Intracellular routing of β-catenin Hendriksen, J.V.R.B.

Intracellular routing of β-catenin

Hendriksen, J.V.R.B.

Citation

Hendriksen, J. V. R. B. (2008, June 19). Intracellular routing of β-catenin.

Retrieved from https://hdl.handle.net/1887/12965

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/12965

Note: To cite this publication please use the final published version (if

applicable).

Chapter 5

Rapid nuclear export of GFP- ß-catenin via a facilitated mechanism independent of CRM1

Manuscriptinpreparation

5

Rapid nuclear export of GFP- ß-catenin via a facilitated mechanism independent of CRM1

JolitaHendriksen,HellavanderVeldeandMaartenFornerod

DepartmentofTumorBiology,TheNetherlandsCancerInstitute,Amsterdam,TheNetherlands

Stimulation of cells with Wnt results in nuclear translocation of ß-catenin and transcrip- tional activation of target genes. Export of ß-catenin out of the nucleus could be an im- portant mechanism to end transcriptional activation. Currently, there are two mod- els for ß-catenin nuclear export. The first suggests that nuclear export of ß-catenin is mediated by co-transport of APC (or Axin) via the CRM1 nuclear export pathway. The sec- ond model predicts that ß-catenin mediates its own nucleocytoplasmic transport. We ana- lyzed the in vivo nuclear export kinetics of GFP-ß-catenin in Hek293 cells using Fluorescence Loss In Photo bleaching (FLIP). We show that GFP-ß-catenin nuclear export is very fast and exceeds the nuclear exit by diffusion of the small GFP molecule, suggesting that GFP- ß-catenin uses a fascilitated nuclear export mechanism. Furthermore, we find that nu- clear export of GFP-ß-catenin is not influenced by leptomycin B (LMB) treatment that inhibits CRM1-mediated export. We find that ß-catenin interacts with FG repeat nucleo- porins Nup62, Nup153, Nup214 and Nup358. We conclude that ß-catenin mediates its own nuclear export, supporting the idea that its localisation is regulated by retention.

TheWntfamilyofsecretedsignalingmolecules

regulatenumerousprocessesduringanimalde- velopmentandtissuehomeostasis.Deregulation

oftheWntpathwayislinkedtomanydiseases

including cancer (Nusse, 2005; Clevers, 2006).

In canonical Wnt signaling, Wnt signals viaß- catenintotransducethesignalfromtheplasma

membrane to the nucleus. In the nucleus,ß- catenininteractswithtranscriptionfactorsofthe

TCF/Lef family and the co-activators Pygopus

and Legless to regulate the expression of Wnt

targetgenes(Behrensetal.,1996;Molenaaret

al.,1996;vandeWeteringetal.,1997;Takemaru

andMoon,2000;Townsleyetal.,2004).

Nuclearimportandexportmaycontroltheavail- ability and thereby the activity ofß-catenin. ß- Catenin has a molecular weight of 92 kDa and

isthereforeexpectedtodependontheimportin/

exportin system for nucleo-cytoplasmic trans- port. However,ß-catenincontainsnorecogniz- ablenuclearlocalizationsignal(NLS),ornuclear

export signal (NES), which are required for re- ceptor-mediated nucleo-cytoplasmic transport

(MattajandEnglmeier,1998;GorlichandKutay).

ß-Catenincanshuttleinandoutofthenucleus

(Fagottoetal.,1998;Yokoyaetal.,1999;Prieve

and Waterman, 1999; Wiechens and Fagotto,

2001;Eleftheriouetal.,2001;SuhandGumbin- er,2003;Koikeetal.,2004).Micro-injectionand

semi-permeabilizedcellassayshaveshownthat

ß-catenin does not need Importin-ß, CRM1 or

RanGTP to exit the nucleus (Prieve and Water- man,1999;WiechensandFagotto,2001;Elefthe- riouetal.,2001).Structuralsimilaritiesbetween

theß-catenin armadillo repeats and the HEAT

repeatsofimportins,haveledtothehypothesis

thatß-cateninmediatesitsowntransportbyin- teractingdirectlytotheNPCproteins(Cingolani

etal.,1999;Fagottoetal.,1998;Yokoyaetal.,

1999; Wiechens and Fagotto, 2001). Nucleo- cytoplasmic transport ofß-cateninis,however,

still under debate. An alternative model speci- fies, that the APC tumour suppressor exports ß- cateninoutofthenucleus,resultinginthedegra- dationofß-catenininthecytoplasm.Thismodel

isbasedupontheobservationthatAPCshuttles

betweenthenucleusandcytoplasm,andthatß- cateninmimicsthelocalizationofAPCcontain- ing either active or inactive NESs (Henderson,

2000;Neufeldetal.,2000;Rosin-Arbesfeldetal.,

2000). Similar to APC, Axin and LZTS2 shuttle

in and out of the nucleus, and a role for these

proteinsinß-cateninnucleartransporthasbeen

suggested(CongandVarmus,2004;Thyssenet

al.,2006).

In this study, we examined the nuclear export

kinetics of GFP-ß-catenin in living cells, by

performing in vivo nuclear transport assays in

5

92

Hek293 cells. We expressed GFP-taggedß- catenin and performed Fluorescence Loss In

Photo bleaching experiments (FLIP). We found

thatnuclearexportofGFP-ß-cateninisfast,as

full recovery of steady state levels is observed

within 3 minutes. Furthermore, inhibition of the

CRM1 pathway does not affect GFP-ß-catenin

nuclear export. Furthermore, we show thatß- catenin interacts with FG repeat nucleoporins

andthatcouplingofGFPtoß-cateninincreases

thenuclearexportrateofGFP.Weconcludethat

ß-cateninexitsthenucleusonitsownbyinter- actingwithFGrepeatnucleoporins,andmayact

asatransportreceptoritself.

Results and discussion

GFP-ß-catenin is active in Wnt signaling Toanalyzethekineticsofß-catenininlivingcells,

weexpressedacarboxy-terminalfusionofGFP

toß-catenininHek293cells.Inapreviousstudy,

this GFP-ß-catenin fusion protein formed rod- likeaggregatesinthenucleiofCOSandMDCK

cells(Gianninietal.,2000).Weobservedthese

aggregatesaswellinthenucleiofHek293cells

but only when high levels of the plasmid were

transfected and/or after long expression times

(datanotshown).Itisconceivablethattheseß- cateninaggregatesareformedtoprotectthecell

from toxic levels of GFP-ß-catenin (Giannini et

al.,2000).Topreventtheformationoftheseag- gregates in the nucleus, we expressed GFP-ß- cateninatverylowlevelsforashortperiod(i.e.

100ngper3.10⁵cellsfor20hours).Underthese

conditions,theintracellulardistributionofGFP- ß-catenin is uniformly distributed between the

nucleusandcytoplasm,yetthenucleoliwereex- cluded(Fig2Aleftimageanddatanotshown).

To determine the expression of our GFP-ß- catenin fusion protein, we transfected Hek293

cells and analyzed cell lysates 24 hours after

transfectionbywesternblot.Ananti-GFPmono- clonalantibodydetectedonlyfulllengthGFP-ß- catenin(~115kDa)andnofreeGFP(~25kDa),

showingproperexpressionoftheGFP-ß-catenin

fusionprotein(Fig1A).Wenexttestedwhether

theGFP-ß-cateninfusionproteinisfunctionalin

Wnt signaling by performing luciferase assays

usingtheTCF-responsiveTOPreporter(Korinek

et al., 1997). In theß-catenin deficient cell line NCI-H28, GFP-ß-catenin activates the TOP re- porter but not the control FOP (Fig 1B). These

resultsshowthattheGFP-ß-cateninfusionpro- teinisproperlyexpressedandactiveintransac- tivation.

ß-Catenin exists in pools of different mobility

Toanalyzethekineticsofß-cateninnuclearex- portinlivingcells,weexpressedGFP-ß-catenin

at low levels in Hek293 cells and performed

Fluorescence Loss In Photo bleaching (FLIP)

experiments. In these experiments, the entire

nucleusorcytosolisbleachedafterwhichtime

laps images are taken to monitor the recovery

rate.Tocontroltheprecisionofthelaserbeam,

we performed similar bleach experiments on

formalin fixed cells expressing GFP-ß-catenin.

A well-defined spot was permanently bleached showingthattherewasnoleakageofthelaser

beamtoregionsoutsideourindicatedregionof

interest(datanotshown).Manynuclearproteins,

includingtranscriptionfactorsarehighlymobile

as measured by FRAP analysis (Fluorescence

RecoveryAfterPhotobleaching).Althoughthese

proteinsmoverapidly,theirmovementsarenot

asfastasafreediffusingmoleculesuchasGFP

(Misteli,2001).

Figure 1. GFP-ß-catenin is functional. A. Hek293

cellsweretransfectedwithGFP-ß-catenininincreas- ing amounts. After 24 hours, cell lysates were ana- lyzed by western blot using a monoclonal antibody

recognizing GFP. B. NCI-H28 cells were transfected

withtheTOP(darkbars)orthecontrolFOP(lightbars)

luciferase reporter together with pRL-CMV, in pres- enceorabsenceofGFP-ß-catenin.Celllysateswere

analyzed24hoursaftertransfection.Shownarenor- malized relative luciferase values corrected with the

pRL-CMV Renilla luciferase reporter. Bars represent

SEMsofindependentexperiments.

pEGFP-β-Catenin

1 2 3

GFP-β-Catenin

16 25 32 4362 17583

α GFP

0 1 2 3 4 5 6 7 8

+-

+

- -

-

-

+

+

-

NCI-H28 NCI-H28

relative luciferase activity

Figure 1 A

B

WenoticedthatthemobilityofGFP-ß-cateninin

boththenucleusandcytoplasmwasveryhigh,

as we could bleach either compartment effec- tively and uniformly by using 4 point bleaches

of 3 seconds, suggesting that during this short

bleaching time, all GFP-ß-catenin molecules in

thebleachedcompartmenthadpassedthelaser

beam.

Typically, after bleaching, fluorescence dropped inthebleachedcompartmentto~80%,whereas

thecompartmentthatwasnotbleachedlostonly

~20% (data not shown). To measure nuclear

export, we bleached the whole cytoplasm and

used time-lapse microscopy to monitor the re- covery.Thesubsequentincreaseincytoplasmic

fluorescence of GFP-ß-catenin was equal to

the decrease in nuclear fluorescence, suggest- ingthatthecytoplasmicincreasewastheresult

of nuclear transport events. From the recovery

curves,wecalculatedthetauvalue,i.e.thetime

needed for the fluorescence to recover to 63%

between its level after bleach and the plateau

Figure 2. Time-lapse confocal images and recovery curves of a representative FLIP experiment. A.Hek293cells

weretransfectedwithGFP-ß-cateninfor24hours,afterwhichFLIPexperimentswerecarriedoutinimagingmedium

at37°C.Twoinitialimages(T=-6sec)weretakenafterwhichthecytoplasmwasbleachedusing4x3secpoint

bleaches. T = 0 is the first image of the time lapse that was used to monitor the fluorescent recovery. Steady state levels were reached within 160 seconds. Recovery curves were corrected for bleach depth, normalized and fitted to a singleexponentialfunction.B.Timeconstants(tau)ofmultipleexperimentsandexperimentalconditionswereplotted

againstthenuclear/cytoplasmic(N/C)ratioofGFP-ß-cateninbeforebleaching.Observedfromthecytoplasm,taus

fromacytoplasmicbleacharecloseto40secondsindependentoftheN/Cratio(right).Fromanuclearview,tausare

higherandincreaseastheN/Cratiorises(left).

Figure 2

A

B

0 50 100 150 200

20 30 40 50

t(s) Cytoplasmic increase

fluorescence model f(t)

t= -6 sec t= 0 sec t= 160 sec

0 50 100 150 200

80 90 100 110

t(s) Nuclear decrease

fluorescence model f(t)(x)

0.8 1.0 1.2 1.4 1.6 1.8 2.0

05100150200250

Cytoplasmic bleach from nucleus

ratio N/C

Tau

Control LMBRanBP3 RNAi Energy depletion RNAi CD63

0.8 1.0 1.2 1.4 1.6 1.8 2.0

050100150200250

Cytoplasmic bleach from cytoplasm

ratio N/C

Tau

Control LMBRanBP3 RNAi Energy depletion RNAi CD63

0c N

5

94

fluorescence (Fig 2A).

Tomeasureexport,webleachedGFPinthecy- toplasmandmeasuredtherecoverytimeinboth

thenucleusandcytoplasm.Therecoverycurves

were different depending on whether measure- ments were taken from the nucleus or the cy- toplasm.Plottingthetauvaluesoftherecovery

curvesfromthenucleusagainstthenuclear/cy- toplasmicratio,revealsthattherearetwopools

of GFP-ß-catenin; a slow and a fast pool. The

slow pool dominates export measured from a

nuclear view and likely represents retention of

GFP-ß-catenininthenucleus,whereasthefast

pool represents free GFP-ß-catenin. Increasing

nuclear/cytoplasmicratio’scorrelatewithincreas- ingtauvalues,suggestingthattheslowpoolof

GFP-ß-cateninthatisseenfromthenucleus,is

dependentonexpressionlevels.However,when

export was measured from the cytoplasm, we

mostlydetectedthefastpoolwithnoeffectson

expressionlevels(Fig2B).Wethereforestudied

GFP-ß-cateninnuclearexportbymeasuringthe

fluorescent recovery in the cytoplasm.

TheslowpoolofGFP-ß-cateninthatweobserve

in the nucleus may reflect nuclear retention by theformationofrod-likeaggregateformationas

describedinGiannietal.(Gianninietal.,2000).

NuclearlevelsofGFP-ß-cateninindeedriseupon

higher expression levels and increased expres- sion time. Alternatively, the slow and fast pool

of GFP-ß-catenininthenucleusmayrepresent

differentmolecularformsofß-cateninthathave

been described earlier (Gottardi and Gumbiner,

2004).HigherlevelsofGFP-ß-cateninmayshift

theequilibriumtoamolecularformofß-catenin

that is more likely to interact with nuclear pro- teins, such as TCF/Lef and BCL9 resulting in

increased nuclear retention. Although retention

isalikelymechanismtoregulateß-catenin,the

different molecular conformations ofß-catenin

remainspeculativeasnoevidenceforthistheory

wasfoundinthermodynamicexperimentsonß- cateninanditsinteractingpartners(Choietal.,

2006).

Nuclear export of GFP-ß-catenin is fast and mediated by a facilitated transport pathway TomeasuretheexportrateofGFP-ß-cateninin

living cells, we bleached the cytoplasm of low

expressing Hek293 cells and monitored the re- covery time. Full recovery of fluorescence was observedwithin160seconds(2min40sec)after

bleaching, after which steady state levels were

reached. We measured an averaged tau of 45

seconds,indicatingthatnuclearexportofGFP- ß-catenininHekcellsisfast(Fig3).Thenuclear

exportratethatwemeasureforGFP-ß-catenin

inHek293cellsismorethantwiceasfastasde- scribedforYFP-ß-catenininCOScells(Towns- leyetal.,2004).Differencesinexpressionlevels

or cell types may account for this discrepancy.

TocomparethenuclearexportkineticsofGFP- ß-catenintothoseofotherexportsubstrates,we

comparedexportofGFP-ß-catenintothenuclear

exitoffreeGFP,thatcanfreelypassthroughthe

NPC by diffusion. Interestingly, the nuclear exit

ofthesmallGFPmoleculewas1.5timesslower

thanthatofGFP-ß-catenin,indicatingthatGFP- ß-catenin export is very efficient and mediated by a facilitated process. To confirm these findings in anothercellline,weusedthehumancoloncan- cercelllineSW480,whichshowsconstitutively

activeWntsignalingduetoatruncationinAPC.

NuclearexportofGFP-ß-catenininSW480was

asfastasinHek293cellsand,likewise,thenu- clearexitoffreeGFPwasmuchslowerthanthat

ofGFP-ß-catenin(Fig3).Theseresultsindicate

thatdifferencesintissueoriginandWntactiva- tionbetweenthesetwocelltypesdonotresultin

differentnucleocytoplasmictransportkineticsof

GFP-ß-catenin.Furthermore,thenuclearexport

rateofGFP-ß-cateninisfasterthanthenuclear

exitoffreeGFP,indicatingthatß-cateninexport

is mediated by an active transport mechanism

(Ben-EfraimandGerace,2001).

Nuclear export of ß-catenin does not depen- dent of the CRM1 export pathway

Proteinsofupto30kDacandiffusethroughthe

NPC,yetlargerproteinsgenerallypassbyafa- cilitatedtransportprocessthatrequiresRanand

recognition by transport receptors (Mattaj and

Englmeier, 1998; Gorlich and Kutay, 1999; Frey

andGorlich,2007).TheCRM1nuclearexportre- ceptorisresponsibleforaconsiderablefraction

ofproteinexportoutofthenucleusandrecog- nizesitscargobytheNES.CRM1bindsdirectly

to the NES in a RanGTP-dependent manner

(Fornerodetal.,1997;Kudoetal.,1997;Stade

etal.,1997;Fukudaetal.,1997;Ossareh-Nazari

etal.,1997).

To test the involvement of the CRM1 pathway

inß-catenin nuclear export, we measured the

nuclearexportofGFP-ß-catenininlivingcellsin

thepresenceandabsenceoftheCRM1inhibitor

LMB. The earlier described shuttling substrate

NLS-Rev-NES-GFPwasusedtocontrolforthe

activity of LMB (Henderson and Eleftheriou,

2000).Thissubstratelocalizestothecytosolin

theabsenceofLMBduetoitsNES,butitaccu- mulatesinnucleoliwhenLMBisadded.Hek293

cells were transfected with GFP-ß-catenin or

NLS-Rev-NES-GFP and either mock treated or

incubatedinimagingmediumcontaining50nM

LMB.Within30min,theNLS-Rev-NES-GFPre- portershiftedfromthecytoplasmtothenucleus,

indicatingthatLMBwasfullyactive(Fig4A).We

performed FLIP experiments on Hek293 cells

expressing GFP-ß-catenin, and cultured the

cellsforatleast30minutestoamaximumof1.5

hours under LMB conditions. As shown in Fig- ure4B,themediantauvalueforGFP-ß-catenin

decreasedfrom45secondswithoutLMBto34

seconds with LMB, but this drop was not sig- nificant (Fig 4B). This indicates that blocking the CRM1pathwaydoesnotaffectthenuclearex- portrateofGFP-ß-catenin.Thesteadystatelo- calizationofGFP-ß-cateninwasneitheraffected

byLMBtreatment,notevenaftertreatmentwith

LMBformorethan3hours(datanotshown).Our

results show that in living cells, GFP-ß-catenin

canexitthenucleusindependentlyoftheCRM1

nuclearexportpathway.Theseresultsareinline

with previous studies (Prieve and Waterman,

1999; Wiechens and Fagotto, 2001; Eleftheriou

etal.,2001)andarenotconsistentwitharoleof

APC,AxinorLTZS2inß-cateninnuclearexport

as these proteins all rely on the CRM1 nuclear

exportreceptortoexitthenucleus.

ß-Catenin associates with FG repeat nucleo- porins

AsGFP-ß-catenincanexitthenucleusindepen- dentlyoftheCRM1exportpathway,wesought

forevidencethatß-catenincantranslocatebyit- selfthroughthenuclearpore.Wethereforetest- edwhetherß-catenincaninteractwithFGrepeat

nucleoporins using immobilized GST-tagged X.

laevisß-catenin to pull down interacting pro- teinsfromX. leaviseggextracts,whicharehighly

concentrated in nucleoporins. We analyzed in- teracting proteins by western blot using mono- clonalantibody414thatrecognizesasubsetof

FGrepeatcontainingnucleoporins.Importantly,

bothfulllengthandthearmadillorepeatregion

ofß-catenin specifically interacted with FG re- peatnucleoporinsNup62,Nup153,Nup214and

Nup358.Furthermore,asmallamountofNup214

alsoboundfulllengthGST-ß-catenin(Figure5).

The interaction was not stimulated or weak- ened by non-hydrolysable forms of RanGTP

(data not shown), indicating that theß-catenin

interaction was not mediated by nuclear trans-

Figure 3. Nuclear export of GFP-ß-catenin is faster than that of GFP alone. Hek293(left)andSW480(right)

cellsweretransfectedwitheitherGFPaloneorwithGFP- ß-catenin.After24hours,FLIPexperimentswerecarried

outtomeasurenuclearexportusingcytoplasmicbleach- ing. Kinetics of the cytoplasmic fluorescence recovery were analyzed by timelapse microscopy, corrected for

bleach depth, normalized and fitted to a single exponen- tialfunction.Tausarerepresentedinaboxplot.Indicated

valuesaremedians.PvaluesareaccordingtoMann-Whit- neytests.

Figure 4. A. The CRM1 nuclear export pathway is effi- cientlyblockedbyadditionof50nMLMBfor30minutes.

Hek293cellsweretransfectedwiththeshuttlingsubstrate

NLS-Rev-NES-GFP(HendersonandEleftheriou,2000).At

24hoursaftertransfection,cellswereexposedtoeither

normal imaging medium or imaging medium containing

50nMLMBfor30min.Thereafter,thecellswereimaged

torecordthelocalizationofthereporter.B.Blockadeof

the CRM1 pathway does not influence the nuclear export kineticsofGFP-ß-catenin.Hek293cellsweretransfected

withGFP-ß-cateninandexportkineticsweredetermined

asdescribedinFigure3.

45.13 69.22

61.48

0 20 40 60 80 100 120 140 160

47.44 p= 0.006

p= 0.005

tau (seconds)

n=26 n=16 n=17

n=20 GFP-β−

CAT GFP GFP-β-

CAT GFP

Figure 3

Hek293 SW480

45.13

34.01

0 20 40 60 80 100

τin sec from cytoplasm

GFP-β-CAT GFP-β-CAT + LMB p= 0.08

Figure 4 A

B

no LMB 50 nM LMB

5

96

port receptors. These FG repeat nucleoporins

are commonly known to bind nuclear transport

receptorsincludingthoseoftheImportin-ßand

NTF2family(Moroianuetal.,1995;Baylissetal.,

2000; Fribourg et al., 2001; Conti and Izaurral- de,2001;VasuandForbes,2001;Baylissetal.,

2002).Asß-cateninsharesstructuralhomology

with Importin-ß,itisinterestingthatitinteracts

with these common nucleoporins. Our results

areinlinewithapreviousstudyinwhichrecom- binantX. laevisß-cateninwasshowntobindto

theyeastNup1(Fagottoetal.,1998).Ourdata

contradictastudy(StuhandGumbiner,2003)in

whichß-cateninwastestedforitsabilitytobind

totheFGnupsrecognizedbymAb414,i.e.the

samenupsastestedhere.Inthatstudy,therela- tivebindingofß-catenintoFGrepeatnupswas

comparedtothatofImportin-ß,butnointeraction

could be detected (Suh and Gumbiner, 2003).

The affinity of Importin-ßforFGrepeatnupsis

veryhighcomparedtoothertransportreceptors

(Ben-Efraim and Gerace, 2001). Therefore, it is

imaginable that the relative weak interaction of

ß-cateninwithFGnupsisbelowdetectionlimits when compared to the strong affinity of Impor- tin-ßforFGnups.Moreover,weakinteractionsof

transport receptors with FG repeats have been

suggested to play an important role for efficient translocation through the inner channel of the

NPC(Freyetal.,2006;FreyandGorlich,2007).

In line with our observation thatß-catenin can

exit the nucleus independent of CRM1, we find thatß-catenin interacts with FG repeat nups,

suggesting that ß-catenin mediates its own

translocationthroughtheNPC.

ß-Catenin nuclear export is insensitive to GSK3ß phosphorylation

OurdatasofarhaveshownthatGFP-ß-catenin

canexitthenucleusbyitself,mostlikelybyme- diatingitsownnuclearexportbyinteractingwith

the FG repeat nucleoporins. In recent years, it

hasbeenshownthatnotallß-cateninmolecules

are equally active in transcriptional activation.

Inhibition ofß-catenin degradation by abolish- ingproteosomaldegradationresultsinincreased

ß-cateninlevels,butnotinincreasedtranscrip- tion. However, blocking the phosphorylation of

ß-catenin on its N-terminal GSK3/CK1 phos- phorylation sites does increase transcription.

The use of the ABC antibody, which specifically recognizesß-cateninthatisnotphosphorylated

onitsN-terminus,hasbeenshowntoreportWnt

signalingactivitymorefaithfullythanantibodies

directedagainstanepitopeelsewhereinthepro- tein (Staal et al., 2002; Chan et al., 2002; Hen- driksenetal.,2005).

To test whether N-terminally dephosphorylated

Figure 5. ß-Catenin interacts with FG- rich nu- cleoporins Nup62, Nup153, Nup214 and Nup 358 in vitro. GST(lane2),GST-ß-cateninArm(lane3)and

GST-ß-catenin(lane4)wereusedtopulldowninter- actingnucleoporinsfromX. laeviseggextractsinthe

presence of 2 mM RanQ69L. Bound proteins were

analyzedbywesternblotusingmonoclonalantibody

414,recognizingFGrepeatnucleoporins.

Figure 6. Nuclear import and export kinetics of GFP-tagged ΔGSK3 mutant ß-catenin are not dif- ferent from wild type GFP-ß-catenin. Hek293cells

were transfected with either GFP-ß-catenin or GFP- ΔGSK3 ß-catenin. After 24 hours, import and export kinetics were measured. To measure import, the

nucleus was bleached followed by monitoring the

recovery of fluorescence in the nucleus. Transport ki- neticsweremeasured,analyzedandrepresentedas

inFigure3.

GFP-β-CAT import from nucleus

τin sec from cytoplasm

50 100 150

GFP-β-CAT export from cytoplasm GFP-mutant β-CAT import from nucleus

GFP-mutant β-CAT export from cytoplasm

Figure 6

Nup214 Nup153

Nup62 Nup358

Input GST

1 2 3 4 5

GST -beta-Cat

Arm

GST -beta-Cat Wt

175

83 68 47

Figure 5

ß-catenin exhibits distinct nucleocytoplasmic

shuttling behavior, we mimicked dephospho-ß- cateninbyusingtheΔGSK3mutantß-cateninin

whichall4GSK3phosphorylationsitesonitsN- terminusaremutatedtoalanine.Wetaggedthis

proteintoGFPandperformedFLIPexperiments

tostudyitsnucleartransportkinetics.Inaddition

tomeasuringexport,wealsomeasurednuclear

importofthissubstrateandcomparedittothe

importofwildtypeGFP-ß-catenin.Tomeasure

import, we bleached the nucleus and analyzed

the subsequent nuclear increase in fluorescence.

We did not measure any significant differences in tauvaluesofboththenuclearimportandexport

betweenΔGSK3mutantandwildtypeGFP-ß- catenin (Fig 6). These results suggest that our

ΔGSK3 mutant GFP-ß-catenin, which is more

activeinWntsignaling,entersthenucleusasfast

asGFP-ß-catenin.Furthermore,theresultsimply

that N-terminal phosphorylations on positions

33,37,41and45donotaffecttheabilityofß- catenintointeractwithFGrepeatstomediateits

nucleartransport.However,asalaninemutations

arenotthesameasnaturalnon-phosphorylated

residues,itisnotclearwhetherthismutantisa

good representative of active or dephospho-ß- catenin.Aslongastheexactnatureoftranscrip- tionally active or dephospho-ß-cateninremains

elusive we cannot be totally sure whether the

nuclearimportofthisproteinisenhancedorits

exportdecreased.Preliminaryexperimentsusing

LiCltoblockß-cateninphosphorylation,howev- er,didnotshowanydifferencesinnuclearexport

kineticseither(datanotshown).

In this study, we have confirmed that GFP-ß- catenin can exit the nucleus independently of

CRM1 and hence, independently of APC and

Axin (Prieve and Waterman, 1999; Wiechens

andFagotto,2001;Eleftheriouetal.,2001).We

suggestthatß-catenininteractswithFGrepeat

nucleoporinstomediateitsownnuclearexport,

which explains why GFP-ß-catenin nuclear ex- port is very efficient even during LMB treatment.

Moreover, coupling GFP toß-cateninincreases

thenuclearexportrateofGFP,suggestingthat

ß-catenin not only exits the nucleus similar to

thetransportreceptors,butmayalsocarryalong

substrates. Indeed, it has recently been shown

that Lef1 can function as a natural nuclear im- portsubstrateforß-catenin(AsallyandYoneda,

2005). It is fascinating that not only the mobil- ity of GFP-ß-catenin within the cytoplasm and

nucleusisveryhigh,butalsothatGFP-ß-catenin

iscapableofshuttlingquicklybetweenthesetwo

compartments. Therefore,ß-cateninseemsca- pableofrelocalizingquicklyinthecelltomeetits

binding partners. A previously proposed model

of retention seems applicable here (Tolwinski

and Wieschaus, 2001). In this model, the bind- ingpartnersofß-cateninregulateitssubcellular

localizationandthereforeitsactivity.E-cadherin

bindsß-cateninattheplasmamembrane,APC,

Axin and Dvl in the cytoplasm, and BCL9 and

TCF in the nucleus. Therefore, Wnt signaling

could regulate the availability of these pools of

ß-cateninallowingthecelltorespondquicklyto

theextracellularWntsignal.

Materials and Methods

Plasmids

GST-ß-catenin and GST-ARM ((Wiechens and

Fagotto, 2001)), GFP-ß-catenin (Giannini et al.,

2000),pSUPER,pSUPER-RanBP3,Top-Tkand

Top-Tk, pRL-CMV Renilla plasmids were previ- ouslydescribed(Hendriksenetal.,2005).pEG- FP-N1wasobtainedfromClontech.

In vitro binding studies

GST-ß-catenin binding studies were performed

asdescribedbefore(Hendriksenetal.,2005).

Luciferase reporter assays

NCI-H28 cells were cultured in 12-wells plates

and transfected with 200 ng Top/Fop-Tk, 1 ng

pRL-CMV Renilla and 25 and 100 ng GFP-ß- catenin. Luciferase activity was measured 48 h

after transfection using the Dual-luciferase re- porterassaysystem(Promega).

Western blotting

To detect GFP-ß-catenin25μgcelllysatewas

analyzed by SDS-PAGE. Western blotting was

performed as described before (Hendriksen et

al.,2005).

Cell culture, transfection and photo bleaching experiments

AllcelllineswereculturedinDMEMsupplement- edwith10%fetalcalfserumandpenicillin/strep- tomycin(Gibco-BRL)andweretransfectedusing

Fugene-6(Roche)orLipfofectamine(Invitrogen)

as instructed by the supplier. For FLIP experi- ments,3.105cellsweregrownon40mmcover- slipsandtransfectedwith100ngpEGFP,GFP- ß-cateninorΔGSK3-ß-catenin,upto500ngin

totalusingpcDNA3asstufferDNA.FRAPexperi- mentswereperformed24hoursaftertransfection

usingalivecellchamberat37°Cinbicarbonate- bufferedsaline(containing:140mMNaCl,5mM

KCl,1mMMgCl₂,1mMCaCl₂,23mMNaHCO₃,

10mMglucoseand10mMHEPESatpH7.2).

5

98

Photo bleaching and imaging was done on a

confocal laser scanning microscope (SP2 TCS

AOBS,Leica).Cellsexpressinglowamountsof

tagged-ß-cateninwereselectedandimagedat

4% laser power using the 488 laser line of the

20mWargonlaserandbleachedat100%laser

power. The cytoplasm was bleached uniformly

by using 4 point-bleaches of 3 seconds each.

Using Leica time lapse, 2 images were taken

beforebleachusinganintervalof1.68seconds,

after bleach 50 frames were imaged using a 3

secondinterval.Averagedintensitiesofregions

ofinterestweremeasuredusingImageJandre- covery curves and taus were determined using

Rsoftware.LMBwasusedat50nMfor30min

upto1.5hours(Wolffetal.,1997).Tocontrolfor

LMBactivity,cellsweretransfectedwith1μgof

NLS-Rev-NES-GFP per 40 mm coverslip (Hen- dersonandEleftheriou,2000).

Data analyis

Statistical analysis was done using the R soft- ware package (R Development Core Team,

2005). Nuclear and cytoplasmic decay curves

were fitted to a single exponential function I = I0 -A*(1-exp(-t/tau)),usingtheoptim()function

in R. Quality of the fit was assessed by “good- ness of fit” (R2) = 1- (sum of squared residuals) /(sumofsquareddifferencesfrommean),where

1 equals a perfect fit and 0 no fit. A small number of fits with an R2 of < 0.8 were not used in further analysis. The low quality of these fits could be tracedbacktomovementofcellsduringrecord- ingorverylowsignaltonoise.ThemeanR2of

dataexcludingtheseoutlierswas0.97.

References

Asally,M.,andY.Yoneda.2005.Beta-catenincanact

asanuclearimportreceptorforitspartnertranscription

factor, lymphocyte enhancer factor-1 (lef-1). Exp Cell

Res.308:357-63.

Bayliss,R.,S.W.Leung,R.P.Baker,B.B.Quimby,A.H.

Corbett,andM.Stewart.2002.Structuralbasisforthe

interaction between NTF2 and nucleoporin FxFG re- peats.EmboJ.21:2843-53.

Bayliss,R.,T.Littlewood,andM.Stewart.2000.Struc- turalbasisfortheinteractionbetweenFxFGnucleopo- rin repeats and importin-beta in nuclear trafficking. Cell.

102:99-108.

Behrens,J.,J.P.vonKries,M.Kuhl,L.Bruhn,D.Wedlich,

R.Grosschedl,andW.Birchmeier.1996.Functionalin- teraction of beta-catenin with the transcription factor

LEF-1.Nature.382:638-42.

Ben-Efraim, I., and L. Gerace. 2001. Gradient of in-

creasing affinity of importin beta for nucleoporins along thepathwayofnuclearimport.JCellBiol.152:411-7.

Chan,T.A.,Z.Wang,L.H.Dang,B.Vogelstein,andK.W.

Kinzler.2002.TargetedinactivationofCTNNB1reveals

unexpectedeffectsofbeta-cateninmutation.ProcNatl

AcadSciUSA.99:8265-70.

Choi,H.J.,A.H.Huber,andW.I.Weis.2006.Thermody- namicsofbeta-catenin-ligandinteractions:therolesof

the N- and C-terminal tails in modulating binding affin- ity.JBiolChem.281:1027-38.

Cingolani,G.,C.Petosa,K.Weis,andC.W.Muller.1999.

Structureofimportin-betaboundtotheIBBdomainof

importin-alpha.Nature.399:221-9.

Clevers,H.2006.Wnt/beta-cateninsignalingindevel- opmentanddisease.Cell.127:469-80.

Cong, F., and H. Varmus. 2004. Nuclear-cytoplasmic

shuttling of Axin regulates subcellular localization of

beta-catenin.ProcNatlAcadSciUSA.101:2882-7.

Conti, E., and E. Izaurralde. 2001. Nucleocytoplasmic

transport enters the atomic age. Curr Opin Cell Biol.

13:310-9.

Eleftheriou,A.,M.Yoshida,andB.R.Henderson.2001.

Nuclearexportofhumanbeta-catenincanoccurinde- pendentofCRM1andtheadenomatouspolyposiscoli

tumorsuppressor.JBiolChem.276:25883-8.

Fagotto,F.,U.Gluck,andB.M.Gumbiner.1998.Nucle- ar localization signal-independent and importin/karyo- pherin-independent nuclear import of beta-catenin.

CurrBiol.8:181-90.

Fornerod, M., M. Ohno, M. Yoshida, and I.W. Mattaj.

1997.CRM1isanexportreceptorforleucine-richnu- clearexportsignals.Cell.90:1051-60.

Frey, S., and D. Gorlich. 2007. A saturated FG-repeat

hydrogel can reproduce the permeability properties of

nuclearporecomplexes.Cell.130:512-23.

Frey, S., R.P. Richter, and D. Gorlich. 2006. FG-rich

repeats of nuclear pore proteins form a three-dimen- sionalmeshworkwithhydrogel-likeproperties.Science.

314:815-7.

Fribourg, S., I.C. Braun, E. Izaurralde, and E. Conti.

2001.Structuralbasisfortherecognitionofanucleopo- rinFGrepeatbytheNTF2-likedomainoftheTAP/p15

mRNAnuclearexportfactor.MolCell.8:645-56.

Fukuda, M., S. Asano, T. Nakamura, M. Adachi, M.

Yoshida,M.Yanagida,andE.Nishida.1997.CRM1is

responsible for intracellular transport mediated by the

nuclearexportsignal.Nature.390:308-11.

Giannini, A.L., M.M. Vivanco, and R.M. Kypta. 2000.

Analysis of beta-catenin aggregation and localization

using GFP fusion proteins: nuclear import of alpha-

cateninbythebeta-catenin/Tcfcomplex.ExpCellRes.

255:207-20.

Gorlich,D.,andU.Kutay.1999.Transportbetweenthe

cellnucleusandthecytoplasm.AnnuRevCellDevBiol.

15:607-60.

Gottardi,C.J.,andB.M.Gumbiner.2004.Distinctmo- lecularformsofbeta-cateninaretargetedtoadhesiveor

transcriptionalcomplexes.JCellBiol.167:339-49.

Henderson,B.R.2000.Nuclear-cytoplasmicshuttlingof

APCregulatesbeta-cateninsubcellularlocalizationand

turnover.NatCellBiol.2:653-60.

Henderson,B.R.,andA.Eleftheriou.2000.Acompari- son of the activity, sequence specificity, and CRM1-de- pendence of different nuclear export signals. Exp Cell

Res.256:213-24.

Hendriksen, J., F. Fagotto, H. van der Velde, M. van

Schie,J.Noordermeer,andM.Fornerod.2005.RanBP3

enhancesnuclearexportofactive(beta)-catenininde- pendentlyofCRM1.JCellBiol.171:785-97.

Koike,M.,S.Kose,M.Furuta,N.Taniguchi,F.Yokoya,

Y.Yoneda,andN.Imamoto.2004.beta-Cateninshows

anoverlappingsequencerequirementbutdistinctmo- lecularinteractionsforitsbidirectionalpassagethrough

nuclearpores.JBiolChem.279:34038-47.

Korinek, V., N. Barker, P.J. Morin, D. van Wichen, R.

deWeger,K.W.Kinzler,B.Vogelstein,andH.Clevers.

1997.Constitutivetranscriptionalactivationbyabeta- catenin-Tcf complex in APC-/- colon carcinoma. Sci- ence.275:1784-7.

Kudo,N.,S.Khochbin,K.Nishi,K.Kitano,M.Yanagida,

M. Yoshida, and S. Horinouchi. 1997. Molecular clon- ingandcellcycle-dependentexpressionofmammalian

CRM1,aproteininvolvedinnuclearexportofproteins.

JBiolChem.272:29742-51.

Mattaj, I.W., and L. Englmeier. 1998. Nucleocytoplas- mic transport: the soluble phase. Annu Rev Biochem.

67:265-306.

Misteli, T. 2001. Protein dynamics: implications for

nuclear architecture and gene expression. Science.

291:843-7.

Molenaar,M.,M.vandeWetering,M.Oosterwegel,J.

Peterson-Maduro, S. Godsave, V. Korinek, J. Roose,

O.Destree,andH.Clevers.1996.XTcf-3transcription

factormediatesbeta-catenin-inducedaxisformationin

Xenopusembryos.Cell.86:391-9.

Moroianu,J.,M.Hijikata,G.Blobel,andA.Radu.1995.

Mammaliankaryopherinalpha1betaandalpha2beta

heterodimers:alpha1oralpha2subunitbindsnuclear

localizationsignalandbetasubunitinteractswithpep- tiderepeat-containingnucleoporins.ProcNatlAcadSci

USA.92:6532-6.

Neufeld, K.L., F. Zhang, B.R. Cullen, and R.L. White.

2000. APC-mediated downregulation of beta-catenin

activityinvolvesnuclearsequestrationandnuclearex- port.EMBORep.1:519-23.

Nusse,R.2005.Wntsignalingindiseaseandindevel- opment.CellRes.15:28-32.

Ossareh-Nazari, B., F. Bachelerie, and C. Dargemont.

1997.EvidenceforaroleofCRM1insignal-mediated

nuclearproteinexport.Science.278:141-4.

Prieve,M.G.,andM.L.Waterman.1999.Nuclearlocal- izationandformationofbeta-catenin-lymphoidenhanc- er factor 1 complexes are not sufficient for activation of geneexpression.MolCellBiol.19:4503-15.

Rosin-Arbesfeld, R., F. Townsley, and M. Bienz. 2000.

TheAPCtumoursuppressorhasanuclearexportfunc- tion.Nature.406:1009-12.

Staal,F.J.,M.NoortMv,G.J.Strous,andH.C.Clevers.

2002.WntsignalsaretransmittedthroughN-terminally

dephosphorylatedbeta-catenin.EMBORep.3:63-8.

Stade,K.,C.S.Ford,C.Guthrie,andK.Weis.1997.Ex- portin 1 (Crm1p) is an essential nuclear export factor.

Cell.90:1041-50.

Suh, E.K., and B.M. Gumbiner. 2003. Translocation of

beta-catenin into the nucleus independent of interac- tionswithFG-richnucleoporins.ExpCellRes.290:447- 56.

Takemaru,K.I.,andR.T.Moon.2000.Thetranscriptional

coactivatorCBPinteractswithbeta-catenintoactivate

geneexpression.JCellBiol.149:249-54.

Thyssen,G.,T.H.Li,L.Lehmann,M.Zhuo,M.Sharma,

andZ.Sun.2006.LZTS2isanovelbeta-catenin-inter- actingproteinandregulatesthenuclearexportofbeta- catenin.MolCellBiol.26:8857-67.

Townsley,F.M.,A.Cliffe,andM.Bienz.2004.Pygopus

andLeglesstargetArmadillo/beta-catenintothenucle- us to enable its transcriptional co-activator function.

NatCellBiol.6:626-33.

van de Wetering, M., R. Cavallo, D. Dooijes, M. van

Beest, J. van Es, J. Loureiro, A. Ypma, D. Hursh, T.

Jones,A.Bejsovec,M.Peifer,M.Mortin,andH.Clev- ers.1997.Armadillocoactivatestranscriptiondrivenby

the product of the Drosophila segment polarity gene

dTCF.Cell.88:789-99.

Vasu,S.K.,andD.J.Forbes.2001.Nuclearporesand

nuclearassembly.CurrOpinCellBiol.13:363-75.

Wiechens,N.,andF.Fagotto.2001.CRM1-andRan- independentnuclearexportofbeta-catenin.CurrBiol.

11:18-27.

Wolff,B.Sanglier,J.J.Wang,Y.1997.LeptomycinBis

aninhibitorofnuclearexport:inhibitionofnucleo-cyto-

5

100

plasmic translocation of the human immunodeficiency virus type 1 (HIV-1) Rev protein and Rev-dependent

mRNA.ChemBiol.4:139-47.

Yokoya, F., N. Imamoto, T. Tachibana, and Y. Yoneda.

1999.beta-catenincanbetransportedintothenucleus

inaRan-unassistedmanner.MolBiolCell.10:1119-31.