research secreted and membrane-bound isoforms of protease

Tum

SecHav

Jessic

Abst

Intro

Oneity of tgage frand intant ormicroegressimetasmembstromlease aefficienchemoangiogPro

bound(MMPin proexposeand co(3, 4).ECM-

AuthorMedica

CorresIsrael D633A, BE-mail:

doi: 10

©2010

www.a

Published OnlineFirst August 24, 2010; DOI: 10.1158/0008-5472.CAN-09-4231
Published OnlineFirst on September 21, 2010 as 10.1158/0008-5472.CAN-09-4231
Canceresearch

or and Stem Cell Biology

reted and Membrane-Bound Isoforms of Protease ADAM9

R

e Opposing Effects on Breast Cancer Cell Migration

a L. Fry and Alex Toker

ractTum

crine fADAMcell-cetherebpressADAMADAMby AD

Proteolysbound gro

s' Affiliatiol Center, Ha

ponding Aeaconessoston, MAatoker@bid

.1158/0008-

American A

acrjourna

Downlo

or cell migration is mediated by cell-autonomous signaling mechanisms as well as paracrine and auto-actors secreted by activated stromal cells in the tumor microenvironment. Like other members of the(a disintegrin and metalloproteinase) family, the integrin-binding metalloproteinase ADAM9 modulatesll and cell-matrix interactions as well as ectodomain shedding of cell surface receptors and ligands,y modifying intracellular and extracellular signaling. ADAM9 transcripts are alternatively spliced to ex-a transmembrane protein (ADAM9-L) and a secreted variant (ADAM9-S). In this study, we show that9-S promotes breast cancer cell migration in a manner requiring its metalloproteinase activity, whereas9-L suppresses cell migration independent of its metalloproteinase activity. Suppression of migrationAM9-L requires a functional disintegrin domain and integrin binding. Expression analysis revealedoth ADAM9 isoforms are expressed in breast cancer cell lines and tissues. Therefore, relative levels of
that b
membrane-tethered and secreted variants of ADAM9 are a key determinant in manifestation of aggressivemigratory phenotypes associated with breast cancer progression. Cancer Res; 70(20); OF1–12. ©2010 AACR.

growtthe ecmembding otumorpromoand msimilanase ddisintementthe ceof promain.rich dThe SrregionSrc animporADAMcer prmal ce

duction

of the hallmarks of advanced breast cancer is the abil-he tumor cells to lose their epithelial phenotype, disen-om neighboring cells, degrade the basement membrane,vade surrounding tissues, ultimatelymetastasizing to dis-gans. The interactions between the tumor and its stromalnvironment are a key determinant in breast cancer pro-on from carcinoma in situ to advanced invasive andtatic carcinoma. Tumor cells proteolyze the basementrane and stroma to invade the vasculature, and also alteral signals to stimulate angiogenesis and alternately re-nd tether to the extracellular matrix (ECM) to allow fort migration. Tumor cells also release growth factors andkines that alter the stroma, inducing inflammation,enesis, and mechanisms of tissue repair (1, 2).teases play a major signaling role at the tumor-stromalary. Enzymes comprising the matrix metalloproteinase) family proteolyze basement membrane and functionmigratory signaling mechanisms. Proteolysis by MMPss cryptic sites in ECM constituents such as laminin-5llagen IV that in turn promote tumor cell migration

is of the basement membrane also releaseswth factors such as insulin and fibroblast

with adomai(16).MovariaADA

cysteinnase),contaproteifactorpromo

n: Department of Pathology, Beth Israel Deaconessrvard Medical School, Boston, Massachusetts

uthor: Alex Toker, Department of Pathology, BethMedical Center, 330 Brookline Avenue, EC/CLS-02215. Phone: 617-735-2482; Fax: 617-735-2480;

mc.harvard.edu.

5472.CAN-09-4231

ssociation for Cancer Research.

ls.org

Research. on April 10, 201cancerres.aacrjournals.org aded from

h factor (5, 6). MMPs proteolytically cleave and releasetodomains of multiple signaling molecules from the cellrane in a process known asmembrane shedding (7). Shed-f substrates such as transforming growth factor-β (8) andnecrosis factor-α (9) activates signaling pathways thatte cell migration and survival. The ADAM (a disintegrinetalloproteinase) family of proteases has functionalrity to the MMPs in their zinc-binding metalloprotei-omains, and is also referred to as the metalloproteinase,grin, cysteine-rich (MDC) family. ADAMs control base-membrane proteolysis and shedding of proteins fromll membrane. Comprising 40 isoforms, the ADAM familyteases also contains an integrin-binding disintegrin do-ADAMs function in cell adhesion through their cysteine-omains that bind to syndecans (10) and fibronectin (11).c homology-3 (SH3)–binding domain in the cytoplasmics of some ADAMs mediates signaling by activation ofd Grb proteins (12). Recent studies point to a functionaltance for some ADAM family members in cancer.12 is expressed in carcinoma and promotes breast can-ogression by inducing the apoptosis of surrounding stro-lls (13). Consequently, ADAM12 protein levels correlatedvanced breast cancer (14, 15). In contrast, the disintegrinn of ADAM15 inhibits angiogenesis and tumor growthoreover, ADAM15 expression decreases integrin-mediatedn cancer cell adhesion and motility (17).M9, also known asMDC9 (metalloproteinase/disintegrin/e-rich protein 9), MCMP (myeloma cell metalloprotei-and meltrin-γ, comprises a subgroup of ADAMs that alsoins ADAM12 and ADAM15 (18). The ADAM9 metallo-nase activity cleaves heparin-binding epidermal growth
(EGF; ref. 19), which binds activates EGF receptor tote tumor growth and angiogenesis (20). ADAM9 also
OF1

9. © 2010 American Association for Cancer

http://cancerres.aacrjournals.org/

cleave(22, 23integr(26), aRec

regulalines ((8p11-and AluminThe

boundin thesecretnot prthe traproteisecretin Matcells toincreaincrearole ofand inHer

to detlines(RNAiADAMulatioADAMthe firimplicbreast

Mate

AntibAnt

purchfromsearchproduclonalCATGwith L(PolyspcDNAbeen dEA.myfrom taccording prCATGCATGCCAT5′-CCmAD

GAATCAGGand mthe cyinto pprimeGCTTGGTTand thmADApcDNAlete ththe didirect5′-GTG5′-GC3′ KpnGAAGCACTdiscarKpn1the MAldricof showereC C TCTGGGAAG3′ (anCAAA3′ (seTATTSecondvpr aviral p

Cell lAll c

Collechuma(33). BRPMIMDA-were mvine seted wiin McCbovineified w100 ngpenicimethoscriptreceip

Cell l

Fry and Toker

CanceOF2


s Delta-like 1, a Notch ligand (21), as well as ADAM10). The ADAM9 disintegrin domain is a ligand for specificin heterodimers, including multiple β1 (24, 25), α6β4

nd αvβ5 integrins (27, 28).ent studies have revealed that ADAM9 message is up-ted in breast tumors as well as in breast cancer cell29, 30). The chromosomal region of the ADAM9 gene12) is amplified in several breast cancers and cell lines,DAM9 is overexpressed in cell lines encompassing theal, basal A, and basal B gene expression clusters (31).ADAM9 transcript is alternatively spliced intomembrane-and secreted isoforms. Alternative splicing of exon 12ADAM9 message leads to a shorter ADAM9-S (ADAM9-ed) transcript that contains eight unique amino acidsesent in ADAM9-L (ADAM9-long). ADAM9-S also lacksnsmembrane and cytoplasmic domains and is a secretedn (26). ADAM9 is implicated in cancer cell phenotypes, asion of ADAM9-S promotes invasion of breast cancer cellsrigel assays (26). Furthermore, adhesion of keratinocyterecombinant ADAM9 disintegrin/cysteine-rich domain

ses cell migration, and overexpression of ADAM9 alsoses pro-MMP9 expression (25). However, the functionalADAM9-L and ADAM9-S in modulating cell migrationvasion in breast cancer cells has not been evaluated.e, we use specific ADAM9-L and ADAM9-S antibodiesect expression of both variants in breast cancer celland breast cancer tissues. Using RNA interference) and gene replacement, we show that ADAM9-S and9-L do not function in a redundant manner in the reg-n of cell migration, whereby ADAM9-S promotes and9-L attenuates migration. These studies identify forst time an ADAM as a migration suppressor, and haveations for the functional roles of ADAM9 proteins incancer progression.

rials and Methods

odies and reagentsi-myc antibody purified from the 9B11 hybridoma wasased from Cell Signaling Technology. Anti–β-actin wasSigma-Aldrich. Anti–ADAM9-L was from BioMol Re-Laboratories, Inc. Anti–ADAM9-S was designed and

ced in collaboration with Pro-Sci, Inc. as a rabbit poly-antibody based on the immunization peptide sequenceLSLKFHAPF. Transient transfections were performedipofectamine 2000 (Invitrogen) or polyethyleimineciences, Inc). pcDNA3-mA9L, pcDNA3-mA9L.EA,3-mADAM9L.myc, and pcDNA4-hADAM9S.myc haveescribed previously (12, 26). Point mutants mADAM9L.c, ADAM9S.EA.myc, and ADAM9L.TCE.mycwere derivedhese vectors via Stratagene site-directed mutagenesising to the manufacturer's instructions and the follow-imers: ADAM9S.EA, 5′-CCAAGATTATGACCCAATG-AGCAACAATGG-3 ′ and 5 ′ -CCATTGTTGCT ‐CATTGGGTCATAATCTTGG-3′; mADAM9L.EA, 5′-TGTTGCTCATGCATTGGGGCATAACCTTGG-3′ and
AAGGTTATGCCCCAATGCATGAGCAACAATGG-3′;AM9L.TCE.myc, 5′-GGACCCTGGAGAGCGTGT-
Cell(pH 7.

r Res; 70(20) October 15, 2010

Research. on April 10, 201cancerres.aacrjournals.org Downloaded from

GCGGC-3′ and 5′-GCCGCATTCACACGTCTCTC-GTCC-3′. Deletion mutations in mADAM9L.Δcyt.mycADAM9L.Δdis.myc were derived as follows: To deletetoplasmic domain, a Kpn1 restriction site was insertedcDNA3-mADAM9-L 5′ of the cytoplasmic region withrs 5′-GGCTAACTAGAGAACCCACTGGTACCTG‐ATCG-3′ and 5′-CGATAAGCCAGGTACCAGTG‐CTCTAGTTAGCC-3′. BamHI from the vector backbonee inserted Kpn1 restriction sites were used to isolate theM9L.Δcyt sequence, which was cloned into digested4A/TO/myc/his vector in frame with themyc tag. To de-e disintegrin domain,Kpn1 sites were inserted 5′ and 3′ ofsintegrin domain in pcDNA3-mADAM9L.myc using site-ed mutagenesis using the following sequences: 5′ Kpn1,GAGCAAAGAGCGGTACCATGAATTCAGGAGC-3′ andTCCTGAATTCATGGTACCGCTCTTTGCTCCAC-3′;1, 5′-GAAGGAGTGTGAGGGTACCCCATGCTGTGAAG--3′ and 5′-CTTCCTTCACAGCATGGGGTACCCTCA-CCTTC-3′. Kpn1 digestion was used to isolate andd the disintegrin sequence, and the vector backbonesites were ligated. For RNAi, ADAM9 sequences fromission small interfering RNA (siRNA) project (Sigma-h) were cloned into the pLKO.1 vector directing expressionrt hairpin RNA (shRNA; ref. 32). The following sequencesused: ADAM9shRNA.1, 5′-CCGGCCCAGAGAAGTT-A T A T A T C T CGAGA TA T A T AGGAAC T T C T ‐GTTTTTG-3′ (sense) and 5′-AATTCAAAAACCCAGA-TTCCTATATATCTCGAGATATATAGGAACTTCTCTGG-tisense); ADAM9shRNA.2, 5′-CCGGGCCAGAATAA-GCCTATTCTCGAGAATAGGCTTTGTTATTCTGGCTTT-nse) and 5′-AATTCAAAAAGCCAGAATAACAAAGCC-CTCGAGAATAGGCTTTGTTATTCTGGC-3′ (antisense).d-generation lentiviral packaging plasmids pCMV-dR8.2nd pCMV-VSVG were obtained from Addgene for lenti-ackaging.

inesell lines were obtained from the American Type Culturetion, with the exception of the estrogen-independentn breast cancer SUM-159-PT, which has been describedT549, HCC38, and ZR75-1 cells were maintained in1640 supplemented with 10% fetal bovine serum.MB-231, HEK293T, MDA-MB-468, and NIH3T3 cellsaintained in DMEM supplemented with 10% fetal bo-rum.CAMA-1 cells weremaintained inMEMsupplemen-th 10% fetal bovine serum. SKBR3 cells were maintainedoy's modified medium 5A supplemented with 10% fetalserum. MCF10A cells were maintained in DMEM mod-ith 5% equine serum, 20 ng/mL EGF, 10 μg/mL insulin,/mL cholera toxin, 500 ng/mL hydrocortisone, and 1%llin and streptomycin. Cell lines were verified by multipleds, including DNA bar coding, gene expression, and tran-ome analysis, and kept in culture for <6 months aftert.

ysis
s were lysed in NP40 lysis buffer [20 mmol/L Tris-HCl0), 10% glycerol, 1% NP40, 10 mmol/L EDTA, 150 mmol/L
Cancer Research



NaCl,suppleProteiAssay

ImmuEqu

volumtein Aand thelution6.8), 2pheno

ImmuCell

AssaySDS-Ptransf5% nowereantibomanuCell mwith pbodiessensiti

LentivpLK

HEK2and pafter tharvesat −80conflu4 hourrum apurom

TranspCS

transfADAMprotocTechning). Mafter t

TransPlas

porterinfectewith Lstructfor noafter ttein ex

MigraMig

bers (sinize0.1% bto thetowardepenlowervirally(FishemidsExperstainethe pltion etransfby coumigraing. StInStatwas hserumfectedin ser

IntegCell

descrinized10 mi

Resu

ADAMbreasWe

and Asues.polyclADAMdesignADAMCOOHADAMEqualantibobody sis detevarian114 kDform aby myshowncross-not de

ADAM9 Isoforms in Breast Cancer Cell Migration

www.a


20 mmol/L NaF, 5 mmol/L sodium pyrophosphate]mented with protease inhibitor cocktail (Sigma-Aldrich).n concentration was calculated using Bio-Rad Proteinsolution.

noprecipitational amounts of protein were incubated at 4°C in a 1-mLe with 2 μg antibody for 4 hours before addition of pro-–Sepharose beads. Beads were incubated for 2 hoursen washed with 1% NP40 in PBS three times beforewith 2× Laemmli lysis buffer [125 mmol/L Tris (pH

3% glycerol, 4% SDS, 10% β-mercaptoethanol, brom-l blue] and immunoblotting.

noblottinglysates were equilibrated using the Bio-Rad Proteinsolution, and protein (50–100 μg) was resolved byAGE, transferred to nitrocellulose (Hoefer semidryer system, 160 mA, 80–120 minutes), and blocked innfat milk in 1% TBST (1% Tween in TBS). Membranesincubated with the appropriate antibodies (primarydies used at the concentration recommended by thefacturer, ADAM9-S used at 2 μg/mL) for 14 hours.embranes were washed in 1% TBST and incubatederoxidase-conjugated anti-mouse or anti-rabbit anti-(Jackson ImmunoResearch) and exposed using high-vity enhanced chemiluminescence (Millipore).

iral infectionO.1 plasmid constructs were cotransfected into93T cells with packaging vectors pCMV-dR8.2 dvprCMV-VSVG using polyethyleimine. Forty-eight hoursransfection, lentiviral particles in supernatant wereted, passed through a 0.45-μm filter, and either stored°C or used to infect target cells. Target cells at ∼80%ency were infected with virus alone for a period ofs, at which point they were supplemented with full se-nd allowed to recover for 24 hours before selection inycin in full culture medium for 18 hours (2–5 μg/mL).

fection of siRNA into BT549 cells2-(n)-β-galactosidase reporter plasmid (1 μg) wasected alone or with 200 pmol/L of nontargeting or9 siRNA pools according to the Lipofectamine 2000ol [5 μL Lipofectamine 2000, 500 μL Opti-MEM (Lifeologies)] into 5 × 105 BT549 cells per 6-cm dish (Corn-edium was replaced with full-serum medium 4 hoursransfection.

fection of plasmid DNAmids of interest and the pCS2-(n)-β-galactosidase re-plasmid were cotransfected at a ratio of 6.5:1 into non-d or lentivirally infected cells 24 hours after infectionipofectamine 2000 according to the manufacturers' in-ions. Medium was replaced with full culture mediumninfected cells or puromycin selection medium 4 hours
ransfection. Cells were assayed for migration and pro-pression 24 hours after transfection.
againsand m

acrjournals.org


tion assaysration assays were carried out using Transwell cham-Corning) with 8-μm pore membranes. Cells were tryp-d and resuspended in serum-free medium containingovine serum albumin (BSA). Cells (1 × 105) were addedtop of each Transwell chamber and allowed to migrated cell-conditioned medium for 4 to 16 hours at 37°Cding on the cell type. Cells that had migrated to thesurface of the membrane were fixed and stained. Lenti-infected cells were stained with Hema-3 Protocol Stainr Scientific). Cells cotransfected with expression plas-and β-galactosidase were stained with X-galactosidase.iments were conducted in triplicate, and equal fields ofd cells were counted in each well and normalized toating efficiency or the transfection efficiency. Transfec-fficiency was calculated by counting the number ofected cells per 100,000. Plating efficiency was calculatednting two wells of 100,000 cells plated in parallel to thetion assays to control for minor discrepancies in plat-atistical analysis was performed using ANOVA and thestatistical analysis software. Cell-conditioned mediumarvested from NIH3T3 fibroblast cells grown in fullfor 48 hours to confluency, or HEK293T cells trans-with hADAM9-S expression plasmids, and incubated

um-free medium for 48 hours.

rin-blocking assayss were infected and transfected with plasmid DNA asbed above. Before migration assay, cells were trypsi-and incubated with 5 μg/mL of GOH3 antibody fornutes at room temperature.

lts

9-S is expressed in breast cancer cell lines andt tumorsfirst determined the expression pattern of ADAM9-SDAM9-L in breast cancer cells and breast cancer tis-To detect specific endogenous isoforms of ADAM9, aonal antibody directed toward the COOH terminus of9-L and an ADAM9-S–specific polyclonal antibodyed against the unique COOH terminus of human9-S were used (Fig. 1A). To confirm isoform specificity,-terminal myc-tagged murine ADAM9-L and human9-S were transiently expressed in HEK-293T cells.amounts of protein were immunoblotted with mycdy to validate equivalent expression and ADAM9 anti-pecificity. The results show that, as expected, ADAM9-Lcted by myc and ADAM9-L antibodies as a pro–ADAM9-Lt that comprises the unprocessed form migrating ata and the fully processed mature membrane-boundt 84 kDa (Fig. 1B). Expression of ADAM9-S is detectedc and ADAM9-S antibodies at 67 kDa, as previously(26). Importantly, the ADAM9-L antibody does not

react with ADAM9-S, and the ADAM9-S antibody doestect ADAM9-L (Fig. 1B). Therefore, specific antibodies
t ADAM9-L and ADAM9-S are specific to the secretedembrane-bound variants of ADAM9. It is important to
Cancer Res; 70(20) October 15, 2010 OF3



note tPAGEence oTo

breastHCC3CAMA80% tominedin allcells pHCC3of explight edetectas welteolytexpresand Mexpres

breastobservcell tyTo

cer tisinvasisamplDeacoADAMwell a

Overecell mPre

creaseogenoADAMproteo

FiguresecreteADAM9Equivalfor immof ADAantibod t least

Fry and Toker

CanceOF4


hat ADAM9-L and ADAM9-S migrate slower on SDS-than their predicted molecular mass due to the pres-f cysteine-rich regions as well as glycosylation.evaluate the expression of ADAM9-S and ADAM9-L incancer cell lines, lysates were collected from BT549,8, MDA-MB-231, SUM159PT, MCF10A, MDA-MB-468,-1, ZR75-1, MCF-7, and SKBR3 cells at approximately90% confluence, and ADAM9 expression was deter-by immunoblotting. ADAM9-L expression is detectedcell lines tested (Fig. 1C, dark exposure). Importantly,reviously characterized to overexpress ADAM9 (BT549,8, and CAMA-1; ref. 31) reveal a substantially higher levelression of ADAM9-L compared with other lines (Fig. 1C,xposure). In addition, the variance in ADAM9-L bandsed by immunoblotting reflects differences in processingl as posttranslational modifications that include pro-ic cleavage and glycosylation. ADAM9-S endogenoussion is detected in BT549, HCC38, SUM159PT, MCF10A,

y. BT549 cell lysate was used as control. All results are representative of a

DA-MB-468 cells (Fig. 1C). Thus, endogenous ADAM9-Lsion is detected in cells of both the luminal and basal

Plasmmutan

r Res; 70(20) October 15, 2010


cancer subtypes, whereas ADAM9-S is predominantlyed in basal-type cells. Notably, CAMA-1, a noninvasivepe, expresses ADAM9-L, but not ADAM9-S.evaluate expression of ADAM9 isoforms in breast can-sues, ADAM9-S levels were evaluated in stage II and IIIve ductal carcinoma tumor and normal breast lysatees kindly provided by Dr. Gerburg Wulf (Beth Israelness Medical Center; ref. 34). The results show that9-S is expressed in breast cancers of both grades ass in normal tissue (Fig. 1D).

xpression of ADAM9-S promotes breast cancerigrationvious studies have shown that cancer cells exhibit in-d translocation through Matrigel in the presence of ex-us recombinant ADAM9-S (26). To evaluate the role of9 isoforms in cell migration independent of matrixlysis, motility was assayed using uncoated Transwells.

three independent experiments.

1. Expression of ADAM9 isoforms in breast cancer cell lines and tissues. A, a schematic representation of the membrane-bound (ADAM9-L) andd (ADAM9-S) isoforms of ADAM9. B, HEK293T cells transiently transfected with a vector control (pcDNA4), murine ADAM9L.myc, and humanS.myc were lysed and proteins were resolved on an 8% SDS-PAGE gel. Equivalent lanes were blotted with anti-myc to validate protein expression.ent lanes were blotted with ADAM9-L– and ADAM9-S–specific antibodies. C, a panel of cell lines representing breast cancer subtypes was usedunoblotting with ADAM9-L COOH-terminal–specific antibody, ADAM9-S COOH-terminal–specific antibody, and β-actin. “#” denotes overexpressionM9 and amplified 8p11-12; “*” denotes amplified 8p11.12. D, SDS-extracted lysates from frozen tumor blocks were immunoblotted with anti–ADAM9-S

id constructs directing the expression of wild-type andt murine ADAM9-L and human ADAM9-S (Fig. 2A)

Cancer Research



wereexpreBT549proteisubstinot intransfADAMnor dideficieporte

activitrecapiTo

gratiopressmediuwere aand BUnder

FigurebreastA, schemurinein migrB, migroverexpADAM9for 4 hof transconfirmCOOH-or anti-by arroADAM9endoge*, P < 0of BT54exogenmigratein whiccontainfrom HpcDNAADAM9ADAM9anti-my*, P < 0mean oof at leexperim


www.a


transfected into BT549 breast cancer cells. Over-ssion of wild-type ADAM9-S significantly increasesmigration. In contrast, overexpression of a metallo-nase-deficient ADAM9-S.EA mutant with a Glu to Alatution in the metalloproteinase active site (35) doescrease migration and is comparable with control-ected cells (Fig. 2B). Similarly, overexpression of the9-L variant did not promote BT549 cell migration,d overexpression of the equivalent metalloproteinase-
nt ADAM9L.EA mutant. Therefore, as previously re-d, ADAM9-S promotes cell migration in a protease
hour inincrea

ents; bars, SD.

acrjournals.org


y–dependent manner. However, this phenotype is nottulated by ADAM9-L.confirm that extracellular ADAM9-S stimulates cell mi-n, ADAM9-S, ADAM9S.EA, or vector control was ex-ed in HEK293T cells and serum-free conditionedm was collected after transfection. These supernatantsdded to the bottom chamber of uncoated Transwells,T549 breast cancer cells were added to the top chamber.these conditions, BT549 cells require an extended 16-
cubation period tomigrate. However, BT549 cells exhibitsed migration when exposed to ADAM9-S–containing
2. ADAM9-S promotescancer cell migration.matic of human andADAM9 constructs usedation experiments.ation of BT549 cellsressing ADAM9-L and-S. Cells were migratedin Transwells. Expressionfected proteins wased with anti–ADAM9-Lterminal–specific antibodymyc. ADAM9-L is denotedws. Overexpressed murine-L migrates slower thannous human ADAM9-L..001, ANOVA. C, migration9 cells stimulated withous ADAM9-S. Cells wered for 16 h in Transwellsh the bottom chambered medium collectedEK293T cells expressingcontrol, ADAM9-S, orS.EA. Expression of-S was confirmed byc immunoblotting..05, ANOVA. Columns,f triplicate measurementsast three independent




conditor ADnous Ametall

ADAMTo d

cell musingquencsientlyhigh le

ingly,measutrol oquantphenotulateTo evthe siwere gregion

urereass walacjunconuonuratencinunOH-β-aAD

AM9unADuntesT5troltiviruNAAD4 h.OVAfirmressi-myc.

Fry and Toker

CanceOF6


ioned medium in comparison with control-transfectedAM9S.EA-transfected medium, indicating that exoge-DAM9-S stimulates haptotactic migration through itsoproteinase activity (Fig. 2C).

9 silencing increases breast cancer cell migrationetermine the functional role of endogenous ADAM9 inigration, we performed loss-of-function experimentsa siRNA oligonucleotide pool targeting five distinct se-es spanning both ADAM9 isoforms. We chose to tran-
transfect these siRNAs into BT549 cells, which expressvels of ADAM9 and are also highly migratory. Surpris-
rich dand s

r Res; 70(20) October 15, 2010


ADAM9 siRNA increases migration in BT549 cells asred by Transwell migration assays compared with con-r nontarget siRNA (Fig. 3A). This is concomitant withitative silencing of ADAM9-L protein expression. Thetype of increased migration by ADAM9 siRNA recapi-s that observed on expression of ADAM9-S (Fig. 2A).aluate that this is not due to an off-target effect ofRNA pool, two distinct lentiviral shRNA constructsenerated to target both isoforms in the 3′ untranslatedof ADAM9 mRNA (ADAM9shRNA.1) and the cysteine-

Figinccellβ-gconoligoligmigSileimmCOandwithADimmandimmlysaC, BconlenpcDandforANconexpant

omain (ADAM9shRNA.2). Aftelection, robust silencing of AD

9. © 2010 American Associati

3. ADAM9 silencinges cell migration. A, BT549ere transfected withtosidase alone, or intionwith nontargeting siRNAcleotide, or ADAM9 siRNAcleotide pool. Cells wered for 16 h. *,P<0.01, ANOVA.g was confirmed byoblotting using ADAM9-Lterminal–specific antibodyctin. B, BT549 cells infectedAM9shRNA.1 andshRNA.2 were lysed andoblotted against ADAM9-LAM9-S. ADAM9-S wasoprecipitated from equal cellbefore immunoblotting.49 cells were infected withor ADAM9shRNA.1ses before transfection ofcontrol, ADAM9-L, ADAM9-S,AM9S.EA. Cells weremigrated*, P < 0.05; **, P < 0.001,. ADAM9-L immunoblots silencing. ADAM9-Sion is detected with

er lentiviral infectionAM9-L and ADAM9-S

Cancer Research

on for Cancer


proteias revductioin BT5migraalso reing CAinal aexpresADAMTo

for thestitutiADAM

shRNAdictedConsiwas recomparobustassaysmigraincreaspecifeffectssuppr

FigureCAMA-infectedADAM9lentivirumigrate4 h (BTANOVAimmunoterminacell couColumnmeasurindependent experiments; bars, SD.


www.a


n is observed in cells transduced with both sequencesealed by immunoblotting (Fig. 3B). Importantly, trans-n of both shRNA sequences also results in an increase49 cell migration (Fig. 3D). Moreover, increased celltion resulting from transduction of ADAM9 shRNA isproduced in distinct breast cancer cell lines, includ-MA-1 and ZR75-1 (Fig. 3D). Therefore, in different lum-nd basal breast cancer type cells, silencing ADAM9sion results in enhanced cell migration observed with9shRNA.1 and ADAM9shRNA.2 sequences.further confirm that ADAM9 silencing is responsibleincrease in migration observed with shRNA, a recon-

on experiment was performed by introducing murine9-L and human ADAM9-S in cells harboring ADAM9

Weenhan

acrjournals.org


. The murine ADAM9-L (mADAM9-L) would be pre-to be refractory to silencing using human ADAM9.

stent with this, expression of mADAM9-L proteinsistant to silencing in cells expressing ADAM9 shRNAred with cells with ADAM9 shRNA alone that show areduction of ADAM9-L protein (Fig. 3C). In Transwell, expression of mADAM9-L rescues the increase intion observed with shRNA alone. Therefore, these in migration observed on silencing of ADAM9 isic to ADAM9-L expression and is not due to off-target. The implication is that endogenous ADAM9-L is aessor of migration of breast cancer cells.

3. Continued. D, BT549,1, and ZR75-1 cells werewith control or

shRNA.1 or ADAM9shRNA.2ses. Selected cells wered in uncoated Transwells for549) or 16 h. *, P < 0.03,. Silencing is confirmed byblot with ADAM9-L COOH-l antibody. In all experiments,nts were normalized.s, mean of triplicateements of at least three

also evaluated the contribution of ADAM9-S in theced migration phenotype observed with ADAM9




shRNAbut notion inof expducedADAM(ADAstead,the mand Aitulatemutanthe mover, tof micance

Functof celThe

migranext ddomaitions ointegrdomaito genhumanallelesADAMtype Agratioin cellsmutandispenwas nomain dand domic doTo

and inmotifmainproteimutancontaiform cwe assregulaantiboand reby treabindinis alsotegrinmodespLKOwith s

(compsignifian effeshRNAlane 6by altintegrpathwrecep

Discu

Thephenospecifunderbeen stify noADAMwith pmotener. SADAMThe

ter thwith ireduceno oveprostaW10, o(35). Aparedical anlines sin brespecifof procytochtion oBre

basedpressphenomore icancergroupsion p(31). CADAMIn ourof eacin thethe finOur

gratioblasts

Fry and Toker

CanceOF8


. As observed in Fig. 2A, expression of wild-type,t protease-deficient, ADAM9-S increases cell migra-control BT549 cells (Fig. 3C). In contrast to the effectression of mADAM9-L, the increase in migration in-by ADAM9 silencing is not rescued by expression of9-S that is refractory to silencing, as the shRNA usedM9shRNA.1) targets the 3′ untranslated region. In-expression of ADAM9-S leads to a further increase inigratory phenotype that is additive to ADAM9 shRNADAM9-S. As expected, this additive effect is not recap-d by expression of the ADAM9-S protease-inactive EAt (Fig. 3C). Therefore, ADAM9-S does not phenocopyigration suppression function of ADAM9-L, and more-he implication from these results is that suppressiongration is the dominant role of ADAM9 in breastr cells.

ional role of ADAM9-L domains in suppressionl migrationdata thus far show that ADAM9-S is an enhancer of celltion, whereas ADAM9-L is a suppressor of migration.Weetermined the contribution of the ADAM9-L functionalns in the suppression of migration. We generated muta-r deletions in the metalloproteinase (ADAM9L.EA), dis-in (ADAM9L.Δdis), and cytoplasmic (ADAM9L.Δcyt)ns (Fig. 4A). Again the murine ADAM9-L allele was usederate these mutations, as it is refractory to silencing inbreast cancer cells. Wild-type and mutant mADAM9-Lwere introduced into BT549 cells after silencing with9 shRNA. As already observed, expression of the wild-DAM9-L allele rescued the phenotype of increased mi-n on ADAM9 silencing (Fig. 4B). This was also observedexpressing the metalloproteinase-deficient ADAM9L.EAt, indicating that the proteolytic activity of ADAM9-L issable for the suppression of migration. In contrast, thererescue of increased migration with the cytoplasmic do-eletionmutant (ADAM9L.Δcyt), showing that localizationwnstream signaling mediated by the ADAM9-L cytoplas-main is required for suppression of cellmigration (Fig. 4C).address the role of the ADAM9-L disintegrin domaintegrin-binding activity, point mutations in the ECDas well as whole deletions in the integrin-binding do-were generated. However, expression of these mutantns followed by immunoblotting revealed that bothts are not properly processed from the prodomain-ning form (114 kDa) to the active and mature 84-kDaompared with wild-type ADAM9-L (Fig. 5A). Therefore,essed the contribution of the disintegrin domain in thetion of cell migration using integrin function-blockingdies. BT549 cells were transduced with ADAM9 shRNAconstituted with wild-type murine ADAM9-L, followedtment with GOH3, an antibody that blocks α6 integring to its ligand laminin (Fig. 5B), or control IgG. GOH3routinely used as direct binding ligand to activate in-clustering (36). Interestingly, treatment with GOH3tly increased the migration of BT549 cells with control
. As predicted by the data thus far, silencing ADAM9hRNA increased cell migration compared with control
of chrincrea

r Res; 70(20) October 15, 2010


are lane 5 with lane 1). However, treatment with GOH3cantly increasedmigration (compare lane 6 with lane 5),ct that was quantitatively rescued by expression of the-resistant mADAM9-L allele (compare lane 8 with

). This shows that ADAM9-L suppresses cell migrationering integrin-mediated signals, either at the level ofin binding or by interacting with a parallel signalingay that engages in cross talk downstream of integrintors.

ssion

ADAM family of metalloproteinases controls multipletypes associated with tumor progression, although theic role of individual isoforms in tumorigenesis is poorlystood. Several ADAM isoforms, including ADAM9, havehown to promote cell migration. In this study, we iden-nredundant roles for the secreted and membrane-bound9 variants in breast cancer cell migration. Consistentrevious observations, we show that ADAM9-S can pro-breast cancer cell migration in a cell-autonomous man-urprisingly, this phenotype is not recapitulated by9-L, which functions as a migration suppressor.ADAM9 gene (8p11.23) is located in the 8p11-12 clus-at is frequently amplified or deleted in breast cancer,nstances of abnormal copy number that correlate withd patient survival (37). Although ADAM9-null mice havert phenotypes (38), inmousemodels of breast, colon, andte carcinoma such as MMTV-PyMT, APC/Min/+, andverexpression of ADAM9 in neoplastic regions is detectedDAM9−/−W10 mice have well-differentiated tumors com-with tumors in ADAM9+/+W10 (35). Immunohistochem-alysis of ADAM9 expression in breast tumors and cellupports our finding that secreted ADAM9-S is expressedast cancer tissues (29, 30). In our studies, the ADAM9-ic antibodies were unable to provide a clear detectiontein expression by immunohistochemistry or immuno-emistry, but regardless, this is the first clear identifica-f ADAM9-S expression in tumors and cell lines.ast cancer cell lines are separated into different clusterson gene expression profiles (39). Luminal cell lines ex-genes associated with a more differentiated, noninvasivetype, whereas basal cell lines are less differentiated andnvasive. In a survey of gene expression profiles of 51 breastcell lines, the basal gene cluster was separated into two

s, basal A and B, of which B has a stem cell–like expres-rofile closest to the clinical “triple-negative” tumor profileell lines with ADAM9 amplification and with or without9 mRNA overexpression were identified in this analysis.present study, we find that ADAM9-L is expressed in cellsh gene cluster, whereas ADAM9-S is expressed primarilymore aggressive basal genotype (Fig. 1C), correlating withding that ADAM9-S promotes cell migration.previous studies had identified ADAM9-S as a promi-n factor that is secreted by activated liver myofibro-in the wound-healing response (26). Tumors are sites
onic wounding, and cancer cells adapt to the wound bysing cell migration and angiogenesis. The success of
Cancer Research



tamoxstrom(40). Sing invprostaportansuggescrine f

ADAMactivipresso(Fig. 4cell sulocaliz

Figurecell migindepenrepreseADAM9domainthe rolemigratioinfectedlentiviruwith pcwild-typselectioin unco*, P < 0were imand ADmurine(bottomcells weA9shRNtransfecmADAMmADAMCells wTransw*** P <expressmyc, Awith anis contimmunexperimnormalefficientriplicaleast thexperiments; bars, SD.


www.a


ifen treatment in recurrent breast cancer with a highal content is correlated with elevated levels of ADAM9imilarly, ADAM9 is upregulated in the stroma surround-asive melanoma (41) and colon cancer (26), as well aste cancer cells (42, 43). These studies emphasize the im-ce of ADAM9 function at the tumor-stromal boundary,
ting that ADAM9 may function downstream of para-actors involved in wound healing. We find that whereas
with tgratio

acrjournals.org


9-S promotes cell migration via its metalloproteinasety (Fig. 2A), ADAM9-L functions as a migration sup-r in a manner that is independent of proteolytic activityB). ADAM9-S may proteolyze specific substrates on therface or ECM that are not targeted by ADAM9-L, as itsation is restricted by membrane tethering. Consistent

4. ADAM9-L suppressesration in ametalloproteinase-dent manner. A, schematicntation of murine wild-typeisoforms and functionalmutations used to assessof each domain in celln. B, BT549 cells werewith control or A9shRNA.1ses before transfectionDNA control, mADAM9-Le, or mADAM9L.EA aftern. Cells were migratedated Transwells for 4 h..01, ANOVA. Cell lysatesmunoblotted against actinAM9-L to detect both(top arrows) and humanarrows) variants. C, BT549re infected with control orA.1 lentiviruses beforetion with pcDNA control,9-L wild-type, or9L.Δcyt after selection.ere migrated in uncoatedells for 4 h. *, P < 0.05;0.001, ANOVA. ADAM9-Lion is detected with anti-DAM9 silencing is detectedti–ADAM9-L, and actinrol. ns, nonspecificoreactive protein. In allents, cell counts were

ized to transfectioncy. Columns, mean ofte measurements of atree independent

his, cells expressing ADAM9-S have a higher rate of mi-n than cells exposed to a more diffuse exogenously




addedmightthan afunctithat thisoforneighbfor ceADAMepithewe findispendisintebraneThe

tion into reccreaseadheribinantcollagin extr

vationbreasthancein thecontrolipid rsurfacwith acells, amechaThe

enhanas to hThis iscells omigratis the psive inently uis the

uredingratie trDAMDAMDAMunOH-s infDAsfecDAMls wA inibodoreatesAM9onfumnasurerime ca< 0.05; ***, P < 0.001.

Fry and Toker

CanceOF10


ADAM9-S (Fig. 2A and B), indicating that ADAM9-Sfunction in a cancer cell–autonomous manner rathercting on the tumor stroma. Regardless, the opposingons of ADAM9-L and ADAM9-S in cell migration suggeste two splice variants control tissue homeostasis, withm-specific functions in response to signals from bothoring tumor and stromal cells. A recent shRNA screenll migration enhancers and suppressors identified9 as amigration suppressor in nontumorigenic MCF10Alial cells, consistent with our findings (44). Surprisingly,d that the ADAM9-L metalloproteinase domain issable for this phenotype, whereas the cytoplasmic andgrin domains are required, indicative of a role for mem-localization and cytoplasmic signaling.disintegrin domain of ADAM9 influences cell migra-a cell type–dependent manner. Keratinocytes adheringombinant ADAM9-L disintegrin domain exhibit in-d migration (25), a result also observed with fibroblastsng to recombinant ectodomain (45). In contrast, recom-disintegrin domain inhibits the adhesion of platelets to
en I (46) and interacts with multiple integrins, resultingacellular signal-regulated kinase phosphorylation, acti-
Inmemb

r Res; 70(20) October 15, 2010


of p38MAPK, cPLA2, and MMP-9 synthesis (25, 27). Incarcinoma cells, antibody binding to α6 integrin en-s migration, a phenotype that is significantly enhancedabsence of ADAM9 expression (Fig. 5B). ADAM9-L mayl the localization of integrins or syndecans to specificafts (47), or alternatively modulate endocytosis of celle receptors. ADAM9 has also been shown to interactnd promote the recycling of E-cadherin in colon cancernd thismay represent ametalloproteinase-independentnism for ADAM9-L (48).opposing roles for ADAM9-S and ADAM9-L as migrationcers and suppressors in breast cancer cells raise thequestionow the expression profile of each splice variant is regulated.predicted to have a major effect in tumor progression, asverexpressing ADAM9-L would presumably have a reducedion phenotype. Conversely, any tumors in which ADAM9-Sredominant isoformwould be predicted to have aggres-vasive and metastatic characteristics. Although pres-nknown, it is highly likely that one important mechanismcontrol of alternative splicing of the ADAM9 message.

FigbinmigwermAmAmAimmCOcellor AtranmACelBSantbefLysADto cColmeexpwer*, P

summary, our results show that the secreted andrane-tethered isoforms of ADAM9 have opposing

Cancer Research


5. The ADAM9-L integrin-activity modulates cell

on. A, HEK293T cellsansfected with wild-type9-L, point mutant9L.TCE.myc, and9.Δdis.myc. Lysates were

oblotted with ADAM9-Lterminal antibody. B, BT549ected with control pLKOM9shRNA.1 virus wereted with pcDNA control or9-L wild-type constructs.

ere resuspended in 0.01%PBS with 5 μg/mL GOH3y and incubated for 10 mina 4-h migration assay.were immunoblotted with-L COOH-terminal antibodyirm silencing and expression.s, mean of triplicateements of two independentents; bars, SD. P valueslculated using ANOVA.


functiboth imodelstimulbraneand siis thattetherthe deinterv

Disclo

No p

Ackn

We tCourtnmurineintegrinSettlemdiscussi

Grant

Refe1. De

J P2. Zig

the3. Gia

Qucle

4. Xuticgro

5. Fomede

6. Whtionbaand

7. Muen

8. MizandIL-200

9. Yuprogio

10. Ibahutrigsp

11. Gaandma

12. Semu7–3

13. Kvbre200

14. PonoEp

15. Rocletus

16. Troanggrin


www.a


ons in breast cancer cell migration. The expression ofsoforms in breast cancer tumors and stroma suggests ain which the secreted ADAM9-S acts as a migratoryant, whereas ADAM9-L tethered to the plasma mem-binds cell surface proteins and mediates localizationgnaling. One important implication from these findingsisoform-specific functions of secreted and membrane-ed proteases exist, which has obvious consequences for
velopment of small-molecule inhibitors for therapeutic
NIHResearch

Theof pageaccorda

ention.

sure of Potential Conflicts of Interest

isclosed.Rece

OnlineF

chon-Joseph V, Martel-Renoir D, Mir LM, et al. Evidence of anti-iogenic and antimetastatic activities of the recombinant disinte-domain of metargidin. Cancer Res 2004;64:2062–9.

17. BevitrRG

18. Kletia

19. IzuMDtodlike

20. Onof

21. DyProCh

22. Pame10

23. ToproproJ B

24. MadoJ 2

25. Zigcy20

26. MaADac

7. Kamemydir

8. Malat20

9. LeIncand

0. O'San

1. NelineCe

32. Mohuscr

acrjournals.org


owledgments

hank Dr. Gerburg Wulf for kindly providing tumor tissue samples; Saraeidge for providing ADAM9-L cDNA; Dr. Carl Blobel for providingADAM9-L cDNA; Dr. Isaac Rabinovitz for assistance with the-binding studies; Drs. Marsha Moses, Bruce Zetter, and Jeffreyan for guidance; and all members of the Toker laboratory forons and advice.

Support

grant CA096710 (A. Toker) and Department of Defense Breast CancerProgram Pre-Doctoral Traineeship Award W81XWH 06-1-046 (J.L. Fry).

costs of publication of this article were defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisement innce with 18 U.S.C. Section 1734 solely to indicate this fact.

ived 11/25/2009; revised 07/21/2010; accepted 08/19/2010; publishedirst 08/24/2010.
otential conflicts of interest were d
rencesWever O, Mareel M. Role of tissue stroma in cancer cell invasion.athol 2003;200:429–47.rino P, Loffek S, Mauch C. Tumor-stroma interactions: their role incontrol of tumor cell invasion. Biochimie 2005;87:321–8.nnelli G, Falk-Marzillier J, Schiraldi O, Stetler-Stevenson WG,aranta V. Induction of cell migration by matrix metalloprotease-2avage of laminin-5. Science 1997;277:225–8.J, Rodriguez D, Petitclerc E, et al. Proteolytic exposure of a cryp-site within collagen type IV is required for angiogenesis and tumorwth in vivo. J Cell Biol 2001;154:1069–79.wlkes JL, Thrailkill KM, Serra DM, Suzuki K, Nagase H. Matrixtalloproteinases as insulin-like growth factor binding protein-grading proteinases. Prog Growth Factor Res 1995;6:255–63.itelock JM, Murdoch AD, Iozzo RV, Underwood PA. The degrada-of human endothelial cell-derived perlecan and release of bound

sic fibroblast growth factor by stromelysin, collagenase, plasmin,heparanases. J Biol Chem 1996;271:10079–86.

rphy G. The ADAMs: signalling scissors in the tumour micro-vironment. Nat Rev Cancer 2008;8:929–41.gerd JP, Spieker MR, Doerschuk CM. Early response cytokinesinnate immunity: essential roles for TNF receptor 1 and type I

1 receptor during Escherichia coli pneumonia in mice. J Immunol1;166:4042–8.Q, Stamenkovic I. Cell surface-localized matrix metalloproteinase-9teolytically activates TGF-β and promotes tumor invasion and an-genesis. Genes Dev 2000;14:163–76.K, Albrechtsen R, Gilpin B, et al. The cysteine-rich domain of

man ADAM 12 supports cell adhesion through syndecans andgers signaling events that lead to β1 integrin-dependent cellreading. J Cell Biol 2000;149:1143–56.ultier A, Cousin H, Darribere T, Alfandari D. ADAM13 disintegrincysteine-rich domains bind to the second heparin-binding do-

in of fibronectin. J Biol Chem 2002;277:23336–44.als DF, Courtneidge SA. The ADAMs family of metalloproteases:ltidomain proteins with multiple functions. Genes Dev 2003;17:0.eiborg M, Frohlich C, Albrechtsen R, et al. A role for ADAM12 inast tumor progression and stromal cell apoptosis. Cancer Res5;65:4754–61.ries SE, Zurakowski D, Roy R, et al. Urinary metalloproteinases:ninvasive biomarkers for breast cancer risk assessment. Canceridemiol Biomarkers Prev 2008;17:1034–42.y R, Wewer UM, Zurakowski D, Pories SE, Moses MA. ADAM 12aves extracellular matrix proteins and correlates with cancer sta-and stage. J Biol Chem 2004;279:51323–30.

2

2

2

3

3

ck V, Herold H, Benge A, et al. ADAM15 decreases integrin αvβ3/onectin-mediated ovarian cancer cell adhesion and motility in anD-dependent fashion. Int J Biochem Cell Biol 2005;37:590–603.ino I, Ortiz RM, Huovila AP. ADAM15 gene structure and differen-l alternative exon use in human tissues. BMC Mol Biol 2007;8:90.mi Y, Hirata M, Hasuwa H, et al. A metalloprotease-disintegrin,C9/meltrin-γ/ADAM9 and PKCδ are involved in TPA-induced ec-omain shedding of membrane-anchored heparin-binding EGF-growth factor. EMBO J 1998;17:7260–72.

gusaha PP, Kwak JC, Zwible AJ, et al. HB-EGF is a potent inducertumor growth and angiogenesis. Cancer Res 2004;64:5283–90.czynska E, SunD, Yi H, Sehara-FujisawaA, Blobel CP, Zolkiewska A.teolytic processing of delta-like 1 by ADAM proteases. J Biolem 2007;282:436–44.rkin E, Harris B. A disintegrin and metalloproteinase (ADAM)-diated ectodomain shedding of ADAM10. J Neurochem 2009;8:1464–79.usseyn T, Thathiah A, Jorissen E, et al. ADAM10, the rate-limitingtease of regulated intramembrane proteolysis of Notch and otherteins, is processed by ADAMS-9, ADAMS-15, and the γ-secretase.iol Chem 2009;284:11738–47.himkar RM, Visaya O, Pollock AS, Lovett DH. The disintegrinmain of ADAM9: a ligand for multiple β1 renal integrins. Biochem005;385:461–8.rino P, Steiger J, Fox JW, et al. Role of ADAM-9 disintegrin-steine-rich domains in human keratinocyte migration. J Biol Chem07;282:30785–93.zzocca A, Coppari R, De Franco R, et al. A secreted form ofAM9 promotes carcinoma invasion through tumor-stromal inter-tions. Cancer Res 2005;65:4728–38.radag A, Zhou M, Croucher PI. ADAM-9 (MDC-9/meltrin-γ), amber of the a disintegrin and metalloproteinase family, regulateseloma-cell-induced interleukin-6 production in osteoblasts byect interaction with the α(v)β5 integrin. Blood 2006;107:3271–8.schler S, Wirl G, Spring H, et al. Tumor cell invasiveness corre-es with changes in integrin expression and localization. Oncogene05;24:2032–41.ndeckel U, Kohl J, Arndt M, Carl-McGrath S, Donat H, Rocken C.reased expression of ADAM familymembers in human breast cancerbreast cancer cell lines. J Cancer Res Clin Oncol 2005;131:41–8.hea C, McKie N, Buggy Y, et al. Expression of ADAM-9 mRNA

d protein in human breast cancer. Int J Cancer 2003;105:754–61.ve RM, Chin K, Fridlyand J, et al. A collection of breast cancer cells for the study of functionally distinct cancer subtypes. Cancerll 2006;10:515–27.
ffat J, Grueneberg DA, Yang X, et al. A lentiviral RNAi library forman and mouse genes applied to an arrayed viral high-contenteen. Cell 2006;124:1283–98.



33. FlaSUce

34. Wucascr34

35. Pefor

36. Leelam

37. Chab20

38. Wedisdu

39. PesioPro

40. SieADrescan

41. Zig

ulace

42. ShdiasigPro

43. SuADRe

44. SimulaNa

45. NaMuα(6mo

46. Cotumlag

47. Thsigpro

Fry and Toker

CanceOF12


nagan L, Van Weelden K, Ammerman C, Ethier SP, Welsh J.M-159PT cells: a novel estrogen independent human breast can-r model system. Breast Cancer Res Treat 1999;58:193–204.lf GM, Ryo A, Wulf GG, et al. Pin1 is overexpressed in breastncer and cooperates with Ras signaling in increasing the tran-iptional activity of c-Jun towards cyclin D1. EMBO J 2001;20:59–72.duto L, Reuter VE, Shaffer DR, Scher HI, Blobel CP. Critical functionADAM9 in mouse prostate cancer. Cancer Res 2005;65:9312–9.EC, Lotz MM, Steele GD, Jr., Mercurio AM. The integrin α6β4 is ainin receptor. J Cell Biol 1992;117:671–8.in K, DeVries S, Fridlyand J, et al. Genomic and transcriptionalerrations linked to breast cancer pathophysiologies. Cancer Cell06;10:529–41.skampG,Cai H, Brodie TA, et al. Mice lacking themetalloprotease-integrin MDC9 (ADAM9) have no evident major abnormalitiesring development or adult life. Mol Cell Biol 2002;22:1537–44.rou CM, Jeffrey SS, van de Rijn M, et al. Distinctive gene expres-n patterns in human mammary epithelial cells and breast cancers.c Natl Acad Sci U S A 1999;96:9212–7.uwerts AM, Meijer-van Gelder ME, Timmermans M, et al. HowAM-9 and ADAM-11 differentially from estrogen receptor predict
ponse to tamoxifen treatment in patients with recurrent breastcer: a retrospective study. Clin Cancer Res 2005;11:7311–21.rino P, Mauch C, Fox JW, Nischt R. Adam-9 expression and reg-
48. Hienco

r Res; 70(20) October 15, 2010


tion in human skin melanoma and melanoma cell lines. Int J Can-r 2005;116:853–9.igemura K, Sung SY, Kubo H, et al. Reactive oxygen species me-te androgen receptor- and serum starvation-elicited downstreamnaling of ADAM9 expression in human prostate cancer cells.state 2007;67:722–31.ng SY, Kubo H, Shigemura K, et al. Oxidative stress inducesAM9 protein expression in human prostate cancer cells. Cancers 2006;66:9519–26.pson KJ, Selfors LM, Bui J, et al. Identification of genes that reg-te epithelial cell migration using an siRNA screening approach.t Cell Biol 2008;10:1027–38.th D, Slocombe PM, Webster A, Stephens PE, Docherty AJ,rphy G. Meltrin γ (ADAM-9) mediates cellular adhesion through)β(1) integrin, leading to a marked induction of fibroblast celltility. J Cell Sci 2000;113:2319–28.minetti MR, Martin AC, Ribeiro JU, et al. Inhibition of platelets andor cell adhesion by the disintegrin domain of human ADAM9 to col-en I under dynamic flow conditions. Biochimie 2009;91:1045–52.odeti CK, Albrechtsen R, Grauslund M, et al. ADAM12/syndecan-4naling promotes β1 integrin-dependent cell spreading throughtein kinase Cα and RhoA. J Biol Chem 2003;278:9576–84.rao T, Nanba D, Tanaka M, et al. Overexpression of ADAM9hances growth factor-mediated recycling of E-cadherin in human
lon cancer cell line HT29 cells. Exp Cell Res 2006;312:331–9.
Cancer Research



Published OnlineFirst August 24, 2010.Cancer Res Jessica L. Fry and Alex Toker Have Opposing Effects on Breast Cancer Cell MigrationSecreted and Membrane-Bound Isoforms of Protease ADAM9

Updated version

10.1158/0008-5472.CAN-09-4231doi:

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/early/2010/09/17/0008-5472.CAN-09-4231To request permission to re-use all or part of this article, use this link

Research. on April 10, 2019. © 2010 American Association for Cancercancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-09-4231

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2010/09/17/0008-5472.CAN-09-4231


research secreted and membrane-bound isoforms of protease

Documents