kras activation in gastric adenocarcinoma stimulates epithelial-to-mesenchymal ... · this...

Signal Transduction and Functional Imaging

KRAS Activation in Gastric AdenocarcinomaStimulates Epithelial-to-Mesenchymal Transitionto Cancer Stem–Like Cells and PromotesMetastasisChanghwan Yoon1, Jacob Till2, Soo-Jeong Cho3, Kevin K. Chang1, Jian-xian Lin1,4,Chang-ming Huang4, Sandra Ryeom2, and Sam S. Yoon1

Abstract

Our previous work showed that in amousemodel of gastricadenocarcinoma with loss of p53 and Cdh1 that addingoncogenic Kras (a.k.a. Tcon mice) accelerates tumorigenesisandmetastasis. Here, we sought to examineKRAS activation inepithelial-to-mesenchymal transition (EMT) and generationof cancer stem–like cells (CSC). Transduction of nontrans-formed HFE-145 gastric epithelial cells with oncogenicKRASG12V significantly decreased expression of the epithelialmarker E-cadherin, increased expression of the mesenchymalmarker vimentin and the EMT transcription factor Slug, andincreasedmigration and invasion by 15- to 17-fold. KRASG12V

also increased expression of self-renewal proteins such as Sox2and increased spheroid formation by 2.6-fold. In tumor-derivedorganoids fromTconmice,Kras knockdowndecreasedspheroid formation, expression of EMT-related proteins,migration, and invasion; similar effects, as well as reversal of

chemoresistance, were observed following KRAS knockdownor MEK inhibition in patient tumor-derived gastric adenocar-cinoma cell lines (AGS and KATOIII). KRAS inhibition ingastric adenocarcinoma spheroid cells led to reduced AGSflank xenograft growth, loss of the infiltrative tumor border,fewer lung metastases, and increased survival. In a tissuemicroarray of human gastric adenocarcinomas from 115patients, high tumor levels of CD44 (a marker of CSCs) andKRAS activation were independent predictors of worse overallsurvival. In conclusion, KRAS activation in gastric adenocar-cinoma cells stimulates EMT and transition to CSCs, thuspromoting metastasis.

Implications: This study provides rationale for examininginhibitors of KRAS to block metastasis and reverse chemo-therapy resistance in gastric adenocarcinoma patients.

IntroductionThere are nearly onemillion new gastric cancer cases and nearly

700,000 gastric cancer–related deaths worldwide per year, andthus gastric cancer accounts for almost 10% of all cancerdeaths (1). Gastric adenocarcinomas comprise the vast majorityof gastric cancers. The majority of patients with gastric adenocar-cinoma present with locally advanced or metastatic disease. Theresponse rate of gastric adenocarcinoma to multiagent chemo-therapy can be 50% or greater, but nearly all patients develop

chemotherapy resistance, andmedian survival is extended only to10 to 12 months (2). Thus, new therapies are needed.

Genes encoding the Receptor Tyrosine Kinase (RTK)-RAS sig-naling pathway and the tumor-suppressor TP53 are altered in60% and 50% of gastric adenocarcinomas, respectively (3). TheRAS family of proteins (in humans, HRAS, KRAS, and NRAS) aresmall GTPases involved in cellular signal transduction supportingcell growth and survival (4).KRAS is amplified ormutated in 17%of gastric adenocarcinomas (3). Upon stimulation by upstreamreceptors, KRAS switches froman inactive,GDP-bound form to anactive, GTP-bound form. This conformational change leads to itsbinding with RAF. KRAS recruits RAF to the membrane where ispromotes RAF dimerization and activation. Activated RAF phos-phorylates and activates MEK, and activatedMEK phosphorylatesand activates ERK.

There is some evidence that RTK-RAS signaling is important inthe epithelial-to-mesenchymal transition (EMT) and mainte-nance of gastric cancer stem–like cells (CSC). CSCs, the existenceof which is still somewhat controversial, share properties ofnormal stem cells such as the capacity for self-renewal anddifferentiation (5), and may be the source of metastases (6).Many of the phenotypic differences between CSCs and bulktumor cells that lack stemness can be attributed to epigeneticchanges caused by the EMT program (7). The CSC paradigm canexplain how epigenetic changes can result in phenotypic diversitywithin tumor cells and lead to chemotherapy resistance. As mostconventional chemotherapies do not reliably eradicate CSCs,

1Gastric and Mixed Tumor Service, Department of Surgery, Memorial SloanKettering Cancer Center, New York, New York. 2Department of Cancer Biology,Perelman School of Medicine, the University of Pennsylvania, Philadelphia,Pennsylvania. 3Department of Internal Medicine, Liver Research Institute, SeoulNational University Hospital, Seoul, South Korea. 4Department of GastricSurgery, Fujian Medical University Union Hospital, Fujian, China.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

Corresponding Author: Sam S. Yoon, Memorial Sloan Kettering Cancer Center,H-1209, 1275 York Avenue, New York, NY 10065. Phone: 212-639-7436; Fax: 212-639-7460; E-mail: [email protected]

Mol Cancer Res 2019;17:1945–57

doi: 10.1158/1541-7786.MCR-19-0077

�2019 American Association for Cancer Research.

MolecularCancerResearch

www.aacrjournals.org 1945

on June 10, 2021. © 2019 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst June 19, 2019; DOI: 10.1158/1541-7786.MCR-19-0077

http://crossmark.crossref.org/dialog/?doi=10.1158/1541-7786.MCR-19-0077&domain=pdf&date_stamp=2019-8-14http://crossmark.crossref.org/dialog/?doi=10.1158/1541-7786.MCR-19-0077&domain=pdf&date_stamp=2019-8-14http://mcr.aacrjournals.org/

treatment strategies that target these cells would both reversechemotherapy resistance and prevent relapse. Some evidencelinking RTK-RAS signaling to EMT and CSCs comes from Voonand colleagues, who treated Runx3�/� p53�/� murine gastricepithelial cells with TGFb1 to induce EMT and found an increasein the EGFR/Ras gene expression signature (8). The addition ofEGF or the increased expression of Kras led to increased sphereformation and colony formation in soft agar, suggesting that theEGFR/Ras pathway is involved in the promotion of EMT togenerate CSCs. Although the role of the RTK-RAS pathway inEMT andCSCs has beenmore extensively studied in other types ofcancer, there are relatively few studies specifically in gastricadenocarcinoma.

We have previously shown that oncogenic Kras can increasegastric tumorigenesis and metastasis in a genetically engineeredmouse model (9). In gastric adenocarcinoma driven by Cdh1 andTrp53 loss in gastric parietal cells, 69% ofmice developed diffuse-type gastric adenocarcinoma that metastasized to lymph nodes at1 year (10). Combining that with oncogenic Kras (KrasG12D)increased the penetrance of gastric adenocarcinoma developmentto 100% and reduced survival to 76 days. In these mice, bothintestinal and diffuse primary tumors are observed throughoutthe stomach, as well as lymph node, lung, and liver metastases(identified by a YFP reporter). When Tcon mice are treated with aMEK inhibitor starting at 4weeks of age,median survival increasesfrom 76 to 95 days. In the present study, we investigate whetherKRAS activation promotes EMT and acquisition of CSC pheno-types, including metastatic potential and chemoresistance.

Materials and MethodsCell lines and reagents

AGS (RRID CVCL_0139) and NCI-N87 (RRID CVCL_1603)subsequently referred to as N87 are Lauren intestinal-type gastricadenocarcinoma cell lines, and KATOIII (RRID CVCL_0371),SNU-668 (RRID CVCL_5081), and MKN-45 (RRID CVCL_0434)are Lauren diffuse-type gastric adenocarcinoma cell lines.AGS, N87, and MKN-45 cells were obtained from the AmericaType Culture Collection (ATCC). The ATCC uses morphology,karyotyping, and PCR-based approaches to confirm the identityof human cell lines and to rule out both intraspecies and inter-species contamination. These include an assay to detect species-specific variants of the cytochromeCoxidase I gene (COI analysis)to rule out interspecies contamination and short tandem repeatprofiling to distinguish between individual human cell lines andrule out intraspecies contamination. KATOIII and SNU-668 wereobtained from the Korean Cell Line Bank (KCLB). KCLB usesDNA fingerprinting analysis, species verification testing, Myco-plasma contamination testing, and viral contamination testing.Cancer cell lines were actively passaged for less than 6 monthsfrom the time that theywere received from theATCCorKCLB, andUnited Kingdom Co-ordinating Committee on Cancer Research(UKCCCR) guidelines were followed (11). KATOIII cells weremaintained in DMEM, and all others were maintained in RPMI1640. All media were supplemented with 10% FBS, 100 U/mLpenicillin and 100 mg/mL streptomycin, and L-glutamine 2mmol/L ("regular media").

The immortalized human normal gastric epithelial cell lineHFE145 was a gift from Dr. Hassan Ashktorab and Duane T.Smoot (Howard University, Washington, DC), and was main-tained in RPMI 1640 (12). The HFE145 cell line was tested for

Mycoplasma contamination using the ATCC Mycoplasma PCRtesting service. It was not tested for cell line authentication.

5-Fluorouracil (5-FU) and cisplatin were purchased from USBiological and Enzo Life Sciences, respectively. TheMEK inhibitorPD0325901 (S1036) was purchased from Selleckchem. Accutase(AT104) was purchased from Innovative Cell Technologies, Inc.

Growth as spheroidsCells were resuspended in DMEM-F12 containing 20 ng/mL of

EGF, bFGF, N-2 (1X), and B27 (1X; "spheroidmedia") and platedon Ultra-Low Attachment culture dishes (Corning Life Sciences)as previously described (13). Spheroids were collected after 5 to7 days unless otherwise noted. Protein was extracted for analysis,or cells were dissociated with Accutase (Innovative Cell Technol-ogies) and used for other experiments (14).

KRAS shRNA and expression vectorKRAS was silenced via lentiviral transduction of human

KRAS shRNA (SC-35731-V; Santa Cruz Biotechnology) andmouse Kras shRNA (iV048022; abm Inc.). Scramble shRNAcontrol (SC-108080; Santa Cruz Biotechnology) and GFP (sc-108084, Santa Cruz Biotechnology) constructs were also used.Maximal knockdown occurred 72 to 96 hours after transduction.KRASG12V and KRASWT were overexpressed using KRASG12V len-tiviral activation particles (LVP1139-GP; GenTarget Inc.) andKRASWT lentiviral activation particles (LVP201104; abm Inc.),respectively, following the manufacturer's protocol. Control len-tiviral activation particles (sc-437282; Santa Cruz Biotechnology)served as controls.

RAS activity assaysActivated RAS affinity precipitation assays were performed

according to the manufacturer's protocol (9). Briefly, 250 mg ofcell lysates were incubated with 10 mL of RAS-binding domainagarose beads (#14-278, Upstate, Millipore) overnight at 4�C.After washing 3 times with washing buffer (RIPA buffer with 1mmol/L Na3VO4, 10mg/mL leupeptin, 10mg/mL aprotinin, and25mmol/L NaF), immunoprecipitation reactions were separatedby 12% SDS-PAGE, and Ras was detected by Western blotting(antibody from Cell Signaling Technology, #3965), with b-actin(Sigma, A5441) as loading control.

In vitro assaysSpheroid cells were dissociated using Accutase, andmonolayer

cells were collected with trypsin. To assay proliferation, 1 � 104cells were plated onto 96-well flat bottom plates and maintainedin regular media overnight. An MTT Cell Proliferation Assay Kit(Colorimetric) was used to assess cell number by optical densityafter 3 days (15). Day 1 represents the time of cell plating. Datareflect the mean of six samples. Soft-agar colony formation fromsingle cells was assayed as previously described (13). To measuremigration and invasion, cells (2� 104 cells/well) were suspendedin 0.2 mL of serum-free DMEM and loaded in the upper well ofTranswell chambers (8-mmpore size; Corning); the lowerwellwasfilled with 0.8 mL of DMEM with serum. For the invasion assay,the upper wells of the chambers were precoated with BDMatrigelmatrix (BD Biosciences) and 10 mg/mL growth factor; migrationassays employed noncoated Transwell chambers. After incuba-tion for 48 hours at 37�C, cells on the upper surface of the filterwere removedwith a cotton swab, and invading ormigratory cellson the lower surface of the filter werefixed and stainedwith aDiff-

Yoon et al.

Mol Cancer Res; 17(9) September 2019 Molecular Cancer Research1946



http://mcr.aacrjournals.org/

Quick kit (Fisher Scientific) and photographed at 20x magnifi-cation. Invasiveness andmigration were quantified as the averagenumber of cells on phase-contrast microscopy among 5 micro-scopic fields per well.

Fluorescence-activated and magnetic cell sortingFor FACS, cellswere dissociatedusingAccutase (InnovativeCell

Technologies) and resuspended in PBS containing 0.5% BSA.Cells were stained with FITC-conjugated anti-CD44 (BD555478)or isotype control antibody (BD555742) and analyzed on aFACSCalibur (BD Biosciences) using Cell Quest software.CD44-positive cells were collected using a magnetic cell-sortingsystem (Miltenyi Biotech). Briefly, cells were dissociated usingAccutase, stainedwith CD44-Micro Beads, and passed through anLS magnetic column that retains CD44-positive cells. CD44-positive cells were then eluted from the column after removal ofthe magnet and quantified by immunofluorescence using FITC-conjugated CD44 antibody (555478; BD Biosciences).

Western blot analysisSamples were collected in RIPA buffer (Sigma) containing

Complete Protease Inhibitor Cocktail (Roche Diagnostics),and protein concentration was determined by Bio-Rad ProteinAssay. Western blot analysis was performed using the followingantibodies: KRAS (sc-30), Pan-RAS (sc-32), ERK1 (sc-271270),and c-Myc (sc-40) from Santa Cruz Biotechnology; Sox2 (#2748,#3579), Oct-4 (#2750), Nanog (#4893), Slug (#9585), Snail(#3879), phospho-ERK1/2 (#9101), RAS (#3965), MEK1/2(Ser217/221; #4694), phospho-MEK1/2 (Ser217/221; #9121,#9154), CD44 (#3578, #3570), E-cadherin (#14472), vimentin(#3932), and cleaved caspase-3 (#9661) from Cell SignalingTechnology; N-cadherin (BD610920) and E-cadherin(BD610181) from BD Biosciences; Zeb1 (NBP-1-05987) fromNovus Biologicals; and b-actin from Sigma.

Mouse studiesAll mouse protocols were approved by the MSKCC Institution-

al Animal Care and Use Committee. Atp4b-Cre; Cdh1fl/fl; LSL-KrasG12D; Trp53fl/fl; Rosa26LSLYFP/YFP triple conditional (Tcon)mice were generated as previously described (9). Tissues weredissected from indicated mice. Paraffin-embedded sections weredeparaffinized, and sections were examined by IHC as describedbelow.

For subcutaneous flank tumors, 5 � 106 AGS cells previouslytransduced with sh.KRAS or sh.Scr were resuspended in 100 mL ofHank's balanced salt solution and injected subcutaneously intothe right flank of athymic, 6- to 8-week-old male BALB/c nu/numice following isoflurane anesthesia. Mice were assigned intotreatment groups (5 mice per group) when tumors reached 50 to100 mm3 in volume, designated as day 0. Cisplatin (2 mg/kg) orcarrier (PBS) was injected intraperitoneally once a week. Tumorswere measured 3x per week for 2 weeks, and tumor volume (TV)was calculated as TV ¼ length x (width)2 x 0.52. After mice weresacrificed, tumors were excised and cut into thirds.

For induction of lungmetastases, mice were injected via the tailvein with 2 � 107 cells of monolayers and 1 � 104 cells ofspheroids grown from cells previously transduced with sh.KRASor sh.Scr. After sacrifice at different time points, tumor colonies inthe lungs were counted. Each tumor of allmouse groups was fixedin 10%buffered formalin for 24hours, embedded inparaffin, andprocessed into 5 mm sections.

Mouse organoid studiesTcon3077 and Tcon3944 gastric tumors in Tcon mice

were harvested, and organoids were isolated as previouslydescribed (16). For in vitro culture, organoids were mixed with 50mL ofMatrigel (Cat. 354248, BDBioscience) andplated in 24-wellplates. After polymerization of Matrigel, cells were overlaidwith DMEM/F12 supplemented with penicillin/streptomycin,10 mmol/L HEPES, GlutaMAX, 1x B27, 1x N2 (Invitrogen), and1.25 mmol/L N-acetylcysteine (Sigma-Aldrich), and the follow-ing growth factors, cofactors, and hormones were added: 0.05 mg/mL EGF, 0.1 mg/mL fibroblast growth factor-basic (FGF-Basic),0.01 mmol/L gastrin I, 10 mmol/L nicotinamide, 10 mmol/L Y-27632, SB202190 (all from Sigma Aldrich), 1 mmol/L prosta-glandin E2 (Tocris Bioscience), 0.5 mg/mL recombinant R-spondin 1, 0.1 mg/mL mNoggin, 0.1 mg /mL FGF-10 (all fromPeproTech), and 0.1 mg /mL Wnt3A and 0.5 mmol/L A83-01[a TGFb kinase/activin receptor-like kinase (ALK 5) inhibitor;both from R&D Systems]. Organoids were passaged every week ata 1:5–1:8 split ratio by removing them from Matrigel using BDCell Recovery Solution (BD Biosciences) following the manufac-turer's instructions and transferring them to fresh Matrigel.

ImmunocytochemistrySpheroids were fixed with 4% paraformaldehyde and permea-

bilized with 0.1% Triton X-100 in PBS. Following fixation, cellswere incubated with antibodies against Sox2 (#4900; Cell Sig-naling Technology), E-cadherin (BD610181; BD Biosciences),p-MEK1/2 (#9121; Cell Signaling Technology), Slug (#9585;Cell Signaling Technology), and/ or CD44-FITC (555478; BDBiosciences) in a solution of PBS with 1% BSA and 0.1% TritonX-100 at 4�C overnight. Staining was visualized using anti-mouse AlexaFluor 488 (A11005; Life Technologies) and anti-rabbit AlexaFluor 594 (A11012; Life Technologies) with nucle-ar counterstaining using DAPI Solution, and imaging on aninverted confocal microscope. Images were processed usingImaris 7.6.

Immunohistochemistry and immunofluorescenceFor IHC, formalin-fixed, paraffin-embedded sections were

deparaffinized by xylene and rehydrated. Sections were eitherstained with hematoxylin and eosin (H&E) or immunostainedusing the VECTASTAIN Elite ABC Kit (Vector Laboratories Inc.)following the manufacturer's instructions and standard proto-cols (17). Antibodies included CD44 (#3570), p-MEK1/2 (#3579and #9121), p-ERK1/2 (#9101; from BD Biosciences), and Snail(sc-271977, Santa Cruz Biotechnology).

We performed immunofluorescence as previouslydescribed (18). Antibodies used were as follows: anti-humanCD44 (#3570), anti–p-MEK1/2 (#3579), and anti–p-EKR1/2(#9101). Nuclei were counterstained using DAPI. Stained cellswere visualized using an inverted confocal microscope, andimages were processed using Imaris 7.6.

Tissue microarrayTo evaluate expression of CD44, phospho-MEK1/2, and phos-

pho-ERK1/2 in normal and stomach cancer tissues, commerciallyavailable paraffin-embedded tissue array slides containing 40stomach cancer and 8 corresponding normal tissues (A209 II,ISU ABXIS, Seoul, Republic of Korea) were purchased. Sectionswere deparaffinized and then incubated with anti-human CD44(#3570), anti–p-MEK1/2 (#9121), and anti–p-ERK1/2 (#9101,

KRAS in Gastric Cancer Stem–Like Cells

www.aacrjournals.org Mol Cancer Res; 17(9) September 2019 1947




all from BD Biosciences) in a solution of PBS with 1% BSA and0.1%Triton X-100 at 4�Covernight. Stainingwas visualized usinganti-mouse Alexa Fluor 488 and anti-rabbit AlexaFluor 594, withnuclear counterstaining using DAPI and imaging on an invertedconfocal microscope. Images were processed using Imaris 7.6.

Tissue microarrays (TMA) of tumors from 115 patients werealso constructed using a precision tissue array instrument(FMUUH: Shanghai Outdo Biotech Co. Ltd.; MSKCC: BeecherInstruments, Inc.). All subjects provided written-informed con-sent for tumor analyses prospectively, and Institutional ReviewBoard approval was obtained for this study. The study wasconducted in accordance with the International Ethical Guide-lines for Biomedical Research Involving Human Subjects(CIOMS). These patients had adenocarcinoma arising in thestomach or gastroesophageal junction of Siewert type II or IIIand underwent radical gastrectomy between May 2013 andMarch 2014 at FujianMedical University UnionHospital (Fujian,China). A representative core biopsy (2 mm diameter) wasobtained from each tumor and embedded.

IHC for CD44 and p-MEK1/2 was performed as describedabove. The following information was collected from theFMUUH gastric database on each patient: age, gender, tumorinvasion depth, number of lymph node metastases, and time torecurrence and death. The frequency of staining for each markerwas scored as follows:� 5%positive cells¼ 0, 6% to 25%positivecells¼ 1, 26% to 50% positive cells¼ 2,� 51%positive cells¼ 3.If the total score (intensity x percentage score) was < 3, expressionwas defined as low; and if the score was � 4, it was consideredhigh.

Statistical analysisData are represented as mean � SD unless otherwise noted.

Groupswere comparedusing Instant 3.10 software (GraphPad).Pvalues were calculated using the Student t test. For comparisonsamong more than two groups, treatment groups were comparedwith the control group using one-way ANOVA with Bonferroniadjustment for multiple comparisons.

For human data analyses, continuous values are expressed asmean � SD and analyzed using the Student t test. Categoricalvariables are analyzedusingc2 or Fisher exact test.Overall survivalcurves were plotted by the Kaplan–Meier method and comparedusing the log-rank test. Cox proportional hazards regressionmodeling was used to examine the relationship between CD44and p-MEK1/2 expression and survival while controlling forconfounding covariates. Analyseswere performedusing IBMSPSSsoftware for Windows version 21 (IBM). A P value less than 0.05was considered statistically significant.

ResultsEffects of Kras inhibition in organoids derived from mousegastric adenocarcinomas

Tumor-derived organoids are an in vitro model that canrecapitulate the pathophysiology of the original tumors alongwith preserving cellular heterogeneity and self-renewal capac-ity (19). We developed two organoid cultures from primarygastric tumors in Tconmice, labeled Tcon3077 and TconY3944,as described in the Materials and Methods section. Thesetumor-derived organoids express oncogenic KrasG12D. Krasknockdown in organoid cultures by transduced Kras shRNAwas confirmed by Western blot analysis (Supplementary Fig.

S1A) and led to a mild 9%–11% decrease in in vitro cellproliferation at 5 days (Supplementary Fig. S1B). Kras knock-down also decreased organoid size by 65% to 69%, disruptednormal organoid architecture (Fig. 1A), and decreased expres-sion of EMT-related proteins including N-cadherin, Snail, andSlug (Fig. 1B and C). Kras shRNA also decreased migration andinvasion of organoid cells by 53% to 74% (Fig. 1D; Supple-mentary Fig. S1C).

Because EMT can lead to the acquisition of CSC pheno-types (20), we next examined whether Kras knockdown affectedexpression of self-renewal transcription factors and formation ofspheroids (an in vitromethod for enriching for CSCs).Kras shRNAdecreased expression of Sox2 protein (Fig. 1E) and decreasedspheroid formation by 79% to 89% (Fig. 1F; Supplementary Fig.S1D). Thus, Kras knockdown appears to inhibit EMT and con-version to CSCs in these tumor-derived organoids.

If CSCs are the source of metastases, we reasoned that thenumber of CSCs should be higher in microscopic metastasescompared with primary tumors or macroscopic metastases.Takaishi and colleagues tested several gastric cancer cell lines forCSC markers, and only CD44 expression was associated withtumor formation in immunodeficient mice and spheroid colonyformation in vitro (21). We therefore examined primary gastrictumors, microscopic lung metastases, and macroscopic lungmetastases from Tcon mice for expression of the gastric CSCmarker CD44 and found that the number of CD44-expressingcells in microscopic lung metastases was 2.8-fold higher than inprimary gastric tumors and 5.4-fold higher than in macroscopiclungmetastases (Fig. 1G; Supplementary Fig. S1E).We also foundhigher levels of phosphorylated MEK, a downstream target ofKRAS, in lung micrometastases compared with larger primarytumors and lung macrometastases (Fig. 1G; Supplementary Fig.S1F and S1G). These data suggest that oncogenic Kras in CSCspromotes metastasis in this model of gastric adenocarcinoma.

CD44 expression and RTK-RAS activation are associated withworse survival in patients with resectable gastricadenocarcinoma

To further examine the role of CSCs and RTK-RAS activation inhuman gastric adenocarcinomas, we next assessed expression ofCD44 and activation of MEK and ERK by immunostaining acommercially available TMA containing 40 human gastric ade-nocarcinomas and 8 human normal gastric tissues (A209 II, ISUABXIS). Levels of CD44, phospho-MEK, and phospho-ERK levelswere 3.1- to 3.5-fold higher in tumor tissue compared withnormal tissue (Fig. 2A and B).

To determine if CD44 expression and MEK activationare prognostic factors for survival in patients with gastric adeno-carcinoma, we immunostained a TMA of 115 gastric adenocarci-nomas from patients undergoing curative-intent surgical resec-tion at Fujian Medical University Union Hospital (Fujian,China; Fig. 2C). Clinical characteristics of these patients andpathologic characteristic of their tumors are shown in Supple-mentary Table S1. Patients whose tumors expressed high levels ofCD44 or phospho-MEK had significantly worse overall survival(Fig. 2D and E). The worst overall survival was seen in the 32patientswith both highCD44 andhigh phospho-MEK expressionin their tumors (Fig. 2F). Onmultivariate analysis, both increasedexpression of CD44 and phospho-MEK were independent prog-nostic factors for worse overall survival (Supplementary TableS1). Thus, increased tumor expression of the gastric CSC marker

Yoon et al.





sh.KRAS

Tcon3077 Tcon3944

0

50

100

150

200

Org

anoi

d si

ze (µ

m)

- + - +

A

sh.KRAS

B

E

β-Actin

KRAS

Snail

Slug

Zeb1

Tcon3944organoids

Tcon3077organoids

CD44

N-cadherin

sh.KRAS - + - +

sh.Scr

sh.KRAS

Tcon

3077

IHC

-Sna

il

sh.S

crsh

.Kra

s

Tcon3944organoid

Tcon3077organoid

D

200 µm

Lung micrometStomach

Tumor size 150 µm

50 µm

G Lung macromet

Tumor size 600 µm

p-MEK1/2CD44 p-MEK1/2CD44 p-MEK1/2CD44

Tumor size 1 cm

* *

Sox2

c-Myc

β-Actin

Tcon3077

Oct-4

Nanog

-

Tcon3944

+ - +

* *

F

C

E-cadherin

Tcon3944organoid cells

Tcon3077organoid cells

0

5

10

15

20

25

Num

ber o

f sph

eroi

ds

(>50

μm

)

sh.Scrsh.KRAS

* *Migration Invasion

Pos

itive

cel

ls/fi

eld

0

20

40

60

Tcon3077

sh.Scrsh.KRAS

Figure 1.

KRAS knockdown inhibits EMT in tumor organoids from a gastric adenocarcinoma GEMM.A, H&E-stained organoids derived from primary gastric tumors(Tcon3077 and Tcon3944) arising in Tcon mice. Graph shows organoid size. Scale bar, 50 mm. B,Western blot for EMT-related proteins in Tcon3077 andTcon3944 organoids transduced with sh.Kras or sh.Scr. C, Immunostaining of organoids for Snail. Scale bar, 50 mm.D,Migration and invasion as determined bytranswell assays for Tcon3077 cells after transduction with sh.Kras or sh.Scr. E,Western blot for self-renewal proteins in Tcon3077 and Tcon3944 cellstransduced with sh.Kras or sh.Scr and grown as spheroids. b-actin, loading control. F, Spheroid formation of Tcon3077 and Tcon3944 cells following transductionsh.Kras or sh.Scr. G, Primary gastric tumors, microscopic lung metastasis, and macroscopic lung metastasis from Tcon mice following immunofluorescencestaining for DAPI (blue), CD44 (red), and phospho-MEK1/2 (p-MEK1/2, green). Bars represent SD. � , P < 0.05 compared with sh.Scr control.






CD44 and activation of the RTK-RAS pathway as measured byMEK phosphorylation portend a worse prognosis in gastric ade-nocarcinoma patients undergoing surgical resection.

Oncogenic KRAS promotes EMT and acquisition of CSCphenotypes in gastric epithelial cells

We next examined the functional consequences of (1) onco-genic KRAS activation in human gastric epithelial cells and (2)KRAS inhibition in gastric adenocarcinoma cell lines. Forgastric epithelial cells, we used HFE-145 cells which are immor-talized human, nonneoplastic gastric epithelial cells. The gas-tric adenocarcinoma cell lines examined included three humandiffuse-type gastric adenocarcinoma cell lines (MKN-45, SNU-

668, and KATOIII), two human intestinal-type gastric adeno-carcinoma cell lines (AGS and NCI-N87), and two mousegastric adenocarcinoma cell lines (Tcon3077 and Tcon3944;ref. 9). We performed targeted next-generation sequencing onour human gastric cancer cell lines and found that two gastricadenocarcinoma cell lines had activating KRASmutations. AGScells have a KRASG12D mutation, and SNU-668 cells have aKRASQ61K mutation. The other gastric adenocarcinoma celllines (MKN-45, NCI-N87, and KATOIII) are KRAS wild-type.We first confirmed that KRAS activation was low in normalgastric epithelial (HFE-145) cells compared with gastric ade-nocarcinoma cell lines, as measured by coprecipitation of KRASwith its signaling target, RAF, and expression of phospho-MEK

Figure 2.

Correlation between CD44expression and RTK-RAS pathwayactivation and overall survival inpatients with resectable gastricadenocarcinoma. A and B,Immunofluorescence staining ofcommercially available tissue arrayslide containing 40 stomach cancerand 8 corresponding normal tissuesfor DAPI (blue), CD44 (green),p-MEK1/2 (red), and p-ERK1/2 (red).Scale bar, 50 mm. C, IHC staining forCD44 and phospho-MEK1/2 of tumorTMAs created from tumors frompatients whose gastricadenocarcinomawas resected atFujian Medical University UnionHospital (FMUUH; in Fujian, China).Scale bar, 50 mm. D–F, Kaplan–Meieroverall survival curves stratified byexpression of CD44 (D), phospho-MEK (E), and both CD44 andphospho-MEK (F) in the FMUUHcohort.

Yoon et al.





A

p-EKR1/2

p-MEK1/2

MEK1/2

KRASH

FE-1

45

MK

N45

Human gastric cancer cell lines

NC

I-N87

ERK2

β-Actin

AGS

KA

TOIII

SN

U-6

68

AC

KP

Y 30

77

AC

KP

Y 39

44

Mouse gastric cancer cell lines

B

KRAS genotype: WT WT WT MUT WT MUT MUT MUT

E-cadherinp-MEK1/2

Vect

orK

RA

SG

12V

KR

ASW

THFE

-145

D

C

InvasionMigration

Vect

orK

RA

SG

12V

KR

ASW

T

HFE-145E

G

Zeb1

KRASWTKRASG12V

HFE-145

K-Ras

E-cadherin

N-cadherin

Slug

Snail

Vimentin

β-Actin

--

-+

+-

KRASWTKRASG12V

HFE-145spheroids

CD44

Sox2

Oct4

β-Actin

c-Myc

Nanog

--

-+

+-

IP: RAF agaroseIB:KRAS

HF

KRASWT

DA

PI

CD

44S

ox2

Mer

ge

KRASG12VVector

HFE-145spheroids

0

5

10

15

20

KRASWTKRASG12V

--

-+

+-

HFE-145 Spheroids

Vect

orK

RA

SG

12V

HFE-145

KR

ASW

T

0

10

20

30N

umbe

r of c

ells

per

fiel

d

Migration Invasion

Vector

KRASWTKRASG12V

Num

ber o

f sph

eroi

ds

(>50

μm

)

* **

**

* **

*

Figure 3.

Oncogenic KRAS promotes EMT and acquisition of CSC phenotypes in gastric epithelial cells. A,Western blot for RTK-RAS pathway proteins in gastric epithelialcells and human andmouse gastric cancer cell lines. The presence (MUT) or absence (WT) of an oncogenic KRASmutation is denoted. B,Western blot for EMT-related proteins in HFE-145 cells transduced with oncogenic KRAS (KRASG12V), wild-type KRAS (KRASWT), or control (Vector). C,Morphology of HFE-145 in tissueculture following transduction with KRASG12D, KRASWT, or Vector. Scale bar, 10 mm. D, Confocal photos following immunofluorescent staining of HFE-145 cellstransduced with KRASG12D, KRASWT, or Vector for E-cadherin (green) and p-MEK1/2 (red). Scale bar, 20 mm. E,Migration and invasion assays for HFE-145 cellstransduced with KRASG12D, KRASWT, or Vector as determined by transwell assay. F, Immunofluorescence of HFE-145 spheroids for DAPI (blue), CD44 (green),and Sox2 (red) following transduction with KRASG12D, KRASWT, or Vector. Scale bar, 50 mm. G,Western blot for self-renewal proteins in HFE-145 cells transducedwith KRASG12D, KRASWT, or Vector. H, Photos and graph of HFE-145 cells following transduction of KRASG12D, KRASWT, or Vector and grown in spheroid formationconditions. Scale bar, 50 mm. Bars represent SD. � , P < 0.05.






and phospho-ERK (Fig. 3A). Successful transduction of HFE-145 cells with oncogenic KRAS (�KRASG12V) or wild-type KRAS(KRASWT) was confirmed by Western blot; expression of KRASand levels of phospho-MEK and phospho-ERK were increasedrelative to control cells (Supplementary Fig. S2A). KRASG12V

mildly increased proliferation of HFE-145 cells in vitro (Sup-plementary Fig. S2B), significantly decreased expression ofthe epithelial marker E-cadherin, and significantly increasedexpression of the mesenchymal marker vimentin and theEMT transcription factor Slug (Fig. 3B). Expression of otherEMT transcription factors including Snail and Zeb1 was notsignificantly changed. HFE-145 cells expressing oncogenicKRASG12V also displayed a more spindle cell morphologycompared with control cells (Fig. 3C). Loss of E-cadherinexpression and increased MEK phosphorylation in KRASG12V-expressing HFE145 cells was also apparent by immunofluores-cence (Fig. 3D). Finally, KRASG12V increased migration andinvasion by 15- to 17-fold compared with control HFE-145cells (Fig. 3E). KRASWT transduction had an intermediate effect.

We next examined the effect of KRASG12V on CSC phenotypesand marker expression in HFE-145 cells. By FACS analysis, CD44expression increased from 0.5% in HFE-145 monolayers to 4.4%in control HFE-145 spheroid cells and 6.4% inHFE spheroid cellsexpressing oncogenic KRASG12V (Supplementary Fig. S2C andS2D). KRASG12V also increased expressionof self-renewal proteinssuch as Sox2 as assessed by immunofluorescence (Fig. 3F) andWestern blot (Fig. 3G). Compared with control cells, HFE-145cells expressing oncogenic KRASG12V formed 2.6-foldmore spher-oids (Fig. 3H). Again, KRASWT transduction had an intermediateeffect. Together, these results show that expression of oncogenicKRAS in gastric epithelial cells promotes EMT and acquisition ofCSC properties.

KRAS activity controls CSC phenotypes in human tumor–derived gastric adenocarcinoma cells

We assessed KRAS activation in gastric adenocarcinoma cellsgrown as monolayers and as spheroids via KRAS coprecipitationwith RAF and phosphorylation of MEK and ERK (Fig. 4A). In allgastric adenocarcinoma cell lines, KRAS pathway activation washigher in cells grown as spheroids compared with cells grown asmonolayers, as was expression of CD44 and Sox2 (Supplemen-tary Fig. S3A).

In KATOIII cells, which express wild-type KRAS, transductionof oncogenic KRASG12V increased spheroid formation capacity(Fig. 4B) and expression of CD44 and Sox2 (Fig. 4C). Knock-down of KRAS using lentiviral shRNA in AGS and KATOIII cellsresulted in loss of CD44 and Sox2 expression as measured byWestern blot (Fig. 4D) and immunofluorescence (Supplemen-tary Fig. S3B) as well as a 74% to 85% reduction in spheroidformation in both standard (Fig. 4E) and single-cell assays(Supplementary Fig. S3C).

We next performed a series of similar experiments using theMEK inhibitor PD0325901 rather than using KRAS shRNA. ThisMEK inhibitor reduced CD44 expression and Sox2 expression inAGS and KATOIII cells grown as spheroids as measured byWestern blot analysis (Supplementary Fig. S4A) and by immu-nofluorescence (Supplementary Fig. S4B). PD0325901 alsoreduced spheroid formation by AGS and KATOIII cells at 5 daysby 68% to 91% (Supplementary Fig. S4C), and the size ofspheroids formed from single AGS and KATO III cells by 32%to 42% (Supplementary Fig. S4D).

We next examined AGS and KATO spheroid cells that weresorted into CD44(þ) and CD44(-) expression by FACS. In thesespheroids, 7% to 14% of cells are CD44(þ) (data not shown).PD0325901 attenuated ERK phosphorylation and expression ofSox2 in CD44(þ) cells (Supplementary Fig. S5A), and greatlyreduced their ability to form spheroids (Supplementary Fig. S5B).MEK inhibition in CD44þ spheroid cells also decreased expres-sion of EMT-related proteins, increased expression of the epithe-lial marker E-cadherin (Supplementary Fig. S5C), and dramati-cally reduced migration and invasion (Supplementary Fig. S5D).

As CSCs are generally resistant to chemotherapy (22), weexamined the effects of KRAS inhibition on CSC sensitivity to5-FU and cisplatin. AGS and KATOIII cells grown as monolayerswere sensitive to these agents, whereas these same cells grown asspheroids were relatively resistant (Supplementary Fig. S6A).KRAS shRNA reversed chemotherapy resistance in AGS andKATOIII spheroid cells (Supplementary Fig. S6B). AGS flankxenografts were muchmore sensitive to cisplatin when KRAS wasknocked down (Supplementary Fig. S6C). The effect of chemo-therapy and inhibition of KRAS on AGS flank xenografts wasexamined for proliferation using Ki67 staining, for total apoptosisusing cleaved caspase 3 staining, and for stemness using CD44staining. Cisplatin combined with KRAS shRNA led to dramaticincreases in apoptosis in CD44-positive cells compared withchemotherapy (Supplementary Fig. S6D).

KRAS knockdown inhibits EMT and metastasis in humantumor–derived gastric adenocarcinoma cells

We next examined the effect of reducing KRAS activity on EMTand metastasis. AGS and KATOIII cells were transduced withKRAS shRNAor a scramble control shRNA, andKRAS knockdownwas confirmed by Western blot (Fig. 5A). KRAS knockdownsignificantly decreased expression of the mesenchymal markerN-cadherin, increased expression of the epithelial marker E-cad-herin, and decreased expression of the EMT transcription factorSnail in both cell lines (Fig. 5A and B). Expression of other EMTtranscription factors including Slug andZeb1wasnot significantlychanged. KRAS knockdown also decreased migration and inva-sion by 85% to 90% (Fig. 5C) and reduced colony formation insoft agar by 86% to 87% (Fig. 5D) compared with control cells.Similar results were obtained using the MEK inhibitorPD0325901 (Supplementary Fig. S7A–S7C).

We next assessed the effect of KRAS knockdown on tumorgrowth and invasiveness in vivo by creating flank tumor xenograftsusing AGS spheroid cells transduced with KRAS shRNA or ascrambled control shRNA (Supplementary Fig. S7D). Flanktumors from control AGS cells grew faster than flank tumorsfrom AGS cells with KRAS knockdown (Fig. 5E). Reducing KRASactivity also drastically inhibited invasion. Control AGS cellsformed tumors with an infiltrating leading edge, whereas AGScells with KRAS knockdown had a well-defined border betweenthe tumor and normal surrounding tissues (Fig. 5E). KRAS shRNAreduced the number of infiltrating cells from 8.2 per mm2 to 1.6per mm2 (Fig. 5F).

EMT and acquisition of CSC characteristics promote solidtumor metastasis (7). We examined the effect of KRAS knock-down on experimental lung metastases using a tail vein injectionmodel in athymic nude mice. AGS cells grown as spheroids andinjected into the tail vein of mice produced significantly morelung metastases than the same cells grown as monolayers(Fig. 6A). KRAS knockdown in cells grown as spheroids or as

Yoon et al.





monolayers dramatically reduced the formation of lung metas-tases, leading to longer survival (Fig. 6A and B). Lung metastasesfrom spheroid cells in which KRAS was knocked down haddramatically lower levels of CD44 and phospho-MEK comparedwith scramble control cells (Fig. 6C and D). Taken together, thesedata show that KRAS promotes EMT and metastasis in gastricadenocarcinoma cells.

DiscussionThis study demonstrates that RTK-RAS signaling promotes EMT

in gastric adenocarcinoma cells, leading to the acquisition of CSCphenotypes and metastatic potential. Transduction of nontrans-formed gastric epithelial cells with oncogenic KRAS led to EMT,expression of the CSC marker CD44 and stem cell transcription

KRAS1 2

Myc tag

KRASG12V

Sox2

Oct-4

Vect

or

CD44

β-Actin

KATOIII spheroids

Nanog

c-Myc

C

Mon

olay

er

KATOIII(KRASWT)

Sph

eroi

dsKRAS

IP: RAF agaroseIB:KRAS

p-ERK1/2

ERK-1/2

β-Actin

A

p-MEK1/2

MEK1/2

sh.S

crsh

.KR

AS

AGSspheroids

E

Mon

olay

ers

MKN-45(KRASWT)

Sph

eroi

ds

Mon

olay

ers

AGS(KRASG12D)

Sph

eroi

ds

D KATOIIIspheroids

sh.S

cr

AGSspheroids

sh.K

RA

S

KRAS

sh.S

crKATOIII

spheroidssh

.KR

AS

CD44

Sox2

Oct-4

Nanog

c-Myc

β-Actin

Mon

olay

ers

SNU668(KRASQ61K)

Sph

eroi

ds

4 Days

KATOIII

Vect

or

0 2 4

B

KR

AS

G12

V

0

20

40

60

80

0 2 4Days

VectorKRASG12V

*

*

Sph

eroi

d di

amet

er (μ

m)

0

5

10

15

20

* *

Num

ber o

f sph

eroi

ds(>

50 μ

m)

AGSspheroids

KATOIIIspheroids

sh.Scrsh.KRAS

Figure 4.

RTK-RAS pathway controls CSC phenotypes in human gastric adenocarcinoma–derived spheroids. A,Western blots for RTK-RAS pathway proteins in KATOIII,AGS, MNK-45, and SNU668 cell grown as monolayers or as spheroids. B, Single-cell assay in KATOIII spheroid cells following transduction KRASG12D or Vector.Graph shows diameter of spheroids at selected time points when grown in spheroid formation conditions. C,Western blot analysis of KATOIII cells at selectedtime points following transduction of Myc-tagged KRASG12D or Vector in spheroid formation condition for levels of total KRAS, Myc tag, CD44, and self-renewalproteins. D,Western blot of AGS and KATOIII cells transduced with KRAS shRNA (sh.KRAS) or control (sh.Scr) and grown in spheroid formation conditions forKRAS, CD44, and self-renewal proteins. E, Spheroid formation assay for AGS and KATOIII cells following transduction with sh.KRAS or sh.Scr. Bars represent SD.� , P < 0.05 compared with control.






sh.Scr sh.KRAS sh.Scr sh.KRASMigration Invasion

AG

S

Sph

eroi

dsK

ATO

III

Sph

eroi

ds

A

sh.S

cr

AGSspheroids

sh.K

RA

S

KRAS

sh.S

cr

KATOIIIspheroids

sh.K

RA

S

E-cadherin

N-cadherin

Snail

Slug

Zeb1

β-Actin

Bsh

.Scr

sh.K

RA

S

H&E Snail DAPI

50 µm 0.5 cm

AGS Xenografts

E

0

20

40

60

0

20

40

60

Csh.Scr sh.KRAS

AG

S

KA

TOIII

S

pher

oids

Sph

eroi

ds

D Soft agar

****

50 µm

CD44/Snail

Rel

ativ

e ex

pres

sion

CD44 Snail

KATOIIIspheroids

CD44 Snail

sh.Scr sh.KRAS

AGSspheroids

CD44/Snail

sh.Scr sh.KRAS

**

0

3

6

9

12

sh.KRAS

*

sh.Scr

Num

ber o

f Sna

il (+

)-

infil

tratin

g ce

lls p

er m

m2 F

Mig

rate

d ce

lls

Inva

sive

cel

ls

sh.Scrsh.KRAS

AGSspheroids

KATOIIIspheroids

AGSspheroids

KATOIIIspheroids

sh.Scrsh.KRAS

Col

onie

s pe

r fie

ld

0

10

20

30

AGSspheroids

KATOIIIspheroids

sh.Scrsh.KRAS

**

0

0.5

1

0

0.5

1sh.Scr sh.KRAS

Figure 5.

KRAS knockdown inhibits EMT and infiltrative behavior of human gastric adenocarcinoma–derived cells. A,Western blot for KRAS and EMT proteins in spheroidsof AGS and KATOIII transduced with sh.KRAS or sh.Scr. B, Immunofluorescence photos of AGS and KATOIII spheroids for DAPI (blue), E-cadherin (green), andSnail (red) following transduction with sh.KRAS or sh.Scr. Graphs show relative expression of CD44 and Snail. Scale bar, 50 mm. C and D,Migration and invasionassays (C) and soft-agar assay (D) for spheroids of AGS and KATOIII transduced with sh.KRAS or sh.Scr. E, Tumor and H&E pictures of flank xenografts from AGScells stably transduced with sh.KRAS or sh.Scr. Photos of immunofluorescence staining with infiltrating cells for DAPI (blue) and Snail (green). Dashed lineindicates tumor border. Graph showing number of Snail(þ)-infiltrating cells. F,Number of Snail(þ) cells infiltrating beyond tumor border. Bars represent SD.� , P < 0.05 compared with control.

Yoon et al.





sh.Scr sh.KRAS

Monolayer

H&

E

sh.Scr sh.KRAS

Spheroids

CD

44A

C

p-M

EK

1/2

D

B

0 3 6 9 12 15 18Weeks

100

80

60

40

20

0

Cum

ulat

ive

surv

ival n = 6

n = 6n = 5

n = 5

MonolayerMonolayer / sh.KRASSpheroidsSpheroids

/ sh.Scr

/ sh.Scr/ sh.KRAS

sh.Scr sh.KRAS

Mon

olay

erS

pher

oids

AG

S

0.0

0.5

1.0

1.5

2.0

CD44 p-MEK1/2

Lung metastasis

Monolayer-Monolayer-SpheroidsSpheroids

sh.Scrsh.KRAS

- sh.Scr- sh.KRAS

Rel

ativ

e ex

pres

sion

0

5

10

15

20

25

AGS

* *

Num

ber o

f m

etas

tatic

lung

nodu

les

sh. KRAS - + - +

Monolayers Spheroids

Figure 6.

KRAS knockdown in gastric adenocarcinoma spheroid cells inhibits experimental metastasis. Monolayer (2� 107 cells per mouse) or spheroid AGS cells (1� 104cells per mouse) transduced with sh.KRAS or sh.Scr were injected intravenously into 7-week-old SCID mice through the tail vein. A, Representative images andquantitation of detectable lung nodules on the surface of whole lungs. Dashed lines denote lung nodules. Scale bar, 0.5 cm. B, Kaplan–Meier survival curves formice injected via tail vein with AGSmonolayer and spheroid cells transduced with sh.KRAS or sh.Scr. C, H&E staining and immunofluorescence for CD44 andphospho-MEK1/2 in tumors derived from indicated cells. Dashed lines denote the lung nodules. Scale bar, 100 mm. D,Quantitation of immunofluorescence forindicated proteins in lung metastases. Bars represent SD. � , P < 0.05 compared with control.






factor Sox2, and increased spheroid formation. Gastric adenocar-cinoma cell lines grown as spheroids had enrichment of CD44expression and increased KRAS activity compared with monolay-er cells. Inhibition of KRAS in gastric adenocarcinoma spheroidcells or CD44(þ) cells using shRNA knockdown or pharmaco-logic inhibition diminished expression of Sox-2, reduced spher-oid formation, and reversed chemotherapy resistance. KRASinhibition in gastric adenocarcinoma spheroid cells also reducedexpression of EMT markers including N-cadherin and Snail,greatly diminished migratory and invasive capabilities, attenuat-ed the infiltrative nature of flank xenografts, and reduced exper-imental lung metastases. KRAS inhibition in organoids derivedfrom gastric tumors in amouse model of gastric adenocarcinomawith oncogenic Kras resulted in similar reductions in EMT andCSC phenotypes. Finally, tumor specimens from patients withgastric adenocarcinomawhounderwent surgical resection of theirgastric tumors revealed an association betweenhighCD44expres-sion and RTK-Ras activity in their tumors and worse overallsurvival.

Although the relationship between KRAS activation and CSCfunction has not been extensively studied in gastric adenocarci-noma, there have been some studies in other gastrointestinaltumors. In colorectal cancer, Blaj and colleagues found that highMAPK activity promotes EMT and marks a progenitor cell sub-population that was the predominant source of growing flankxenografts (23). Also, in colorectal cancer, Moon and colleaguesshowed that in cells carrying mutated APC, oncogenic KRASincreases expression of CSC markers (CD44, CD133, andCD166), spheroid formation, and the size of xenografts. Inpancreatic CSCs, inhibition of KRAS led to downregulation ofJNK signaling and loss of self-renewal and tumor-initiating capac-ity (24). In this study, we used spheroid formation and expressionof CD44 as means of identifying gastric adenocarcinoma CSCs.These gastric adenocarcinoma spheroid cells or CD44(þ) cellshave increased activation of the KRAS pathway as measured bybinding to RAF and phosphorylation of MEK and ERK. They alsohave elevated expression of the stem cell factor Slug along withincreased colony formation in soft agar. Inhibition of KRAS ingastric adenocarcinoma cells using either shRNA or MEK inhibi-tion blockedCSCphenotypes including spheroid formation, soft-agar colony formation, and chemotherapy resistance. Thus, KRASactivity may be a common pathway supporting CSC function ingastrointestinal cancers.

Our results also support a link between the EMT program, anaturally occurring transdifferentiation program long known tosupport the proliferation and metastasis of cancer cells (25),and the generation of CSCs. The phenotypic differencesbetween CSCs and bulk tumor cells are predominately due toepigenetic changes which activate the EMT program (7). Thislink between the passage through EMT and the acquisition ofstem-like properties is vital for cancer cells in order to metas-tasize and to survive chemotherapy, and this study highlightsthe role that KRAS plays in this process. In this study, we foundthat blockade of KRAS activity in gastric adenocarcinoma CSCsdownregulates the EMT-related proteins Slug and N-cadherinand reduces migration, invasion, metastasis, and chemothera-py resistance. In Tcon mice, we found that there were moreCD44(þ) cells and higher phosphorylated MEK expression inmicrometastases compared with macroscopic tumors, suggest-ing CD44(þ) CSC with higher RTK-RAS activity may be thesource of metastatic disease, but further studies are needed to

prove this hypothesis. In addition to studies in Tcon mice,which develop primary gastric adenocarcinomas and metasta-ses with oncogenic Kras, we used tumor-derived organoidsfrom Tcon tumors as an in vitro model of Tcon tumors topreserve tumor heterogeneity and self-renewal capacity."

Because RAS GTPases including KRAS are difficult to targetdirectly with drugs because of structure-function considera-tions (26), we inhibited the KRAS pathway using an MEK inhib-itor. Several MEK inhibitors including trametinib, cobimetinib,and binimetinib are currently FDA-approved for use in patientswith BRAF-mutated melanoma (27). The MEK inhibitor used inthis study, PD0325901, is currently in clinical trials for patientswith various solid tumors. The inhibitory effects of PD0325901on spheroid formation, soft-agar colony formation, EMT proteinexpression,migration, and invasionwere similar to those of KRASknockdown by shRNA.

Our results also suggest a potential strategy for targeting ther-apies that inhibit the RTK-RAS pathway to a subgroup of patientswith gastric adenocarcinoma. Such therapies are urgently neededto overcome the challenge of chemotherapy resistance. Theimportance of appropriately selecting patients for targeted ther-apies is demonstrated by the success of trastuzumab plus che-motherapy in patients whose gastric adenocarcinoma overex-presses HER-2, which prolonged survival from 11 to14 months (28), compared with the lack of benefit of combiningcytotoxic chemotherapy with agents targeting the vascular endo-thelial growth factor A or EGF pathways in unselectedpatients (29–31). RTK-RAS pathway inhibition may only beeffective for a subset of gastric adenocarcinomas with highRTK-RAS activity. Our finding that patients with increased tumorlevels of CD44 and increased activation of the RAS pathway hadsignificantly worse overall survival after resection of their tumorssuggests that this may be a subgroup in which targeting the RTK-RAS pathway would be most beneficial.

Our use of tumor-derived organoids allowed us to investigatethe relationship between KRAS activation and CSC properties inheterogeneous tumors in vitro (32). Tumor organoids can bemanipulated under controlled conditions in ways that are notpossible even in genetically engineered mouse models. We gen-erated gastric cancer organoids from gastric tumors that devel-oped in our Tcon genetically engineered mouse model (9). Weperformed many of the CSC and EMT assays on both gastricadenocarcinoma cell lines and organoids and confirmed thatKRAS activation in gastric adenocarcinoma stimulates EMT toCSCs in both settings.

To the best of our knowledge, this is the first study to establishthe importance of KRAS in gastric adenocarcinoma in terms ofpromoting EMT and the generation of CSCs and to demonstratethat inhibition of KRAS may be a strategy for blocking metastaticprogression. Studies were performed using human and mousegastric adenocarcinoma cells, nontransformed gastric epithelialcells, and organoids derived from a GEMM of gastric adenocar-cinoma with oncogenic Kras in vitro, and the relevance of thesestudies was confirmed in analyses of mouse and human tumors.These studies provide rationale for studying inhibitors of theKRAS pathway to block metastasis and reverse chemotherapyresistance in gastric adenocarcinoma.

Disclosure of Potential Conflicts of InterestS. Ryeom has an ownership interest (including stock, patents, etc.) in Attis

Lab. No potential conflicts of interest were disclosed by the other authors.

Yoon et al.





Authors' ContributionsConception and design: C. Yoon, J.-x. Lin, S.S. YoonDevelopment of methodology: C. Yoon, J. Till, S.S. YoonAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Yoon, J. Till, S.-J. Cho, K.K. Chang, S.S. YoonAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Yoon, S.-J. Cho, J.-x. Lin, C.-m. Huang, S. Ryeom,S.S. YoonWriting, review, and/or revision of the manuscript: K.K. Chang, S. Ryeom,S.S. YoonAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): C. Yoon, J. Till, S.S. YoonStudy supervision: S.S. Yoon

AcknowledgmentsWe would like to thank Drs. Hassan Ashktorab and Duane T. Smoot for

providing the HFE-145 cell line. We also thank MSKCC senior editor JessicaMoore for reviewing this article. This study was funded by NIH grants 1R01CA158301-01 (S.S. Yoon) and P30 CA008748 (S.S. Yoon) and by theDeGregorio Family Foundation Grant (S.S. Yoon and S. Ryeom).

The costs of publicationof 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 23, 2019; revised May 7, 2019; accepted June 10, 2019;published first June 19, 2019.

References1. Torre LA, Siegel RL, Ward EM, Jemal A. Global cancer incidence and

mortality rates and trends–an update. Cancer Epidemiol Biomarkers Prev2016;25:16–27.

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

3. Cancer Genome Atlas Research Network. Comprehensive molecular char-acterization of gastric adenocarcinoma. Nature 2014;513:202–9.

4. Samatar AA, Poulikakos PI. Targeting RAS-ERK signalling in cancer:promises and challenges. Nat Rev Drug Discov 2014;13:928–42.

5. Alison MR, Lin WR, Lim SM, Nicholson LJ. Cancer stem cells: in the line offire. Cancer Treat Rev 2012;38:589–98.

6. Ye X, Weinberg RA. Epithelial-mesenchymal plasticity: a central regulatorof cancer progression. Trends Cell Biol 2015;25:675–86.

7. Shibue T, Weinberg RA. EMT, CSCs, and drug resistance: the mechanisticlink and clinical implications. Nat Rev Clin Oncol 2017;14:611–29.

8. Voon DC, Wang H, Koo JK, Chai JH, Hor YT, Tan TZ, et al. EMT-inducedstemness and tumorigenicity are fueled by the EGFR/Ras pathway.PLoS One 2013;8:e70427.

9. Till JE, Yoon C, Kim BJ, Roby K, Addai P, Jonokuchi E, et al. OncogenicKRAS and p53 loss drive gastric tumorigenesis in mice that can beattenuated by E-cadherin expression. Cancer Res 2017;77:5349–59.

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. UKCCCR guidelines for the use of cell lines in cancer research. Br J Cancer2000;82:1495–509.

12. Shin JY, Kim YI, Cho SJ, Lee MK, Kook MC, Lee JH, et al. MicroRNA 135asuppresses lymph node metastasis through down-regulation of ROCK1 inearly gastric cancer. PLoS One 2014;9:e85205.

13. Yoon C, Park DJ, Schmidt B, Thomas NJ, Lee HJ, Kim TS, et al. CD44expression denotes a subpopulation of gastric cancer cells in whichHedgehog signaling promotes chemotherapy resistance. Clin Cancer Res2014;20:3974–88.

14. Bisson I, Prowse DM. WNT signaling regulates self-renewal and differen-tiation of prostate cancer cells with stem cell characteristics. Cell Res 2009;19:683–97.

15. Yoon SS, Eto H, Lin CM, Nakamura H, Pawlik TM, Song SU, et al. Mouseendostatin inhibits the formation of lung and liver metastases. Cancer Res1999;59:6251–6.

16. Vlachogiannis G,Hedayat S, Vatsiou A, Jamin Y, Fernandez-Mateos J, KhanK, et al. Patient-derived organoids model treatment response of metastaticgastrointestinal cancers. Science 2018;359:920–6.

17. Cho SJ, Kook MC, Lee JH, Shin JY, Park J, Bae YK, et al. Peroxisomeproliferator-activated receptor gamma upregulates galectin-9 and predictsprognosis in intestinal-type gastric cancer. Int J Cancer 2015;136:810–20.

18. Yoon C, Cho SJ, Aksoy BA, Park DJ, Schultz N, Ryeom SW, et al. Chemo-therapy resistance in diffuse-type gastric adenocarcinoma is mediated byRhoAactivation in cancer stem-like cells. ClinCancer Res 2016;22:971–83.

19. McCracken KW, Cata EM, Crawford CM, Sinagoga KL, Schumacher M,Rockich BE, et al. Modelling human development and disease in plurip-otent stem-cell-derived gastric organoids. Nature 2014;516:400–4.

20. Scheel C, Weinberg RA. Cancer stem cells and epithelial-mesenchymaltransition: concepts and molecular links. Semin Cancer Biol 2012;22:396–403.

21. Takaishi S, Okumura T, Tu S, Wang SS, Shibata W, Vigneshwaran R, et al.Identification of gastric cancer stem cells using the cell surface markerCD44. Stem Cells 2009;27:1006–20.

22. Alison MR, LinWR, Lim SM, Nicholson LJ. Cancer stem cells: in the line offire. Cancer Treat Rev 2012;38:589–98.

23. Blaj C, Schmidt EM, Lamprecht S, Hermeking H, Jung A, Kirchner T, et al.Oncogenic effects of high MAPK activity in colorectal cancer mark pro-genitor cells and persist irrespective of RASmutations. Cancer Res 2017;77:1763–74.

24. OkadaM, Shibuya K, Sato A, Seino S, Suzuki S, SeinoM, et al. Targeting theK-Ras–JNK axis eliminates cancer stem-like cells and prevents pancreatictumor formation. Oncotarget 2014;5:5100–12.

25. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition.J Clin Invest 2009;119:1420–8.

26. Cox AD, Fesik SW, Kimmelman AC, Luo J, Der CJ. Drugging the undrug-gable RAS: mission possible? Nat Rev Drug Discov 2014;13:828–51.

27. ZhaoY,Adjei AA. The clinical development ofMEK inhibitors.Nat RevClinOncol 2014;11:385–400.

28. Bang YJ, Van Cutsem E, Feyereislova A, Chung HC, Shen L, Sawaki A, et al.Trastuzumab in combination with chemotherapy versus chemotherapyalone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomisedcontrolled trial. Lancet 2010;376:687–97.

29. Lordick F, Kang YK, Chung HC, Salman P, Oh SC, Bodoky G, et al.Capecitabine and cisplatin with or without cetuximab for patients withpreviously untreated advanced gastric cancer (EXPAND): a randomised,open-label phase 3 trial. Lancet Oncol 2013;14:490–9.

30. Ohtsu A, Shah MA, Van Cutsem E, Rha SY, Sawaki A, Park SR, et al.Bevacizumab in combination with chemotherapy as first-line therapy inadvanced gastric cancer: a randomized, double-blind, placebo-controlledphase III study. J Clin Oncol 2011;29:3968–76.

31. Waddell T, Chau I,CunninghamD,GonzalezD,Okines AF,OkinesC, et al.Epirubicin, oxaliplatin, and capecitabine with or without panitumumabfor patients with previously untreated advanced oesophagogastric cancer(REAL3): a randomised, open-label phase 3 trial. Lancet Oncol 2013;14:481–9.

32. Driehuis E, CleversH.CRISPR/Cas 9 genome editing and its applications inorganoids. Am J Physiol Gastrointest Liver Physiol 2017;312:G257–65.






2019;17:1945-1957. Published OnlineFirst June 19, 2019.Mol Cancer Res Changhwan Yoon, Jacob Till, Soo-Jeong Cho, et al. and Promotes Metastasis

Like Cells−Epithelial-to-Mesenchymal Transition to Cancer Stem KRAS Activation in Gastric Adenocarcinoma Stimulates

Updated version

10.1158/1541-7786.MCR-19-0077doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2019/06/19/1541-7786.MCR-19-0077.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/17/9/1945.full#ref-list-1

This article cites 32 articles, 10 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/17/9/1945.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://mcr.aacrjournals.org/content/17/9/1945To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-19-0077http://mcr.aacrjournals.org/content/suppl/2019/06/19/1541-7786.MCR-19-0077.DC1http://mcr.aacrjournals.org/content/17/9/1945.full#ref-list-1http://mcr.aacrjournals.org/content/17/9/1945.full#related-urlshttp://mcr.aacrjournals.org/cgi/alertsmailto:[email protected]://mcr.aacrjournals.org/content/17/9/1945http://mcr.aacrjournals.org/

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 200 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 900 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed false >> > ]>> setdistillerparams> setpagedevice

kras activation in gastric adenocarcinoma stimulates epithelial-to-mesenchymal ... · this...

Documents