cells and confirmed gastric tumors,displayedmuch less E-cadherin expression throughout the sectionas well as a loss of membrane localized b-catenin in YFPþ tumorscells lacking E-cadherin expression as indicated by immunostain-ing (Fig. 5A).

As b-catenin can be degraded in the cytoplasm, we performedgene set overlap analysis to further investigate whether loss ofmembrane localized b-catenin expression in stomach tumorsfrom Tcon mice led to increased b-catenin activation. We com-pared genes upregulated in Tcon stomachs with tumors versusDcon-Ecad heterozygous stomachs likely without tumors, to apublicly available data set of transcription factor target gene sets(Table 2). The first hit was a set of genes containing binding sitesfor the b-catenin binding partner LEF1, in their promoters. Fur-thermore, when genes upregulated in Tcon versus Dcon-Ecadheterozygous stomachs were compared with oncogenic signaturegene sets, the results contained several genes sets whose upregula-tion are due to the expression of LEF1, Wnt signaling, andb-catenin. Finally, to confirm increased b-catenin activation instomach tumors from Tcon mice, we investigated the expressionof several canonical WNT/b-catenin target genes including LEF1,CD44, CyclinD1, and c-Myc, by qRT-PCR (Fig. 5B). All genesweresignificantly upregulated in Tcon compared with Dcon-Ecadheterozygous stomach tissue. Collectively, these data are consis-tent with an upregulation of b-catenin target genes in the absenceof E-cadherin.

Table 1. Gene set overlap analysis comparing cancer hallmark gene sets and oncogenic signature gene sets with genes upregulated in Tcon versus Dcon-Ecad hetstomachs

Name Description FDR q-value

Cancer hallmark gene setsHALLMARK_KRAS_SIGNALING_UP Genes upregulated by KRAS activation 3.36e�61

Oncogenic signatureKRAS.600_UP.V1_UP Upregulated in four lineages of epithelial cell lines overexpressing oncogenic KRAS 1.47e�19

KRAS.DF.V1_UP Upregulated in epithelial lung cancer cell lines overexpressing oncogenic KRAS 9.39e�18

KRAS.600.LUNG.BREAST_UP.V1_UP Upregulated in epithelial lung and breast cancer cell lines overexpressing oncogenic KRAS 2.68e�17

KRAS.LUNG.BREAST_UP.V1_UP Upregulated in epithelial lung and breast cancer cell lines overexpressing oncogenic KRAS 2.55e�15

KRAS.BREAST_UP.V1_UP Upregulated in epithelial breast cancer cell lines overexpressing oncogenic KRAS 4.49e�11

KRAS.LUNG_UP.V1_UP Upregulated in epithelial lung cancer cell lines overexpressing oncogenic KRAS 1.34e�10

KRAS.300_UP.V1_UP Upregulated in four lineages of epithelial cell lines overexpressing oncogenic KRAS 8.06e�09

KRAS.PROSTATE_UP.V1_UP Upregulated in epithelial prostate cancer cell lines overexpressing oncogenic KRAS 2.75e�08

KRAS.50_UP.V1_UP Upregulated in four lineages of epithelial cell lines overexpressing oncogenic KRAS 2.78e�07

Abbreviation: FDR, false discovery rate.

Till et al.





DiscussionGiven the high rate of mutation and amplification of the

receptor tyrosine kinase (RTK)/Ras pathway in gastric cancer, wehave generated a mouse model of gastric adenocarcinoma thatincludes expression of oncogenic Kras. In our Tcon mouse modelof gastric adenocarcinoma, we combined oncogenic Kras expres-sion with E-cadherin and p53 loss in the gastric parietal celllineage to reveal a clinically relevant model of gastric adenocar-cinoma that progresses rapidly and metastasizes widely to lymphnodes, lung, and liver. This work has demonstrated the impor-tance of E-cadherin as a gatekeeper for primary gastric tumori-genesis as mice generated with oncogenic Kras, loss of Trp53, andone wild-type allele of Cdh1, referred to as Dcon-Ecad heterozy-gousmice,was sufficient to significantly inhibit tumor growth andprogression and prolong overall survival. As of 123 days, only20% of Dcon-Ecad heterozygous mice demonstrated gastrictumors with no metastases identified in distant organs. RAS

activity was measured in gastric cancer cell lines generated fromTcon tumors and from Dcon-Ecad tumors by a RAF-RBD pull-down of RAS-GTP. Our data indicate a significant increase inRAS activity in gastric cancer cell lines lacking E-cadherin expres-sion (from Tcon mice) as compared with gastric cancer cell linesexpressing E-cadherin (from Dcon-Ecad heterozygous mice).Furthermore, we show that phospho-ERK expression was sig-nificantly higher in the absence of E-cadherin supporting thenotion E-cadherin may act as a gatekeeper to oncogenic Krasand Trp53 loss-driven gastric cancer. Gene expression analysisidentified increased oncogenic KRAS signaling only in theabsence of E-cadherin with increased phospho-MEK expressiondetected by Western blot only in gastric tumors isolated fromTcon mice and not from mice with expression of E-cadherin(Dcon-Ecad heterozygous mice).

Expression of oncogenic Kras and loss of Trp53 is sufficient todrive rapid development of lung cancer (11), pancreatic ductal

Figure 5.Increased b-catenin activation in gastric tumors from Tcon mice. A, Representativeimmunofluorescence images of YFP (green), E-cadherin (white), and b-catenin (red)expression with DAPI (blue) in stomachs isolated from mice with the indicatedgenotypes. Dotted yellow boxes indicate enlarged insets and arrows indicatenontumor YFP- areas. B, Expression of canonical b-catenin target genes inTcon versus Dcon-Ecad heterozygous (het) stomachs. Data are normalized toHPRT expression and are displayed as mean fold change over Dcon-Ecad hetstomachs þ SD (n ¼ 6).






adenocarcinoma (12), squamous cell carcinoma (13), acutemyeloid leukemia (14), and undifferentiated pleomorphicsarcoma (15, 16) in the mouse. However, in the gastric parietalcell lineage, despite the expression of oncogenic Kras and Trp53loss, E-cadherin expression dramatically decreases tumorigen-esis. These data implicate Cdh1 loss as necessary for oncogenicKRAS and P53 loss to promote tumorigenesis and metastases inthese cells. Our microarray data show that gene sets upregu-lated by oncogenic Kras activity are overrepresented in genesupregulated in stomachs from Tcon versus Dcon-Ecad hetero-zygous mice. The increased phospho-MEK expression in stom-ach tumors from Tcon mice further confirms increased RASactivity. The levels of Kras mRNA were similar in the stomachsof both mouse models (data not shown) thus, one wouldexpect similar levels of RAS activity in these two models. Ourdata suggest that the presence of E-cadherin attenuates onco-genic KRAS signaling and its effector pathways in the Dcon-Ecad heterozygous mouse model, thus indicating that E-cad-herin functions downstream of oncogenic KRAS.

The observed loss of E-cadherin expression in the 20%of gastrictumors that arise inDcon-Ecad heterozygousmice suggests that E-cadherin is a gatekeeper to tumorigenesis in the gastric parietal celllineage. Our data demonstrate that the presence of a wild-typeCdh1 allele in our model delays oncogenesis significantly and ofthe tumors that do form, the tumors show dramatic loss of E-cadherin expression. This concept of a gatekeeper in gastrointes-tinal tumors is not without precedent as loss of the Apc gene hasalso been shown to be necessary for colon carcinoma develop-ment even in the presence of oncogenic Kras and Trp53 loss inLgr5þ colon cells. In addition, colon tumorswere shown to regressupon reexpression of Apc (29). Furthermore, it is intriguing thatAPC and E-cadherin functionality overlap in their regulationb-catenin (30, 31). Collectively, our data support the hypothesisthat canonical WNT signaling is necessary for oncogenic Kras andTrp53 loss-driven tumorigenesis in the gut. Similarly, Hayakawaand colleagues, (32) have shown in Mist1þ gastric stem cells thatoncogenic-Kras alone drives intestinal metaplasia but is insuffi-cient to drive gastric adenocarcinoma development. When pairedwithAPC loss, intestinal-type gastric adenocarcinomadevelops by120 days (32). Both E-cadherin and APC share regulatory controlof the canonicalWNT signalingmolecule b-catenin. Our data andpublished data on APC suggest that loss of either E-cadherin orAPC lead to increased b-catenin signaling, which may be crucialfor mutant Kras and Trp53 loss–driven oncogenesis in both thegastric parietal cell lineage and in Lgr5þ colon stem cells. Thus,regulation of b-catenin activation may be a critical node fortumorigenesis throughout the gut.

Whereas Hayakawa and colleagues proposed Mist1þ stem cellsas the cell of origin in both intestinal- and diffuse-type gastriccancer (32), our model and that of Shimada and colleagues (10)

suggest that the gastric parietal cell lineage, including both pre-parietal cells and mature parietal cells, may also act as a cell oforigin for gastric adenocarcinoma. Specifically, our modeldemonstrates that both intestinal- and diffuse-type lesion canarise from this, Atp4bþ lineage. It is possible that together thesedata suggest a Mist1þ;Atp4bþ pre-parietal cell as the cell of originin gastric adenocarcinoma. Thepresence ofMist1þ;Atp4bþ cells inthe stomach remains to be investigated.

Both our Tcon and Dcon-Ecad heterozygous mouse modelsof gastric adenocarcinoma offer the potential to investigatethe molecular mechanisms driving gastric cancer via both E-cadherin and the RTK/Ras signaling axis. The Tcon mice mayalso serve as a useful model for testing new therapies targetingboth primary gastric cancer and metastatic lesions. Both therapid progression of gastric cancer and widespread metastasesset the Tcon model apart from other available mouse modelsof gastric cancer. Furthermore, our model represents the firstmouse model of mixed-type gastric adenocarcinoma andoffers the only in vivo tool to investigate this gastric adeno-carcinoma subtype.

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

Authors' ContributionsConception and design: S.S. Yoon, S. RyeomDevelopment of methodology: B.-J. Kim, S.S. Yoon, S. RyeomAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.E. Till, P. Addai, E. Jonokuchi, S.S. Yoon, S. RyeomAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.E. Till, K.C. Roby, L.H. Tang, S.S. Yoon, S. RyeomWriting, review, and/or revision of the manuscript: J.E. Till, L.H. Tang,S.S. Yoon, S. RyeomAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J.E. Till, C. Yoon, S. RyeomStudy supervision: S.S. Yoon, S. Ryeom

AcknowledgmentsThe authors would like to thank Dr. Anil Rustgi and members of the Ryeom

lab for helpful comments and suggestions.

Grant SupportThis work was supported by NIDDK Center Grant P30-DK050306

(S. Ryeom), The Garrett B. Smith Foundation (S. Ryeom), NCI R01 CA 118374(S. Ryeom),MSKCCFunctionalGenomics InitiativeRapid Response PilotGrant(S.S. Yoon), NCI Cancer Center Support Grant P30 CA008748 (S.S. Yoon)

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 January 27, 2017; revised June 12, 2017; accepted July 21, 2017;published OnlineFirst July 31, 2017.

Table 2. Gene set overlap analysis comparing transcription factor target gene sets and oncogenic signature gene sets with genes upregulated in Tcon versusDcon-Ecad het stomachs

Name Description FDR q-value

Transcription factor target gene setsCTTTGT_V$LEF1_Q2 Genes with promoter regions containing the LEF1 motif 3.27 e�87

CAGGTG_V$E12_Q6 Genes with promoter regions containing the TCF3 motif 1.44 e�58

Oncogenic signature gene setsLEF1_UP.V1_UP Upregulated in DLD1 cells overexpressing LEF1 2.86 e�19

WNT_UP.V1_UP Upregulated in C57MG cells overexpressing WNT1 1.42 e�8

BCAT.100_UP.V1_UP Upregulated in HEK293 cells expressing active CTNNB1 3.83 e�8

Abbreviation: FDR, false discovery rate.


Till et al.




References1. Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, et al.

GLOBOCAN 2012 v1.0. Cancer Incidence and Mortality Worldwide IARCCancerBase No. 11 [Internet]. Lyon, France: International Agency forResearch on Cancer; 2013. Available from: http://globocan.iarc.fr.

2. CrewKD,Neugut AI. Epidemiology of gastric cancer.World J Gastroenterol2006;12:354.

3. Wagner AD, Grothe W, Haerting J, Kleber G, Grothey A, Fleig WE. Che-motherapy in advanced gastric cancer: a systematic review and meta-analysis based on aggregate data. J Clin Oncol 2006;24:2903–9.

4. Lauren P. The two histological main types of gastric carcinoma: diffuse andso-called intestina-type carcinoma. An attempt at a histo-clinical classifi-cation. Acta Pathol Microbiol Scand 1965;64:31–49.

5. Lauwers GY. Epithelial neoplasms of the stomach. In: Odze RD, GoldblumJR, editors. In: Odze and Goldblum surgical pathology of the GI tract, liver,biliary tract and pancreas. Third edition. Philadelphia, PA: Saunders/Else-vier; 2015. Available from: https://www.clinicalkey.com/#!/content/book/3-s2.0-B978145570747800025X?scrollTo¼%2523s0045.

6. Cerami E,Gao J,DogrusozU,Gross BE, Sumer SO, Aksoy BA, et al. The cBiocancer genomics portal: an open platform for exploring multidimensionalcancer genomics data. Cancer Discov 2012;2:401–4.

7. Bass AJ, Thorsson V, Shmulevich I, Reynolds SM,Miller M, Bernard B, et al.Comprehensive molecular characterization of gastric adenocarcinoma.Nature 2014;513:202–9.

8. Graziano F, Humar B, Guilford P. The role of the E-cadherin gene (CDH1)in diffuse gastric cancer susceptibility: from the laboratory to clinicalpractice. Ann Oncol 2003;14:1705–13.

9. Hayakawa Y, Fox J, Gonda T, Worthley D, Muthupalani S, Wang T. Mousemodels of gastric cancer. Cancers 2013;5:92–130.

10. Shimada S, Mimata A, Sekine M, Mogushi K, Akiyama Y, Fukamachi H,et al. Synergistic tumour suppressor activity of E-cadherin and p53 in aconditional mouse model for metastatic diffuse-type gastric cancer. Gut2012;61:344–53.

11. Jackson EL, Olive KP, Tuveson DA, Bronson R, Crowley D, Brown M, et al.The differential effects of mutant p53 alleles on advanced murine lungcancer. Cancer Res 2005;65:10280–8.

12. BardeesyN, Aguirre AJ, ChuGC, Cheng K -h, Lopez LV,Hezel AF, et al. Bothp16Ink4a and the p19Arf-p53 pathway constrain progression of pancreaticadenocarcinoma in the mouse. Proc Natl Acad Sci 2006;103:5947–52.

13. Caulin C, Nguyen T, LangGA,Goepfert TM, Brinkley BR, CaiW-W, et al. Aninducible mouse model for skin cancer reveals distinct roles for gain- andloss-of-function p53 mutations. J Clin Invest 2007;117:1893–901.

14. Zhao Z, Zuber J, Diaz-Flores E, Lintault L, Kogan SC, Shannon K, et al. p53loss promotes acute myeloid leukemia by enabling aberrant self-renewal.Genes Dev 2010;24:1389–402.

15. Kirsch DG, Dinulescu DM, Miller JB, Grimm J, Santiago PM, Young NP,et al. A spatially and temporally restricted mouse model of soft tissuesarcoma. Nat Med 2007;13:992–7.

16. Mito JK, Riedel RF, Dodd L, Lahat G, Lazar AJ, Dodd RD, et al. Cross speciesgenomic analysis identifies a mouse model as undifferentiated pleomor-phic sarcoma/malignant fibrous histiocytoma. PLoS One 2009;4:e8075.

17. Boussadia O, Kutsch S, Hierholzer A, Delmas V, Kemler R. E-cadherin is asurvival factor for the lactating mouse mammary gland. Mech Dev2002;115:53–62.

18. Jackson EL, Willis N, Mercer K, Bronson RT, Crowley D, Montoya R, et al.Analysis of lung tumor initiation and progression using conditionalexpression of oncogenic K-ras. Genes Dev 2001;15:3243–8.

19. Marino S, Vooijs M, van der Gulden H, Jonkers J, Berns A. Induction ofmedulloblastomas in p53-null mutant mice by somatic inactivation of Rbin the external granular layer cells of the cerebellum. Genes Dev2000;14:994–1004.

20. Srinivas S,Watanabe T, Lin CS,William CM, Tanabe Y, Jessell TM, et al. Crereporter strains produced by targeted insertion of EYFP and ECFP into theROSA26 locus. BMC Dev Biol 2001;1:4.

21. Syder AJ, Karam SM, Mills JC, Ippolito JE, Ansari HR, Farook V, et al.A transgenic mouse model of metastatic carcinoma involving transdiffer-entiation of a gastric epithelial lineage progenitor to a neuroendocrinephenotype. Proc Natl Acad Sci U S A 2004;101:4471–6.

22. Wang X, Seed B. A PCR primer bank for quantitative gene expressionanalysis. Nucleic Acids Res 2003;31:e154.

23. Spandidos A, Wang X, Wang H, Dragnev S, Thurber T, Seed B, et al. Acomprehensive collection of experimentally validated primers for Poly-merase Chain Reaction quantitation ofmurine transcript abundance. BMCGenomics 2008;9:633.

24. Spandidos A, Wang X, Wang H, Seed B. PrimerBank: a resource of humanand mouse PCR primer pairs for gene expression detection and quantifi-cation. Nucleic Acids Res 2010;38:D792–9.

25. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarraysapplied to the ionizing radiation response. Proc Natl Acad Sci U S A2001;98:5116–21.

26. Geer LY, Marchler-Bauer A, Geer RC, Han L, He J, He S, et al. The NCBIBioSystems database. Nucleic Acids Res 2010;38:D492–6.

27. SubramanianA, TamayoP,Mootha VK,Mukherjee S, Ebert BL,GilletteMA,et al. Gene set enrichment analysis: a knowledge-based approach forinterpreting genome-wide expression profiles. Proc Natl Acad Sci 2005;102:15545–50.

28. Mruk DD, Cheng CY. Enhanced chemiluminescence (ECL) for routineimmunoblotting: An inexpensive alternative to commercially availablekits. Spermatogenesis 2011;1:121–2.

29. Dow LE, O'Rourke KP, Simon J, Tschaharganeh DF, van Es JH, Clevers H,et al. Apc restoration promotes cellular differentiation and reestablishescrypt homeostasis in colorectal cancer. Cell 2015;161:1539–52.

30. Mohammed MK, Shao C, Wang J, Wei Q, Wang X, Collier Z, et al. Wnt/b-catenin signaling plays an ever-expanding role in stem cell self-renewal,tumorigenesis and cancer chemoresistance. Genes Dis 2016;3:11–40.

31. Liu X, Chu KM. E-cadherin and gastric cancer: cause, consequence, andapplications. BioMed Res Int 2014;2014:637308.

32. Hayakawa Y, Ariyama H, Stancikova J, Sakitani K, Asfaha S, Renz BW, et al.Mist1 expressing gastric stem cells maintain the normal and neoplasticgastric epithelium and are supported by a perivascular stem cell niche.Cancer Cell 2015;28:800–14.





http://globocan.iarc.fr

https://www.clinicalkey.com/#!/content/book/3-s2.0-B978145570747800025X?scrollTo=%2523s0045




2017;77:5349-5359. Published OnlineFirst July 31, 2017.Cancer Res Jacob E. Till, Changhwan Yoon, Bang-Jin Kim, et al. Mice That Can Be Attenuated by E-Cadherin ExpressionOncogenic KRAS and p53 Loss Drive Gastric Tumorigenesis in

Updated version

10.1158/0008-5472.CAN-17-0061doi:

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/77/19/5349.full#ref-list-1

This article cites 30 articles, 11 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/77/19/5349.full#related-urls

This article has been cited by 3 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/77/19/5349To 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-17-0061

http://cancerres.aacrjournals.org/content/77/19/5349.full#ref-list-1

http://cancerres.aacrjournals.org/content/77/19/5349.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/77/19/5349


oncogenic kras and p53 loss drive gastric tumorigenesis ......tumor and stem cell biology oncogenic...

Documents