HER2 Signaling Drives DNA Anabolism and Proliferation through SRC-3 Phosphorylation and ...

14 Pages·2016·2.08 MB·English

Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Cancer
MolecularandCellularPathobiology Research
HER2 Signaling Drives DNA Anabolism and
Proliferation through SRC-3 Phosphorylation and
E2F1-Regulated Genes
BryanC.Nikolai1,RainerB.Lanz1,BrianYork1,SubhamoyDasgupta1,NicholasMitsiades1,2,3,
Chad J.Creighton2,4, Anna Tsimelzon5, Susan G. Hilsenbeck5,DavidM. Lonard1,
Carolyn L. Smith1, and Bert W.O'Malley1
Abstract
Approximately 20% of early-stage breast cancers display scripts regulated by HER2 signaling are highly enriched with
amplificationoroverexpressionoftheErbB2/HER2oncogene, E2F1bindingsitesanddefineagenesignatureassociatedwith
conferringpoorprognosisandresistancetoendocrinetherapy. proliferativebreasttumorsubtypes,cell-cycleprogression,and
Targeting HER2þ tumors with trastuzumab or the receptor DNA replication. We show that HER2 signaling promotes
tyrosinekinase(RTK)inhibitorlapatinibsignificantlyimproves breast cancer cell proliferation through regulation of E2F1-
survival, yet tumor resistance and progression of metastatic driven DNA metabolism and replication genes together with
disease still develop over time. Although the mechanisms of phosphorylationandactivityofthetranscriptionalcoactivator
cytosolic HER2 signaling are well studied, nuclear signaling SRC-3.Furthermore,ouranalysesidentifiedacyclin-dependent
components and gene regulatory networks that bestow kinase (CDK) signaling node that, when targeted using the
therapeutic resistance and limitless proliferative potential are CDK4/6 inhibitor palbociclib, defines overlap and divergence
incompletely understood. Here, we use biochemical and bio- of adjuvant pharmacologic targeting. Importantly, lapatinib
informaticapproachestoidentifyeffectorsandtargetsofHER2 andpalbociclibstrictlyblockdenovosynthesisofDNA,mostly
transcriptional signaling in human breast cancer. Phosphory- through disruption of E2F1 and its target genes. These results
lationandactivityoftheSteroidReceptorCoactivator-3(SRC- have implications for rational discovery of pharmacologic
3)isreduceduponHER2inhibition,andrecruitmentofSRC-3 combinationsinpreclinicalmodelsofadjuvanttreatmentand
toregulatoryelementsofendogenousgenesisimpaired.Tran- therapeutic resistance. CancerRes;76(6);1463–75.(cid:1)2016AACR.
Introduction domain proteins that function, in part, as signal adaptors
and amplifiers for downstream kinase activity. Ras/Raf/MAPK
TheErbB/HERreceptortyrosinekinase(RTK)familyofmem-
and PI3K/AKT pathways are subsequently activated and the
branegrowthfactorreceptorsactivatesmultiplesignalingpath-
cascadingsubstratesofthesephospho-signaling eventsbecome
waysandisoftenassociatedwithcancer.Tyrosinekinaseactivity
effectorsofcelldivision,evasionofapoptosis,andgeneraltumor-
in this family of receptors is accomplished through homo- or
igenicity(1,2).
heterodimerization of ErbB family members (EGFR/ERBB1,
Steroid Receptor Coactivator-3 (SRC-3/AIB1/NCOA3) is a
HER2/ERBB2,ERBB3,andERBB4)uponligandbindingand/or
potent transcriptional coregulator for various nuclear receptors
membranejuxtaposition,andsubsequentauto-andcross-phos- andtranscriptionfactors,includingE2F1,NFkB,andmembersof
phorylationofintracellulardomains(1).Phosphorylatedrecep- theETSfamily(3–6).SRC-3isamplifiedand/oroverexpressedin
tortyrosineresiduesserveasdockingplatformsforvariousSH2 awidevarietyoftumors,includingamplificationinupto10%
and overexpression in up to 60% of breast tumors (3, 7–10).
Importantly,SRC-3isacoactivatingpartnerforestrogenreceptor
1Department of Molecular and Cellular Biology, Baylor College of (ERa),thecognateeffectorforthemitogenicactionofestrogenic
Medicine,Houston,Texas.2DepartmentofMedicine,BaylorCollege
of Medicine, Houston, Texas. 3Center for Drug Discovery, Baylor steroidsinthemammaryepithelium(11).ERaexpressionisan
CollegeofMedicine,Houston,Texas.4DanL.DuncanCancerCenter importantbiomarkerforprognosisandtreatmentandispresent
DivisionofBiostatistics,BaylorCollegeofMedicine,Houston,Texas. in (cid:2)70% of breast tumors at diagnosis, most of which receive
5Lester and Sue Smith Breast Center, Baylor College of Medicine,
endocrinetherapytoblockestrogensynthesisorreceptoractivity
Houston,Texas.
(12,13).ThefunctionofSRC-3inbreastcancerhasbeeninten-
Note: Supplementary data for this article are available at Cancer Research
sively studied because it is a major coactivator for ERa and a
Online(http://cancerres.aacrjournals.org/).
prominent,overexpressedoncogene.Thecoactivationpotential
C.L.SmithandB.W.O'Malleycontributedequallytothisarticle. ofSRC-3istightlyregulatedbyposttranslationalmodifications
CorrespondingAuthor:BertW.O'Malley,DepartmentofMolecularandCellular that alter protein activity, stability, and subcellular localization
Biology,BaylorCollegeofMedicine,OneBaylorPlaza,Houston,TX77030. (14,15).InthecontextofERaandNFkB-regulatedtranscription,
Phone:[phone];Fax:[phone];E-mail:[email protected]
SRC-3isdifferentiallyactivatedviaphosphorylationatmultiple
doi:10.1158/0008-5472.CAN-15-2383 residuesinresponsetoestradiolorTNFa(14,16).Collectively,
(cid:1)2016AmericanAssociationforCancerResearch. thesestudiespositionSRC-3asapotential"signalintegrator"of
www.aacrjournals.org 1463
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Nikolaietal.
variouscytokineandhormoneelicitedcell-signalingeventsthat Samplepreparation,immunoblotting,andRT-PCR
specifydistinctpatternsofgeneexpression(17). Forproteinanalysis,cellswerelysed(20mmol/LTris–HClpH
Clinicalstudieshaveshownthatbreastcancerpatientswith 8.0,150mmol/LNaCL,1mmol/LEDTA,and0.5%NP-40)for1
tumorsexpressinghighlevelsofbothHER2andSRC-3display hour at 4(cid:4)C, followed by centrifugation at 12,000 (cid:3) g for 10
resistance to endocrine therapy and have reduced disease-free minutes.Thesupernatantwasremovedtoafreshtubeanddiluted
survival (12, 18, 19). Furthermore, SRC-3 knockout mice are forquantitationusingtheBradfordmethod.Samplesweresep-
resistanttoHER2/neu-mediatedbreasttumorigenesisanddis- aratedusingNuPAGE4%to12%Bis–Trisgels(LifeTechnologies)
play significantly reduced phosphorylation of key signaling andtransferredtonitrocelluloseusingmethanoltransferbuffer.
components, including HER2, AKT, and JNK (20). In MCF-7 Membraneswereblockedin5%fat-freemilkreconstitutedinPBS
breast cancer cells, SRC-3 phosphorylation is influenced by plus 0.5% Tween-20 (PBST) for 1 hour at room temperature.
HER2 (21). Together, these studies suggest HER2 and SRC-3 Primary antibodies were diluted in 1% fat-free milk in PBST
are cooperating oncogenes whereby HER2 signaling events andblotsincubatedforseveralhoursatroomtemperatureor
influence SRC-3 activity on discrete gene programs. However, overnight at 4(cid:4)C.Species-specific secondary antibodiesconju-
the transcriptional targets of HER2 and SRC-3 are largely gated with horseradish peroxidase were used with traditional
undescribedanditisunknownwhetherintrinsicSRC-3activity enhanced chemiluminescense and X-ray film. Densitometric
andspecificphosphorylationsitesareaffectedbyHER2.Ampli- quantification was performed using ImageJ and numbers in
fication of HER2 occurs in approximately 20% of all breast figures represent the area under curve for each band, normal-
tumors, and overexpression of HER2 and SRC-3 is associated ized to actin in the identical lane. TRIzol reagent (low pH
withendocrineresistanceandextremelypoorprognosis.Given phenol;LifeTechnologies)wasusedtocollectandextracttotal
the link between HER2 signaling and SRC-3 in breast cancer RNA. RNA concentration was determined using a NanoDrop
andtheimportanceofSRC-3phosphorylationforitscoactiva- spectrometer and integrity was evaluated using Lab-on-chip
tion of steroid receptors and other transcription factors, the (Agilent).Reversetranscriptionusingrandomhexamerpriming
combined effects of HER2 and SRC-3 on gene regulation are (BioRadiScript)wasperformedbeforeqPCRonABIOneStep
understudiedinbreastcancer.Additionally,nucleareffectorsof with Roche Universal Library hydrolysis probes. Data were
HER2signalingthatregulatetranscriptionoftumorigenicgene analyzedusingtheDDCTmethod(seewww.nursa.org/qpcr_tu-
networks are incompletely understood. torials/forreview).Detaileddescriptionsofprimersequences
Here,westudytheeffectsofHER2-mediatedsignalingevents are listed in Supplementary Table S2.
on SRC-3 activity in breast cancer cells with amplified HER2
andSRC-3genes.HER2depletionattenuatessite-specificphos- RNAi,transfections,andluciferaseassays
phorylation of SRC-3, leading to a reduction in its intrinsic siRNA was custom designed and/or ordered as prevalidated
activity.Transcriptomicanalysisidentifiesgenesthatareregu- duplexes (Life Technologies; see Supplementary Table S2 for
lated by HER2 signaling, with a massive enrichment for tran- details).BT-474cellswereplatedon6-wellplatesatadensityof
scriptsinvolvedinmitoticcell-cycleprocessesandDNArepli- 4 (cid:3) 105 per well 1 day prior to transfection in reduced serum
cation.AstrikingmajorityofgenesdownregulateduponHER2 media(OPTI-MEM;LifeTechnologies)atafinalconcentrationof
depletion contain canonical and evidence-based binding sites 40nmol/LsiRNA.LuciferasereporterandplasmidDNAtransfec-
for the E2F1 transcription factor. Moreover, bioinformatics tions with siRNA are performed similarly with Lipofectamine
reveals a common HER2/SRC-3/E2F1 target gene (CCND1) 2000 (Life Technologies). Cells were harvested in Tris–EDTA–
thatissignificantlycorrelatedwithsensitivitytoRKTinhibitors NaCl (TEN) buffer, pelleted at 3500 (cid:3) g, and suspended in
andidentifiesaCDKsignalingnodeamendabletopharmaco- reporterlysisbuffer(Promega).Reporteractivitywasmeasured
logictargeting.OurresultssuggestthatSRC-3andE2F1servea using a Thermo Luminoskan Ascent. Relative light unit (RLU)
membrane-to-nuclear conduit function for HER2 signaling, wascalculatedrelativetototalproteinquantity.Detaileddescrip-
resulting in an anabolic DNA gene signature exploitable by tionsofreagentsandcatalognumbersarelistedinSupplementary
combinatorial therapies. TableS2.
Microarrayanalysis,motifscanning,andbioinformatics
Materials and Methods
RNAwaspreparedfromBT-474cellsasdescribedaboveand
Cellculture,growthassays,andreagents hybridized to Affymetrix GeneChip Human Genome U133A
BT-474andSKBR-3cellswereculturedinDMEM(Gibco/Life 2.0ArraychipsbytheGenomicandRNAProfiling(GARP)core
Technologies) supplemented with 10% fetal bovine serum facility at the Baylor College of Medicine, and analysis was
(FBS) and 2 mg/L insulin. Cells were obtained from ATCC, performed using BRB Array Tools (http://linus.nci.nih.gov/
andcellidentitywasverifiedusingshorttandemrepeat(STR) BRB-ArrayTools.html), dChip (http://www.hsph.harvard.edu/
analysisroutinelybytheTissueCultureCoreatBaylorCollege cli/complab/dchip), and ArrayAnalyzer (http://www.solution-
ofMedicine,Houston,TX.Forgrowthassays,5(cid:3)104cellswere metrics.com.au/products/s-plus_arrayanalyzer/default.html).
platedperwellon12-or24-wellmultiwellplatesincomplete MotifscanningwasperformedusingP-scanVer.1.2.2(http://
mediaandallowedtosettle,andthefollowingdaywererinsed 159.149.160.51/pscan/)with(cid:5)950toþ50criteriaofmatched
oncewithphenolred-freeDMEMbeforetreatmentinidentical probesetsthatwereatleast2-folddownregulateduponHER2
conditions supplemented with 5% charcoal-stripped FBS. depletion.Bioinformaticswereperformedusingacollectionof
For Figs. 1A, C, E and 6A and B, crystal violet staining with resourcesintheDanL.DuncanCancerCenterandDepartment
colorimetric absorbance was used to measure cell density. ofMolecularandCellularBiology.Briefly,theGalaxy/Cistrome
For Fig. 6C and D, MTS (Promega) assay was used. Detailed suite, NCBI Gene ID, and existing data deposits were used in
descriptionsofreagentsarelistedin SupplementaryTableS2. combination with SQL-based database organization to queue
1464 CancerRes;76(6)March15,2016 CancerResearch
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
HER2RegulatesDNAReplicationthroughE2F1TargetGenes
Figure1.
BT-474cellsareestrogenresponsive
anddisplayHER2-dependantSERM
resistance.A,BT-474breastcancer
cellgrowthcurveupontreatmentwith
0.1%ethanol(Veh),17-b-estradiol
(E2),4-hydroxytamoxifen(4HT),or
raloxifene(Ral).B,transcriptlevelsof
canonicalERatargetgenesare
elevatedupontreatmentwithE2.C,
depletionofHER2usingsiRNAresults
inlossofestrogenmitogenicactivity
andSERMresistance.D,inductionof
SERM-inducedapoptosisinBT-474
cellsuponknockdownofHER2.E,
depletionofeitherHER2orSRC-3in
BT-474cellsgrownincompletemedia
decreasescellviability.
anddisplayintegratedtranscriptomeandcistromedatabased supernatant was incubated with 4-mg SRC-3 antibody (Cell
on expression and protein binding, respectively. Gene ontol- Signaling5E11)orcontrolIgGovernight.Agarosebeadswere
ogy was displayed using DAVID and pathway mapping by used to precipitate immune complexes for 1 hour before
Kyoto Encyclopedia of Genes and Genomes (KEGG). Gene washing in buffers of increasing stringency as follows: 1X in
signature analysis involving the Kessler and colleagues (22) 20mmol/LTris–HCl(pH8.0),150mmol/LNaCl,2mmol/L
and The Cancer Genome Atlas (TCGA; ref. 23) datasets was EDTA,and1%TritonX-100;1(cid:3)in20mmol/LTris–HCl(pH
carried out as previously described (24). Microarray data are 8.0),500mmol/LNaCl,2mmol/LEDTA,and1%TritonX-100;
deposited in the Gene Expression Omnibus as accession 1(cid:3) in 10 mmol/L Tris–HCl (pH 8.0), 1 mmol/L EDTA, 1%
GSE71347. deoxycholate, 1% NP-40, and 0.25 M LiCl; and 2(cid:3) with TE.
Beadswerecentrifugedat1,000(cid:3)gfor1minutetopellet.DNA
Chromatinimmunoprecipitation waselutedfrombeadsin100mmol/LNaHCO3and1%SDSat
ChIP was performed as described with minor modifications 65(cid:4)C overnight. Extraction was performed using PCR product
(25).Briefly,cellswerefixedin1%formaldehydebeforequench- purification columns (Qiagen).
ingwith125mmol/Lglycine.Cellswerewashedtwicewithice-
coldTENbufferandscrapedintotubesonice.Cellpelletswere Tritium(3H)-labeledthymidineincorporation
eitherflashfrozenandstoredat(cid:5)80(cid:4)Corresuspendedin500mL Cellswereplatedon6-wellplatesatadensityof2(cid:3)105perwell
ofTris–EDTA(TE)/1%SDS.Allstepsbeyondthispointweredone andtreatedwithlapatinib,palbociclib,orbufalinfor24hoursin
oniceorat4(cid:4)C.Thissuspensionwassonicatedfor8cyclesof normalgrowthmedia.FourmCof3H-labeledthymidine(Perkin
10 pulses at amplitude 90 and duty cycle 3.5 on a Branson Elmer) was added to the media, and cells were grown for two
Sonifier250.Chromatinextractswereclearedbycentrifugation additionalhours.Mediawereremoved,andcellswerefixedfor5
and diluted 1:10 in a buffer containing 20 mmol/L Tris–HCl minutes with 25% acetic acid in methanol, rinsed once with
(pH8.0),150NaCl,2mmol/LEDTA,and1%Triton-X-100to methanol,andproteinprecipitatedusing5%trichloricacidfor
reduce SDS concentration. Four micrograms of normal rabbit 5minutes.Wellswerewashedanadditional3timeswithmeth-
IgGantibody(SantaCruzBiotechnology)wasusedwith50mL anolandallowedtodry.Radioactivity wasrecoveredusing1N
Protein A/G Agarose plus salmon sperm DNA (Millipore) NaOHandequilibratedwithanequalvolumeof1NHClbefore
to preclear the diluted chromatin for 1 hour. The cleared the addition of biodegradable counting scintillant and liquid
www.aacrjournals.org CancerRes;76(6)March15,2016 1465
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Nikolaietal.
scintillation using a Beckman scintillation counter. For E2F1 (pBIND-SRC-3) and a pGL5-Luciferase reporter was used to
rescueexperiments,cellswereplatedon12-wellplatesatadensity interrogate whether HER2 influences intrinsic SRC-3 activity.
of5(cid:3)104perwellandinfectedusing25virionspercellfor24 UponHER2knockdown,intrinsicSRC-3activityissignificantly
hours,thentreatedwithlapatinibfor24hoursfollowedby2mC reduced(Fig.2D).Theseresultscanbepharmacologicallyreca-
of 3H-thymidine treatment for 2 hours and processing as pitulated using an inhibitor of HER2, AG-825 (tyrphostin;
described. Fig.2E).Additionally,transienttransfectionofHER2issufficient
toenhanceSRC-3transcriptionalactivity(Fig.2F,compare200vs.
300ngofpBind-SRC-3).Thesedatacorroboratepreviousstudies
Results
suggestingaroleforSRC-3inHER2tumorigenesisandhighlight
BT-474cellsdisplayHER2-dependentestrogenresponsiveness the ability of this oncogenic signaling pathway to elicit strong
andresistancetoantiestrogens effects on transcriptional processes. Taken together, our results
Resistancetoselectiveestrogenreceptormodulators(SERM)in show that HER2 signaling has specific effects on SRC-3 phos-
HER2þ breast cancer is incompletely understood. Specifically, phorylationandiscapableofincreasingtheintrinsicactivityof
whetherHER2expression isrequiredforsensitivitytoestrogen SRC-3.
andSERMsinHER2-amplifiedcellsremainsunanswered.TheBT-
474breastcancercelllinerepresentstheidealmodelsystemto HER2depletiondrasticallyaltersthetranscriptionallandscape
studyeffectsofendogenouslyamplifiedHER2andSRC-3onco- ThetranscriptionalconsequencesofHER2actionundoubtedly
genes on estrogen signaling because it is also positive for ERa play a significant role in its oncogenic activity, but relatively
expression(datanotshown;ref.3).Wefirsttestedthemitogenic fewstudieshavecomprehensivelyinvestigatedthetranscriptional
activityof17b-estradiol(E2)andeffectsoftheSERMs4-hydro- targets of HER2 signaling. To define the HER2-regulated tran-
xytamoxifen (4HT) and raloxifene (Ral) on cell growth and scriptome,weperformedmicroarrayanalysisoncellstransfected
gene expression. Cell density was significantly increased by E2 with control or HER2 siRNA (Fig. 3A). HER2 knockdown
relative to vehicle treatment, while tamoxifen and raloxifene decreased AKT and Raf phosphorylation, indicating disruption
failed to decrease cell density relative to the vehicle control, ofthedownstreamHER2signalingcascade(Fig.3B).Substantial
indicating resistance (Fig. 1A). Likewise, transcript levels of transcriptionalchangesuponHER2depletionareevidentinthe
canonical ERa-target genes were significantly induced upon E2 heatmapshowninFig.3C.GeneOntologyanalysisoftranscripts
treatment, but not affected by treatment with either SERM downregulated upon HER2 depletion indicated an impressive
(Fig.1B).UponHER2knockdown,E2isnolongermitogenicin decreaseoftranscriptsinvolvedincellcycleandmitoticprocesses
BT-474cells(Fig.1C,E2panel).Moreover,siRNAagainstHER2 (Fig.3D),andRT-PCRconfirmedlossoftranscriptsinvolvedinS-
significantlysensitizedcellstotreatmentwith4HTorRal(Fig.1C, phase DNA replication, Ras signaling, and transcription factor
4HTandRalpanels;notethedecreaseincelldensityovertimein genes (Supplementary Fig. S1A). Moreover, parsing this HER2-
siHER2transfectedcells).TheseresultsconfirmthatBT-474cell regulatedtranscriptomewithgenesignaturesassociatedwithRas/
growth requires HER2 expression for estrogen sensitivity and Raf/MAPKorPI3K/AKTsignalingrevealedsignificantoverlapand
SERM resistance and indicate that HER2 protects cells from divergence between these signaling pathways (Supplementary
SERM-inducedapoptosis.Indeed,whenHER2isdepletedusing Fig.S1B–S1D).Confirmingaroleforcell-cycledisruptionwhen
siRNA,caspase-3iscleaved(activated)upontreatmentwith4HT HER2 is depleted, a reduction in G to S phase transition is
1
orRal,butnotvehicleorestrogen(Fig.1D,compareTandRtoV observedusingflowcytometrytoanalyzeDNAcontent(Supple-
orEinsiHER2lanes).TheactionofSERMsispostulatedtooccur, mentaryFig.S2).
atleastinpart,throughdisruptionoftheinteractionbetweenERa Next,wecomparedgenesthatweredownregulateduponHER2
anditstranscriptionalcoactivators,suchasSRC-3(26).Depletion knockdownwithpubliclyavailablebreastcancergeneexpression
of SRC-3 also induced caspase-3 activation (Fig. 1D) and patterns in TCGA in relation to HER2 and SRC-3 (NCOA3)
decreasedsurvivalofbreastcancercellsgrowninsteroidreplete expression levels (Supplementary Fig. S3). Strikingly, tumor
media(Fig.1E),demonstratingafunctionaldependencyonboth samples with the highest expression of HER2 (Supplementary
HER2andSRC-3survivalpathwaysinthiscellmodel. Fig. S3, top) or SRC-3 (Supplementary Fig. S3, bottom) also
displayedhighexpressionofHER2-sensitivegenesidentifiedby
HER2signalingaffectsSRC-3phosphorylationandactivity microarrayuponHER2depletion.Jointlyconsidered,thesedata
Steroid Receptor Coactivator-3 is a potent coactivator for indicate that loss of HER2 in breast cancer cells results in a
numeroustranscriptionfactorsinvolvedinbreastcancerandthe substantialdecreaseoftranscriptsinvolvedincellularprolifera-
preferenceofSRC-3for,andactivitywith,differenttranscription tionthatrelatestoexpressionofHER2andSRC-3inalargedataset
factorsisinfluencedbycombinatorialphosphorylation(14).To ofbreasttumorsamples.
determineifHER2signalingaffectsSRC-3phosphorylationand BecauseHER2ispresentatextracellularmembranesandthus
function, siRNA was used to deplete HER2 in BT-474 cells. acts as an indirect effector of nuclear transcription, we next
KnockdownofHER2resultsindecreasedSRC-3phosphorylation investigatedtranscriptionfactor–bindingmotifsintheproximal
atT24,S543,S857,andS860(Fig.2A–C).Importantly,thislossof promoterregionsofdownregulatedgenes.Interrogationof1-kb
phosphorylationdoesnotappeartobeuniformforSRC-3protein upstreampromotersequencesofputativetargetgenesidentified
becauseS505phosphorylationremainsunchangeduponHER2 anE2F1 consensussequence asthemost significantlyenriched
depletion(Fig.2A).Thesedataareconsistentwithalackofchange transcriptionfactormotif(Fig.3E).TranscriptlevelsforE2F1and
intotalSRC-3proteinlevelsuponHER2knockdown,asS505has its target genes CCND1 and CDC25A have similar expression
beendescribedtofunctionaspartofaphospho-degron(27). patterns in our microarray analysis (Fig. 3F) and HER2 knock-
BecauseSRC-3doesnotdirectlybindDNA,asystemusinga downdepletestheproteinlevelsofthesegeneproducts(Fig.3G–
GAL4-DNA binding domain fused to the N-terminus of SRC-3 I).UponHER2knockdown,E2F1proteinlevelswereabolished
1466 CancerRes;76(6)March15,2016 CancerResearch
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
HER2RegulatesDNAReplicationthroughE2F1TargetGenes
Figure2.
HER2affectsphosphorylationand
activityofSRC-3.AandB,BT-474
cellsweretransfectedwithsiRNA
targetingdifferentregionsofHER2
transcript,andcelllysateswere
immunoblottedusingphospho-
specificSRC-3antibodies.SRC-3
phosphorylationisdecreasedupon
HER2depletionatT24,S543,S860,
andS857,whilephosphorylationof
S505isunaltered.C,densitometry
quantificationofS857
phosphorylation.D,BT-474cells
transfectedwithpBind-SRC-3/GAL4-
LuciferasereportersystemandsiRNA
targetingHER2.KnockdownofHER2
usingtwoseparatesiRNAsdecreases
intrinsicSRC-3activity.E,AG-825
(tyrphostin)inhibitionofHER2
decreasespBind-SRC-3activity.F,
cotransfectionofHER2relievesthe
activationplateauofpBind-SRC-3
inHeLacells.(cid:6),P<0.05forall
experiments.
(Fig. 3G), which also was observed upon SRC-3 knockdown, published chromatin immunoprecipitation-sequencing (ChIP-
indicatingthatHER2andSRC-3signalinglikelyconvergeonE2F1 seq) inbreast cancercells (31)werecompared with transcripts
targetgenes(Fig.3H).QuantitativeRT-PCRscreeningof HER2 fromourBT-474siHER2microarray.Briefly,chromosomalcoor-
target genes revealed that transcripts for additional mitogenic dinatesofsignificantlydownregulatedgeneswerecomparedwith
transcription factors FOXM1 and ETV4, as well as the G to S E2F1bindingsiteintervalswithintheextendedpromoterregion
1
checkpoint gene CCND1, are significantly decreased in cells (EPR,(cid:5)7.5toþ2.5Kb)oranywherewithinthegeneandinclud-
transfectedwithsiRNAagainstSRC-3orHER2(Supplementary ing10-kbgene-flanking regions (10kb). Astrikingmajority of
Fig.S4C).TheseresultssuggestthatdepletionofHER2orSRC-3 genesregulatedbyHER2(>75%)inourmicroarrayareboundby
reducestheexpressionofmultipletranscriptionfactors,including E2F1 (Fig. 4A). Atypical E2F7 and E2F8 ChIP-seq binding in
E2F1, which are known to regulate the transcription of genes cancer cells also shows significant overlap with the HER2 gene
involvedinmitoticprogressionandproliferation(28–30). signature, and E2F8 may be a HER2 and SRC-3 target gene
(SupplementaryFig.S5A–S5C).Transcriptandproteinlevelsof
AmajorityofgenesregulatedbyHER2containE2F1binding E2F7andE2F8werereducedwithsiRNAtransfectionofsiHER2or
sites siSRC-3(SupplementaryFig.S5D–S5F).Theseresultsareintrigu-
All E2F family gene transcripts except E2F4 and E2F6 were ingconsideringaproposedroleofatypicalE2Fsinmaintenanceof
decreaseduponHER2knockdowninthemicroarray,andE2F7 genome size and cancer, but E2F7 and E2F8 function was not
andE2F8wereamongthemostdepletedtranscriptsintheentire further explored in this study (32, 33). Additionally, HER2-
array (Table 1 and Supplementary Fig. S3). Given the large regulated genes with E2F1 promoter binding sites displayed
number of cell-cycle and mitotic genes downregulated upon overlap with PI3K/AKT and Ras/Raf/MAPK gene signatures
HER2depletionandtheenrichmentforE2F1motifs,weinves- (SupplementaryFig.S6A).
tigated E2F1 transcription factor binding to DNA regulatory TheCCND1geneisflankedbyE2F1bindingsitesincludinga
elementsofHER2-regulatedgenes.BindingsitesforE2F1from knownSRC-3bindingelementdownstream,colloquiallytermed
www.aacrjournals.org CancerRes;76(6)March15,2016 1467
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Nikolaietal.
Figure3.
TheHER2-regulatedtranscriptome
revealslossofgenesinvolvedincell-
cycleregulationdownstreamofthe
HER2/SRC-3/E2Fsignalingaxis.A,
schematicofmicroarrayexperimental
groupsandHER2knockdownin
biologictriplicates(right,
immunoblot).B,HER2knockdown
resultsindisruptionofAKTandRas/
Rafsignalingcascades,asevidenced
bydecreasedAKTandc-Raf
phosphorylation.DecreaseofSRC-3
phosphorylationwasconsistently
observeduponHER2knockdown.
C,heatmapofsignificantlyaltered
probesetsacrosstreatmentgroups.
D,topgeneontologytermsreturned
fromHER2-downregulatedgenelist
specifycell-cycleandmitotic
processes.E,predictedE2F1binding
sitesareenrichedinthepromoter
regionofHER2-downregulatedgenes.
F,enlargedheatmapshowing
decreasedlevelsofE2F1,CCND1,and
CDC25AuponHER2knockdown.G,
knockdownofHER2reducesE2F1
proteinexpression.H,knockdownof
SRC-3reducesE2F1protein
expression.I,proteinlevelsofgenes
fromFwithcontrolandHER2siRNA
transfectioninBT-474cells.
enhancer2(enh2;Fig.4B;ref.34).Thisenhancerregionisbound suggestedfromcistromicandtranscriptomicanalyses,comparing
by E2F1 and SRC-3 in BT-474 cells (Fig. 4C), and upon HER2 ourHER2genesignatureexpressionscoreswithtwolargebreast
knockdown,SRC-3recruitmenttoCCND1enh2issignificantly tumordatasetsrevealsastrongcorrelationwithE2F1expression
reduced(Fig.4D).Moreover,treatmentwiththeHER2inhibitor (Tableinsets, Fig.5Aand B).Additionally, theHER2-regulated
lapatinib results in decreased SRC-3 recruitment, and H3K27 genesignaturecorrelateswiththemostproliferativebreastcancer
acetylation(butnotH3K14ac)islostfromtheenhancerregion subtypes: basal, HER2þ, and luminal B (Fig. 5A and B, histo-
(Fig.4E).Takentogether,thesedatashowthatHER2-mediated grams).Tumorsconsideredtobeslowgrowing(luminalAand
generegulationinvolvesmechanismsofE2F1andSRC-3recruit- normal-like)displayanegativecorrelationwithourHER2gene
menttogeneregulatoryelementsinvolvedinmitogenicexpres- signature,asdothepredictorsofluminalAsubtype,ERaandPR
sionofCCND1. (Fig. 5). Collectively, gene signatures from patient samples are
consistent with our in vitro findings and indicate that HER2
TheHER2transcriptomesignaturecorrelateswithE2F1and signaling mediates a proliferative regulatory gene network in
proliferativebreastcancersubtypes breastcancer.
BT-474cellshavepreviouslybeencharacterizedasaLuminalB Perturbationanalysis(36)ofHER2signaturegeneswithE2F1
subtypemodelforbreastcancerowingtoHER2amplificationand binding sites using functional annotation with the Genomic
ERaexpression(35),andweobserveresponsivenesstoestrogen Regions Enrichment Annotation Tool (GREAT; ref. 36) reveals
butHER2-dependentresistancetoantiestrogens(seeFig.1A).As asignificantenrichmentofoncogenicgenenetworksindiverse
1468 CancerRes;76(6)March15,2016 CancerResearch
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
HER2RegulatesDNAReplicationthroughE2F1TargetGenes
Table1. HER2regulatesDNAanabolismandreplication Table1.HER2regulatesDNAanabolismandreplication (Cont'd)
siHER2log2FC GeneID Genesymbol siHER2log2FC GeneID Genesymbol
E2Ftranscriptionfactors Celldivisioncycleproteins
(cid:5)2.6 144455 E2F7 (cid:5)0.3 11140 CDC37
(cid:5)1.9 79733 E2F8 (cid:5)0.3 64866 CDCP1
(cid:5)0.5 1875 E2F5 0.2 55664 CDC37L1
(cid:5)0.4 1871 E2F3 0.4 10602 CDC42EP3
(cid:5)0.3 1869 E2F1 0.4 55536 CDCA7L
(cid:5)0.3 1870 E2F2 0.4 8881 CDC16
1874 E2F4 0.5 51362 CDC40
1876 E2F6 0.8 8555 CDC14B
CyclinD1 Ribonucleotidereductases
(cid:5)0.9 595 CCND1 (cid:5)2.3 6241 RRM2
(cid:5)0.6 896 CCND3 (cid:5)0.9 6240 RRM1
894 CCND2 Originrecognitioncomplex
Go,Ichi,Ni,Sanfactors (cid:5)1.4 23594 ORC6
(cid:5)1.1 84296 GINS4 (cid:5)1.1 4998 ORC1
(cid:5)1 64785 GINS3 (cid:5)0.2 4999 ORC2
(cid:5)1 9837 GINS1 23595 ORC3
(cid:5)0.3 51659 GINS2 5000 ORC4
ReplicationproteinA 5001 ORC5
(cid:5)0.4 6117 RPA1 Chromodomain(cid:5)helicaseDNAbindingproteins
(cid:5)0.2 6118 RPA2 (cid:5)0.6 9557 CHD1L
6119 RPA3 (cid:5)0.3 1108 CHD4
29935 RPA4 (cid:5)0.2 55636 CHD7
TK,DHFR,SHMT,andTYMS 0.3 55349 CHDH
(cid:5)1.9 7083 TK1 0.4 84181 CHD6
(cid:5)1.5 1719 DHFR 1105 CHD1
(cid:5)0.8 6470 SHMT1 1106 CHD2
6472 SHMT2 1107 CHD3
(cid:5)0.7 7298 TYMS 26038 CHD5
IMPsynthesisenzymes 57680 CHD8
(cid:5)1.1 5471 PPAT 80205 CHD9
(cid:5)0.3 10606 PAICS MCMDNAreplicationfactors
(cid:5)0.2 5631 PRPS1 (cid:5)1.9 55388 MCM10
0.6 2618 GART (cid:5)1.4 4172 MCM3
DNAligasesandPCNA (cid:5)1.4 4175 MCM6
(cid:5)1.4 5111 PCNA (cid:5)1.4 4171 MCM2
(cid:5)1 3978 LIG1 (cid:5)1.2 4174 MCM5
(cid:5)0.4 3980 LIG3 (cid:5)0.9 4176 MCM7
3981 LIG4 (cid:5)0.9 84515 MCM8
Topoisomerases (cid:5)0.7 4173 MCM4
(cid:5)1.6 7153 TOP2A 254394 MCM9
(cid:5)0.5 11073 TOPBP1 ReplicationfactorC
(cid:5)0.4 7150 TOP1 (cid:5)1.2 5984 RFC4
0.5 116447 TOP1MT (cid:5)1.1 5985 RFC5
7155 TOP2B (cid:5)0.9 5983 RFC3
7156 TOP3A (cid:5)0.9 5982 RFC2
8940 TOP3B 5981 RFC1
Celldivisioncycleproteins DNAprimases
(cid:5)2.1 8318 CDC45 (cid:5)1.5 145270 PRIMA1
(cid:5)1.9 990 CDC6 (cid:5)1.2 5557 PRIM1
(cid:5)1.9 83879 CDCA7 (cid:5)0.6 5558 PRIM2
(cid:5)1.7 8317 CDC7 DNApolymerases
(cid:5)1.6 113130 CDCA5 (cid:5)2 5427 POLE2
(cid:5)1.5 993 CDC25A (cid:5)1.3 23649 POLA2
(cid:5)1.5 83461 CDCA3 (cid:5)1.2 10721 POLQ
(cid:5)1.4 991 CDC20 (cid:5)1.2 55703 POLR3B
(cid:5)1.2 55143 CDCA8 (cid:5)0.9 5425 POLD2
(cid:5)1 157313 CDCA2 (cid:5)0.8 5424 POLD1
(cid:5)1 994 CDC25B (cid:5)0.8 10714 POLD3
(cid:5)1 995 CDC25C (cid:5)0.7 5426 POLE
(cid:5)0.7 55038 CDCA4 (cid:5)0.6 51728 POLR3K
(cid:5)0.5 8697 CDC23 (cid:5)0.5 5422 POLA1
(cid:5)0.5 56882 CDC42SE1 (cid:5)0.4 5441 POLR2L
(cid:5)0.5 996 CDC27 (cid:5)0.4 661 POLR3D
(cid:5)0.4 998 CDC42 (cid:5)0.3 56655 POLE4
(cid:5)0.4 10435 CDC42EP2 (cid:5)0.3 5433 POLR2D
(cid:5)0.3 55561 CDC42BPG (cid:5)0.3 5434 POLR2E
(Continuedonthefollowingcolumn) (Co(nCtoinnutiendueodnothnethfoellfoowlloinwgincgolpumagne))
www.aacrjournals.org CancerRes;76(6)March15,2016 1469
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Nikolaietal.
Table1.HER2regulatesDNAanabolismandreplication (Cont'd)
functionwithCCND1iswelldescribedandrecentclinicaltrials
siHER2log2FC GeneID Genesymbol
havereportedoptimisticresultsandrecentFDAapprovalofthe
DNApolymerases CDK4/6-specific inhibitor palbociclib (PD 0332991, Pfizer;
(cid:5)0.3 5436 POLR2G
(cid:5)0.3 5423 POLB ref. 40), we questioned whether this drug could similarly alter
(cid:5)0.3 11128 POLR3A cell-cycle gene transcription and sensitize breast cancer cells to
(cid:5)0.2 55718 POLR3E other pharmacologic agents. As shown in Fig. 6A, palbociclib
(cid:5)0.2 171568 POLR3H preventedthemitogeniceffectsofestradiolsimilartoknockdown
(cid:5)0.2 26073 POLDIP2
of HER2 (see Fig. 1C). Additionally, combination treatment of
0.3 246721 POLR2J2 palbociclibwith4HTsignificantlyreducedcellsurvival(Fig.6A).
0.4 5430 POLR2A
0.4 57804 POLD4 Integratorsofcellularsignaling,suchastranscriptionalcoacti-
0.5 84172 POLR1B vators,mayrepresentanadditional"weaklink"(andthereforea
0.5 51426 POLK therapeutictarget)incancergeneregulationnetworksowingto
0.5 51082 POLR1D theirpleiotropiceffectsonmultipleproliferativepathways.When
0.6 84265 POLR3GL
palbociclib treatment is used in combination with bufalin, a
0.7 11232 POLG2
recentlydescribedinhibitorofSRC-3(41),weobserveasignif-
0.9 11201 POLI
391811 POLD2P1 icantcombinatorialeffectonthedecreaseincellsurvival(Fig.6B).
84271 POLDIP3 Similarly,cellularsensitivitytolapatinibisaugmentedbypalbo-
54107 POLE3 cicliborbufalininadose-dependentmanner(Fig.6CandD).
5428 POLG Notably,bufalinisactiveinthenanomolarrangeandsensitizes
5429 POLH
cellstolowerconcentrationsoflapatinib(Fig.6D).Similarresults
27343 POLL areobservedintheHER2þ/ERa(cid:5)breastcancercelllineSKBR-3,
27434 POLM
353497 POLN suggestingthatthesepathwayssignalcooperativelyinothercell
models(SupplementaryFig.S8).Palbociclibtreatmentreduced
NOTE: Selection of genes from enriched gene ontology terms involved in
nucleotidesynthesis,DNAreplicationforkformation,andreplicationproces- transcriptlevelsofcell-cyclegenesE2F1,CDC6,andCDC45,but
sivity.Numbersrepresent log fold change, with negativevalues indicating not CCND1 (Fig. 6E). Transcript levels of E2F1, CDC6, and
2
decreaseuponHER2knockdownandpositivenumbersindicatingincreased CCND1,butnotCDC45,aredepleteduponlapatinibtreatment,
expression.Boxedgenesymbolsarevalidatedtargetsrepresentativeofalarger suggestingthattheHER2pathwayconvergesonsimilartranscrip-
networkoftranscriptionaltargetsdownstreamofHER2signaling.
tional targets as CDK signaling (Fig. 6F). Additional genes
involvedincell-cycleregulationweredecreaseduponpalbociclib
cancertypes,includingresistancetogefitinib(SupplementaryFig. (i.e.,E2F7,E2F8,andMCM10)orlapatinib(i.e.,E2F7,E2F8,and
S6B).Importantly,CCND1isinvolvedinbreastcancersensitivity ETV4) treatment, indicating that these compounds influence
totheHER/EGFRfamilyinhibitorsgefitinib,lapatinib,andafa- broad cell-cycle transcriptional networks (Supplementary Fig.
tinib(GenomicsofDrugSensitivityinCancer;ref.37).Takenin S9A and S9B). Treatment with lapatinib also decreases protein
context,theseresultssuggestthatHER2/E2F1signalingregulates levelsofCCND1,CDC6,CDC45,MCM10,andribonucleotide
cyclinD1andothergenesthatmaypredictsensitivitytotherapyin reductase (RRM2), a rate-limiting enzyme in DNA nucleoside
proliferativebreastcancersubtypes. metabolism(SupplementaryFig.S9C).
CyclinD1/CDKcross-talkwithHER2convergeson AcutelapatiniborpalbociclibtreatmentrestrictsDNAsynthesis
transcriptionalregulationofcell-cyclegenes inHER2þbreastcancercells
Proliferationiscriticalintheetiologyandassociatedpathoge- GiventhebroadeffectofHER2andCDKsignalingpathwayson
nicityofcancer.Individualoncogenesandtumorsuppressorsdo transcriptionalregulationofgenesinvolvedincell-cycleandDNA
noteasilydefinethebatteriesofgenesthatregulateproliferation replication(Fig.6G),wereasonedthatpharmacologicallyblock-
per se, because cell mitosis, DNA replication, and escape from ingHER2orCDKshouldalsoblockreplicativesynthesisofDNA
differentiationpathwaysinvolvelargegenenetworksworkingin inHER2þbreastcancercells.Indeed,lapatinibpotentlyrestricts
concert.Importantly,becausethetumorhasbecomedependent DNAsynthesisatlownanomolarconcentrationsafter24hoursof
on cellular events resulting from amplification or mutation of treatment, before any major changes occur to cellular health
such genes to maintain a proliferative advantage, these genes (SupplementaryFig.S9D–S9G)butconcomitantwithdecreased
representfragilityinthenetworkofcancercellsignaling(2,38). phosphorylationofHER2andcell-cycleproteinexpression(Sup-
Indeed,thishasbecomethebasisfordirectedtherapiestargeting plementaryFig.S9H).Tritiated(3H)-thymidineincorporationto
oncogenicsignalingfromtyrosinekinasesHER2andEGFR(2). directly measure DNA synthesis revealed that acute (24-hour)
WeusednetworkanalysisofCCND1andanotherHER2/E2F1- treatment with lapatinib or palbociclib blocks de novo DNA
regulated gene from the microarray, CDC6, to identify a fragile replication,withoutaffectingcellularsize,shape,orviability(Fig.
componentofthissignalingnetworkamendabletopharmacologic 6H). Bufalin treatment blocks DNA synthesis to a lesser extent
targeting.NetworkanalysisofCCND1andCDC6revealedstrong (Fig.6H),suggestingthecombinedeffectofbufalinwithlapatinib
overlapwiththeHER2transcriptome.Remarkably,50%ofgenesin orpalbocicliboncellularviability(seeFig.6BandD)isconsistent
theCCND1-associatednetwork,and59%ofCDC6networkgenes withitsmechanisticroleofinducingapoptosis(41).Consistent
are regulated byHER2inour microarray(SupplementaryTable with the effect of lapatinib and palbociclib on transcription,
S1). Not surprisingly, members of the cyclin-dependent kinase protein levels of E2F1, CDC6, and CCND1 are reduced in BT-
(CDK)familyandtheirregulatorsareenrichedintheCCND1and 474andSKBR-3breastcancercellsafteracutetreatment(Fig.6I).
CDC6networks(SupplementaryFig.S7;ref.39),whichisconsis- Moreover, the lapatinib-induced reduction of replication fork
tent with KEGG pathway analysis. Because CDK4 and CDK6 proteinsandblockofDNAreplicationcanberescuedbyectopic
1470 CancerRes;76(6)March15,2016 CancerResearch
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
HER2RegulatesDNAReplicationthroughE2F1TargetGenes
Figure4.
CooperativeregulationofcyclinD1by
HER2andSRC-3.A,E2F1ChIP-seq
peakswerecalledfrompublished
resourcesandalignedtohg19for
parsingwithHER2-regulatedgenes
fromthemicroarray.Thepiegraph
indicatesthatamajorityofHER2-
regulatedgenescontainedatleastone
E2F1bindingsiteintheextended
promoterregion(EPR).B,E2F1and
SRC-3bindingfrompublishedChIP-
seqattheenhancer2(enh2)regionof
CCND1(stripedbar).C,E2F1andSRC-3
bindingbytheChIPassaytothe
CCND1enhancerregion.D,HER2
depletiondecreasesSRC-3bindingto
theCCND1enhancer.E,lapatinib
treatmentreducesCCND1transcript
levelsanddismissesSRC-3and
H3K27acfromtheCCND1enhancer.
expression of E2F1 (Fig. 6J). Collectively, these results demon- Discussion
stratethatHER2andCCND1/CDKsignalingpathwaysregulate
DNAreplicationthroughE2F1anditsassociatedtranscriptional WedescribeanE2F1transcriptionalnetworkdownstreamof
targets, many of which are directly involved in nucleotide bio- HER2signalingthatisinfluencedbySRC-3phosphorylationin
synthesis and replicative processivity. Moreover, our findings breast cancer cells. Overexpression of SRC-3 contributes to the
demonstratethepotentialforcombinedtargetingofHER2and etiologyandprogressionofbreastcancer,andmicelackingSRC-3
CDKsignaling pathwaysorDNAreplication maybearational areresistanttooncogeneandcarcinogen-inducedbreasttumor-
modelthatwarrantsfurtherinvestigationinpreclinicalstudies. igenesis(20,42,43).Moreover,SRC-3isacoactivatorforE2F1
www.aacrjournals.org CancerRes;76(6)March15,2016 1471
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.
Published OnlineFirst February 1, 2016; DOI: 10.1158/0008-5472.CAN-15-2383
Nikolaietal.
Figure5.
HER2-regulatedtranscriptional
signatureofBT-474cellsstrongly
correlateswithE2F1andproliferative
breastcancersubtypes.Genes
upregulatedordownregulatedby
HER2depletionwereinversely
correlatedwithexpressionsignatures
fromtwolargebreastcancerdatasets
fromKesslerandcolleagues(A;
ref.22)andTCGA(B).ERandPRare
negativelycorrelatedwiththeHER2
signatureandE2F1displaysahighly
significantpositivecorrelation,as
expectedfrommotifanalysisinFig.4.
Breastcancersubtypesthat
correlatedwiththeHER2gene
signature(basal,HER2þ,andluminal
B)aregenerallyconsideredtohave
ahighproliferativeindex.
transcription, resulting in breast cancer cell proliferation and critical to investigate how oncogenes such as SRC-3 and HER2
antiestrogen resistance (5). However, the involvement of E2F1 hijack E2F-mediated gene regulation to achieve wanton DNA
inHER2-driventranscriptionalprogramsinbreastcancerisrel- replicationandcellularproliferation.
ativelyunderstudied.Recently,severalstudieshavereportedE2F Cross-talkandoverlapinvolvingHER2andestrogensignaling
activators (typically considered to be E2F1-3) in thecontext of havebeenextensivelystudiedwithimplicationsforresistanceto
HER2/neu, c-Myc, or Ras oncogene–induced breast tumors in endocrineandadjuvanttherapiesinbreastcancer(48).Mechan-
mousemodels(44,45).AndrechekandWuandcolleaguesfound ismsproposedincludephosphorylation-mediatedexportofERa
thatE2f1andE2f3(andtoalesserextentE2f2)knockoutmiceare fromthenuclearcompartment,reductionofERaexpression,and
protectedfromHER2/neutumorigenesis,andE2F1-3transcripts highexpressionofSRC-3(13).Importantly,thegenomiclocifor
are coordinately elevated in human tumors with high HER2 HER2 and SRC-3 in the BT-474 cells used in this study are
expression(44,45).Moreover,E2F1isknowntobeinvolvedin amplified,whilemaintainingERaexpressionandestrogensen-
afeed-forward autoregulatoryloop withSRC-3(46,47).These sitivitydespiteSERMresistance(Fig.1;ref.3).Thisresistanceis
reports are in agreement with our findings and highlight the dependentonHER2expression,andcellsundergocaspase-medi-
intricaciesofaproliferativeandantiapoptoticE2F1transcription- atedapoptosisuponHER2depletionandtreatmentwithtamox-
al network downstream of oncogenic HER2 activity in breast ifenorraloxifene.Interestingly,BT-474cellsdevoidofHER2fail
cancer cells that has been underappreciated. More research is to respond to the proliferative estrogen signal, indicating sub-
neededtounderstandautoregulatorymechanismsandcross-talk stantialcoordinationbetweenthesetwopathwaysinbreastcancer
between all members of the E2F family. Moreover, it will be cells. Considering that 70% of breast tumors are ERþ, future
1472 CancerRes;76(6)March15,2016 CancerResearch
Downloaded from cancerres.aacrjournals.org on April 4, 2019. © 2016 American Association for Cancer Research.