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 2

RanBP3 enhances nuclear export of active ß-catenin indepen- dently of CRM1

JCB,Vol.171,No.5,p785-797December5,2005

2

RanBP3 enhances nuclear export of active ß-catenin independently of CRM1

JolitaHendriksen

,FrancoisFagotto

,HellavanderVelde

,MartijnvanSchie

,Jas- prienNoordermeer

,andMaartenFornerod

1DepartmentofTumorBiology,NetherlandsCancerInstitute,1066CXAmsterdam,Netherlands 2DepartmentofBiology,McGillUniversity,Montreal,Quebec,CanadaH3A2T5

3DepartmentofMolecularCellBiology,LeidenUniversityMedicalCenter,2333ALLeiden,Netherlands

ß-Catenin is the nuclear effector of the Wnt signaling cascade. The mechanism by which nucle- ar activity of ß-catenin is regulated is not well defined. Therefore, we used the nuclear marker RanGTP to screen for novel nuclear ß-catenin binding proteins. We identified a cofactor of chro- mosome region maintenance 1 (CRM1)–mediated nuclear export, Ran binding protein 3 (RanBP3), as a novel ß-catenin–interacting protein that binds directly to ß-catenin in a RanGTP-stimulated manner. RanBP3 inhibits ß-catenin–mediated transcriptional activation in both Wnt1- and ß- catenin–stimulated human cells. In X. laevis embryos, RanBP3 interferes with ß-catenin–induced dorsoventral axis formation. Furthermore, RanBP3 depletion stimulates the Wnt pathway in both human cells and D. melanogaster embryos. In human cells, this is accompanied by an increase of dephosphorylated ß-catenin in the nucleus. Conversely, overexpression of RanBP3 leads to a shift of active ß-catenin toward the cytoplasm. Modulation of ß-catenin activity and localization by RanBP3 is independent of adenomatous polyposis coli protein and CRM1. We conclude that RanBP3 is a direct export enhancer for ß-catenin, independent of its role as a CRM1-associated nuclear export cofactor.

TheWntsignalingpathwayregulatesavarietyof

processesduringhomeostasisanddevelopment,

includingcellularproliferation,cellfatedecision,

axisformation,andorgandevelopment(Nusse,

1999). Deregulation of the pathway is implicat- edinmanyhumancancers(Polakis,2000).The

key effector protein of the Wnt pathway is the

transcriptional activator ß-catenin. Cytoplasmic

ß-catenin is efficiently trapped in a multiprotein complexcontainingadenomatouspolyposiscoli

(APC;Grodenetal.,1991;Kinzleretal.,1991),

Axin(Zengetal.,1997;Behrensetal.,1998),and

glycogensynthasekinase3ß(GSK3ß;Heetal.,

1995).IntheabsenceofaWntsignal,thiscom- plexrapidlyphosphorylatesß-catenin,targeting

itfordegradation(Hartetal.,1998;Ikedaetal.,

1998; Itoh et al., 1998; Sakanaka et al., 1998).

Wnt binding to the Frizzled/LRP (low-density

lipoprotein receptor–related protein) receptors

results in inhibition of the APC–Axin–GSK3ß

complex by activation of Dishevelled (Boutros

and Mlodzik, 1999; Wharton, 2003) and by re- cruitment of Axin to the plasma membrane by

LRP (Mao et al., 2001; Tolwinski et al., 2003).

Thisresultsinanincreaseinnonphosphorylated

ß-cateninthatformsactivetranscriptionalcom- plexesinthenucleuswithTcellfactor(TCF)/lym-

phocyteenhancerbindingfactor(LEF)transcrip- tionfactors(Behrensetal.,1996;Molenaaretal.,

1996;Staaletal.,2002).

Nuclearactivityofß-cateninisregulatedbysev- eralmechanisms.IntheabsenceofaWntsignal,

TCF proteins occupy and repress promoters of

their target genes by recruiting repressor pro- teinslikeGroucho,CtBP(COOH-terminalbind- ing protein), and histone deacetylases (Cavallo

etal.,1998;Levanonetal.,1998;Rooseetal.,

1998; Waltzer and Bienz, 1998; Brannon et al.,

1999;Chenetal.,1999).Interactionofß-catenin

with TCF/LEF transcription factors results in

activation of these genes. BCL-9/Legless and

Pygopushavebeenshowntobeessentialcom- ponentsoftheß-catenin–TCFtranscriptioncom- plexes(Krampsetal.,2002;Parkeretal.,2002;

Thompsonetal.,2002).ß-Cateninalsointeracts

with chromatin remodeling and histone modifica- tionproteinssuchasBrg1(Brahma-relatedgene

1)andCBP(CREBbindingprotein)/p300topro- motetargetgeneactivation(HechtandKemler,

2000;TakemaruandMoon,2000;Barkeretal.,

2001). Furthermore, ICAT (inhibitor of ß-catenin

and TCF4) and Chibby are identified as nuclear proteinsthatrepressWntsignalingbycompet-

2

ingwithTCFforbindingtoß-catenin(Tagoetal.,

2000;Takemaruetal.,2003).

Inthisstudy,weaimedtoidentifynewmodulators

ofß-catenininthenucleus.Weusedthenuclear

markerRanGTPtoselectfornuclearfactorsthat

directly bind ß-catenin and identified Ran bind- ing protein 3 (RanBP3). We show that RanBP3

inhibits ß-catenin–TCF4–mediated transactiva- tioninhumancelllinesbyrelocalizationofactive

ß-cateninfromthenucleustothecytoplasm.In

addition,weshowthatRanBP3causesventral- ization and inhibits ß-catenin–induced double

axis formation in X. laevis embryos. Loss of D.

melanogasterRanBP3resultsincuticledefects

and expands the Engrailed protein expression

domain.WeconcludethatRanBP3functionsas

anoveltypeofinhibitorofß-cateninandidentify

itsgeneasacandidatehumantumorsuppressor

in the commonly deleted chromosomal region

19p13.3.

Results

RanBP3 interacts directly with ß-catenin in a RanGTP-stimulated way

Tostudytheinteractionbetweenß-cateninand

nuclear transport factors, we used GST-tagged

ß-catenintopulldowninteractingproteinsfrom

X.laeviseggextracts.Interactingproteinswere

initiallyanalyzedbyWesternblotusingmAb414,

which recognizes a phenylalanine glycine (FG)–

richepitopepresentinmultiplenucleoporins.FG

repeat–containingnucleoporinsNup62,Nup153,

and Nup358 were specifically bound by full- length ß-catenin and by the central armadillo

(ARM)repeatregion(Chapter5,Fig5).Interest- ingly,wefoundastronginteractionbetweenß- catenin and two unknown proteins of ~80 and

90 kD that were recognized by mAb414 (Fig. 1

A,lanes3and4).Theseproteinsinteractedwith

full-lengthß-cateninandtoalesserextentwith

theARMrepeats(ARM1–12).ThemAb414reac- tivityindicatedthatthesetwoproteinscontained

FGrepeats.TwoisoformsofRanBP3stoodout

as possible candidates for these two unknown

proteins because they contain FG repeats and

havethecorrectsizes.Indeed,recombinanthu- manRanBP3-acomigratedwiththep90protein

andwasrecognizedbymAb414(Fig.1A,lane

5). To confirm that RanBP3 was one of these new ß-catenin–interacting proteins, we repeated the

pull-down experiment using HeLa nuclear ex- tractsandanmAbrecognizinghumanRanBP3.

ThebisoformofRanBP3wasmoreabundantin

HeLa nuclear extracts and copurified with GST- tagged full-length and the ARM repeats of ß-

catenin(Fig.1B).Tomimicnuclearconditions,2

µMofanonhydrolysablemutantofthesmallGT- PaseRan(RanQ69L-GTP)wasadded,resulting

in increased interaction between ß-catenin and

RanBP3(Fig.1B,lanes2and4).Inthepresence

of RanQ69L-GTP, the less abundant a isoform

of human RanBP3 also bound to full-length ß- catenin(Fig.1B,lane2).

To investigate whether the binding between ß- catenin and RanBP3 was direct, we performed

pull down assays with GST-taggedß-catenin

and recombinant RanBP3. Human RanBP3-b

interacted directly with GST-ß-catenin with an

optimum at 0.5 µM RanBP3 (Fig. 1C, lane 3).

These binding characteristics resemble the in- teraction of RanBP3 with CRM1, which shows

optimalbindingat0.2µMRanBP3(Englmeieret

al.,2001).Furthermore,weusedaRanBP3mu- tantthatcannotbindtoRanGTPduetoapoint

mutationinitsRanGTPbindingdomain(RanBP3

“wv”mutant(Englmeieretal.,2001).Thismutant

interacted only very weakly with ß-catenin and

lostitsabilitytobindatanoptimumconcentra- tion(Fig.1C,lanes5-7).Thesedatasuggestthat

RanGTP increases the affinity of RanBP3 for ß- catenin. To confirm the RanGTP dependency, RanBP3wasboundtoß-catenincolumnsatthe

optimalconcentrationof0.5µMinthepresence

ofRanGTPandelutedeitherintheabsenceor

presenceoftherecombinantRancofactorsRan- BP1andRanGAP(lanes8-11).Whilevirtuallyno

RanBP3 was eluted with buffer only, significant amountsweredetectedafterelutioninthepres- enceofeither0.5µMRanBP1,0.2µMRanGAP

oracombinationofthese.

RanBP3 inhibits transcription of a TCF re- sponsive reporter

Wnt signaling ultimately results in the stabiliza- tionofß-cateninthatformsactivetranscriptional

regulationcomplexeswithtranscriptionfactorsof

theTCF/Leffamily.Awell-establishedfunctional

read-outofWntsignalingmakesuseofTCFre- sponsiveluciferasereporterconstructs(Korinek

etal.,1997).Totestthefunctionalrelevanceof

theinteractionbetweenß-cateninandRanBP3,

we transfected human embryonic kidney cells

(HEK293) with reporter constructs that contain

either three optimal TCF binding sites (TOP) or

three mutated binding sites (FOP). Transfection

ofaWnt1plasmidresultedinastrongactivation

of the TOP reporter but not of the FOP control

(Fig.2B).Co-transfectionofincreasingamounts

ofRanBP3repressedWnt1/ß-catenintransacti- vation dose-dependently (Fig 2B). A mutant of

RanBP3 that cannot interact with RanGTP and

Figure 1. Identification of RanBP3 as an interaction partner of ß-Catenin. A.Pulldownexperimentusingimmo- bilizedGST(lane2),GST-taggedß-CateninARMrepeats1-12(lane3)andfulllengthß-Catenin(lane4)incubated

withXenopuseggextract(input,lane1).BoundproteinswereanalyzedbyWesternblotusingMab414recognizinga

subset of nucleoporins. Two unknown proteins p80 and p90 are marked with an arrow. B. Identification of p80 and p90 asthebandaisoformsofRanBP3.PulldownexperimentasinA,incubatedwithHeLanuclearextractsandanalyzed

usingRanBP3antibody.C.RanBP3bindsdirectlytoß-Catenin.GST-taggedfulllengthß-Catenin(lanes2-11)was

incubatedwith2µMRanGTPand0.2µM(lanes2and5),0.5µM(lanes3,6,8-11)or2.0µM(lanes4and7)wildtype

(lanes2-4,8-11),or“wv”mutant(lanes5-7)RanBP3-b.Boundproteinswereelutedasindicatedabovethelanesand

visualizedwithsilver(lanes1-7)orCoomassie(lanes8-11)staining.

binds ß-catenin with less affinity (Fig. 1C) was lessactivethanwildtypeRanBP3(Fig.2B).To

investigate whether RanBP3 inhibits Wnt sig- naling downstream or upstream of ß-catenin,

wemimickedWntsignalinginHEK293cellsby

expressing ß-catenin. RanBP3 could still specifi- cally inhibit activation of the TOP reporter (Fig.

2C),whiletheRanBP3“wv”mutantwaslessef- fective. These experiments show that RanBP3

inhibits TCF-dependent transcription by acting

eitheronß-cateninitself,oronregulatorsdown- stream of ß-catenin. We confirmed that the ex- pressionlevelsofourwildtypeand“wv”mutant

RanBP3constructswereequalbyanalyzingcell

lysatesfromtransfectedHEK293cellsonWest- ernblot(Fig.2A).

The interaction of recombinant ß-catenin with

RanBP3(Fig.1C)impliesthatRanBP3canbind

N-terminallyunphosphorylatedß-catenin,which

isthoughttobethesignaling-competentformof

the protein. To test whether this is the case in vivo,weusedaß-cateninmutantthatcontains

alanines in all four N-terminal GSK3ß phos- phorylation sites (ß-cateninΔGSK3ß, Barth et al.,1999),andthereforeisconstitutivelyactive.

This mutant stimulated expression of the TCF

reporterto2-3foldhigherlevelsthanwildtype

ß-catenin (data not shown). Co-expression of

wild-type RanBP3 lead to a significant reduction in transactivation by ß-cateninΔGSK3ß (Fig. 2D).

Again,theRanBP3RanGTP-bindingmutantwas

less able to repress ß-cateninΔGSK3ß mediated µM BP3

0.2 0.5 2.0 0.2 0.5 2.0 wild type WV mutant

500 mM NaCl elution

inputRanBP3

C

RanBP1 RanBP1+ RanGAP

RanGAP

Control

Specific elution

1 2 3 4 5 6 7 8 9 10 11

B

1 2 3 4 5 6

1 2 3 4 5 6

RanBP3-a RanBP3-b 62 -

83 -

RanQ69L

-

-

-

GST-β

Cat wt GST-β

Cat Arm GST Input

A

1 2 3 4 5

p80 p70

Input GST GSTβ-CatArm GST β-Catwt Rec.RanBP3-a

pull-down Hendriksen et al., Figure 1

2

Wnt1

wt RanBP3

TOPFOP +- ₊- ⁺- -_{+ +}- _{+ + + +}- - - - _{+ + + + +}- - - - -

+ + + +

+ + + + + + +

- +

-

mut RanBP3 - -

- -

RanBP3

300200100 50 25 0

2 4 6 8 10

C B

A

β-Actin RanBP3 mutRanBP3

50 100

200

+ + + + +300 + + + +

-

0 2 4 6 8 10 12 14

TOPFOPTOP TOP TOP FOP

+ + +

- +

-- - - - ⁺ -

- - - - - ⁺

Hendriksen et al., Figure 2

30020010050 25

D

TOPFOP RanBP3

+ +

+ +

+ + +

+ + -

- - ⁺ - - - - - - - -

+

+ + + +

+

+ + +

- +

- - -

- -

100 200 0

1 2 3 4 5 6 7

+-

+ 200

E

β-Gal

wtBP3mutBP3

RanBP3 mut RanBP3 wt β-Catenin wt RanBP3

100

RanBP3 c-Myc β-Actin

∆GSK3-Cat

relativeluciferaseactivity relativeluciferaseactivity

relativeluciferaseactivity

wt RanBP3 mut RanBP3

1 2 3

Figure 2. Expression of RanBP3 inhibits ß-Catenin/TCF mediated transcriptional activation.A.Wildtypeand

“wv”mutantRanBP3areexpressedatequallevels.HEK293cellsweretransfectedwithindicatedconstructs(ng)

andlysateswereanalyzed48hrspost-transfectionbyWesternblotwithindicatedantibodies.B.RanBP3represses

Wnt-1inducedß-Catenin/TCFmediatedtranscriptionalactivationdose-dependently.HEK293cellsweretransfected

withTOP(blackbars)orthecontrolFOP(greybars),Wnt1,anddecreasingamountsofRanBP3wtor“wv”mutant

asindicated(ng)andluciferaseactivitywasmeasuredafter48hours.CandD.RanBP3repressestranscriptional

activation induced by wild type ß-Catenin (C) or ΔGSK3-ß-Catenin (D). HEK293 cells were transfected with indicated constructsandluciferaseactivitywasmeasured48hrsaftertransfection.Inallexperiments,normalizedrelativelucif- erasevaluesareshownascorrectedwithpRL-CMVRenilla.Barsrepresentstandarderrorsofmeansofindependent

experimentsE.RanBP3inhibitstheexpressionoftheendogenousWnttargetc-Myc.HCT116coloncarcinomacells

expressing Δ45-ß-Catenin were transfected with GFP and ß-galactosidase, RanBP3 wt or mutant plasmids. 2 days after transfection, GFP-positive cells were sorted using flow cytometry, lysed in sample buffer and analyzed by West- ernblotusingindicatedantibodies.

transactivation.

To address whether RanBP3 could also affect

expression of endogenous target genes of ß- catenin/TCF, we expressed RanBP3 in human

coloncarcinomacelllineHCT116.Thiscellline

harbors an activating mutation in ß-catenin (Δ45

catenin)andthereforeexpressesincreasedlev- els of the target gene c-Myc (He et al., 1998).

ExpressionofwtRanBP3decreasedc-Mycpro- teins levels compared to control cells (Fig 2E,

lanes2and3).Althoughexpressedinhigherlev- els,thewvmutantRanBP3waslesscapableof

Figure 3. Reduction of RanBP3 by RNAi results in increased ß-Catenin/TCF mediated transcription activation.

A.WesternblotshowingthatdifferentshorthairpinRNAs(shRNA)againstRanBP3reduceRanBP3proteinlevels

inHEK293cells.cellsweretransfectedwithshRNAsandpHA262-PURwasco-transfectedtointroducepuromycin

resistence.24hrsaftertransfection,cellsweregrownonpuromycinmediumfor48hrsandcelllysateswereprepared

andanalyzedonWesternblotwithindicatedantibodiesB.RNAiagainstRanBP3increasesWnt1-inducedß-Catenin/

TCF-mediatedtranscription.HEK293cellsweretransfectedwithindicatedconstructsandactivityofTOP(blackbars)

andFOP(greybars)weremeasured72hoursaftertransfectionC.RNAiagainstRanBP3increasesß-Catenin/TCF

driven transcription in HEK293 cells that transiently express an active form of ß-Catenin (ΔGSK3-ß-Catenin). Cells weretransfectedwithindicatedconstructsandluciferaseactivitywasmeasuredafter72hours.D.Co-expression

ofCRM1withRanBP3shRNAconstructsdoesnotaffectß-Catenin/TCFmediatedtranscriptioninWnt1transfected

cells.HEK293cellsweretransfectedwithindicatedconstructsand72hoursaftertransfectionluciferaseactivitywas

measured.Forallexperiments,relativeluciferaselevelsareshownascorrectedwithCMV-Renilla-luc.Errorbarsin

Brepresentstandarddeviationsoftechnicalreplicatesofarepresentativeexperiment.BarsinCandDrepresent

standarderrorsofmeansofindependentexperiments.

0 2 4 6 8 10 12 14 16

+ + + + + + +

+ + + + + + +

- - - - - - -

GFP BP3-3 GFP

TOPFOP RNAi

D B

- -

+ +

- - - - - -

BP3-3 CRM1 0

2 4 6 8 10 12 14

TOPFOP RNAiWnt1

+ + + +

- - -

BP3-2 BP3-8

- -

BP3-3

+

-- -- --

-- -- --

+ + + +

+ +

+ +

+ +

-

+ +

+ +

+

+ + +

- - - -

-

-

BP3-4 BP3-9 BP3-12

BP3-3 GFP

- - - -

- - -

1 2 3

β-Actin shRNA 42 -

80 - kDa

RanBP3

4 5 6

A

0 5 10 15 20 25

TOPFOP RNAi

+

+ + +

- - - - -

BP3-3 +

-

-

+ + + + +

+ +

- - - - -

BP3-12 + + ++

C

Hendriksen et al., Figure 3

relativeluciferaseactivity

relativeluciferaseactivity relativeluciferaseactivity

∆GSK3-Cat

RanBP3

RanBP3 RanBP3

GFP 3 4 GFP 9 12

2

decreasingc-Myclevels.

Reduction of RanBP3 results in increased transactivation of a TCF responsive reporter.

In addition to studying the effects of RanBP3

overexpression, we studied the effects of Ran- BP3depletion.WeexpressedshorthairpinRNAs

(shRNAs)directedagainstdifferentuniqueparts

of RanBP3 that are present in all isoforms of

RanBP3. We obtained several shRNA RanBP3

constructs that downregulate RanBP3 protein

levelsinHEK293cells(Fig.3A).

Whenweco-expressedWnt1andRanBP3shR- NAs, we observed significant increases in TCF/

LEFreporteractivitycomparedtotheGFPRNAi

control(Fig.3B).TotestwhetherRanBP3deple- tionalsoactsonN-terminallydephosphorylated

ß-catenin, we cotransfected ß-cateninΔGSK3ß withanti-RanBP3shRNAexpressionconstructs

(Fig. 3C). Reduction of RanBP3 increased re- porter activity, confirming that RanBP3 can act on the N-terminally dephosphorylated or “acti- vated”formofß-catenin.IntheabsenceofWnt

signaling,depletionofRanBP3didnotresultin

increasedreporteractivity(Fig.3D),arguingfor

a specific effect on ß-catenin. The direct bind- ing of RanBP3 to ß-catenin that we observed

(Fig.1C)indicatedthatRanBP3mayactonthe

WntsignalingpathwayindependentlyofCRM1,

thathasbeenreportedtoplayaroleinß-catenin

nuclearexportviainteractionwithAPC(Hender- son,2000;Neufeldetal.,2000;Rosin-Arbesfeld

et al., 2000). Increased expression of CRM1 is

abletocompensateforreductionofCRM1nu- clearexportatreducedRanBP3levels(Tauraet

al.,1998);(Noguchietal.,1999)).Therefore,we

expressedCRM1incombinationwithWnt1and

RanBP3 shRNAs. As shown in Fig. 3D, CRM1

overexpression did not reverse the effects of

RanBP3 depletion, indicating that the mecha- nism by which Wnt signaling is modulated by

RanBP3isindependentofCRM1-mediatednu- clearexportofß-catenin.

RanBP3 downregulates ß-catenin-mediated transactivation independently of APC.

To further address the question whether Ran- BP3 represses ß-catenin transcriptional activa- tion by stimulating export of ß-catenin via the

APC/CRM1 pathway, we expressed RanBP3 in

human colorectal cancer cell lines that express

C-terminal truncations of APC. First, we tested

DLD1 cells, which express APC1-1417 that re- tainssomeß-cateninbindingsitesbutlacksall

C-terminal nuclear export signals (NES) which

aretheonesmosthighlyconservedinevolution.

AsshowninFig.4A,ß-catenin/TCFactivityisal- readyhighinthesecells.ExpressionofRanBP3

wtor“wv”mutantcouldstilldosedependently

downregulatetranscriptionalactivity,themutant

againbeingalesspotentinhibitor(Fig.4A).As

APCinDLD1cellscanstillbindtoß-catenin,and

NESshavealsobeenreportedintheN-terminus

ofAPC,werepeatedtheexperimentinCOLO320

cells.ThesecellsexpressaveryshortAPCtrun- cation (1-811) that lacks all ß-catenin binding

sites.ß-catenin/TCFactivitywasmuchhigherin

Figure 4. RanBP3 antagonizes Wnt/ß-Catenin transactivation in APC mutated colon carcinoma cells. Luciferase assay showing that RanBP3 inhib- its ß-Catenin-mediated transactivation in colon car- cinoma cell lines DLD1 and COLO320. A. APC type

I truncated human colon carcinoma cell line DLD1

(APC1-1417)wastransfectedwithluciferasereporter

constructs and increasing amounts of RanBP3 ex- pressionconstructsasindicated.DLD1cellsexpress

a truncated APC protein that lacks all its C-terminal

NESs.B.LuciferasereporterassayasinA,carriedout

intheAPCtypeItruncatedhumancoloncarcinoma

celllineCOLO320(APC1-811).Thesecellsexpressa

shortAPCproteinthatlacksallß-Cateninbindingand

regulatorysites.Relativeluciferaseactivitywasmea- sured48hposttransfection.Barsrepresentstandard

deviationsofarepresentativeexperiment.

0 5 10 15 20 25

wt RanBP3 TOPFOP

+ + +

+ +

+

+ - -

- ⁺ +

+ - - +

- + -

+

mut RanBP3 -

-

+ - -

- - - - ⁺ - ⁺ - - - ⁺ - ⁺ DLD1

100 200 300400 100200 300400

A

0 200 400 600 800 1000 1400

wt RanBP3 TOPFOP +

mut RanBP3 -- - +

- -

+ +

- +

- +-

+- +- +

- - + COLO320

300 600 300 600

B

1200

Hendriksen et al., Figure 4

relativeluciferaseactivity relativeluciferaseactivity

thesecellsthaninDLD1cellscorrelatingwiththe

severityoftheAPCmutation(Fig.4B,Rosin-Ar- besfeld et al., 2003). Nevertheless, transfection

oftheRanBP3expressionconstructscauseda

significant downregulation of transcription (Fig.

4B).Therefore,themechanismbywhichRanBP3

inhibitsß-cateninisindependentofanuclearex- portfunctionofAPC.

RanBP3 influences subcellular localization of active ß-catenin

TostudythemechanismbywhichRanBP3inhib- its Wnt signaling, we tested the possibility that

RanBP3 influences the stability of ß-catenin. We transfected HEK293 cells with or without Wnt1

in combination with shRNA constructs. Total

ß-catenin levels were virtually unchanged af- terexpressionofWnt1aloneorincombination

withshRNAagainstRanBP3(Fig.5A).Whenthe

sameblotwasprobedwithanti-activeß-catenin,

recognizing N-terminally desphosphorylated ß- catenin,weobservedanincreaseinWnt1trans- fectedcellsbutnoeffectsofRanBP3(Fig.5A).

From this we conclude that RanBP3 depletion

doesnotaffectß-catenindegradation.

We next prepared nuclear and cytoplasmic ex- tracts from HEK293 cells transfected with or

withoutWnt1andRNAiagainstGFPorRanBP3.

Totalß-cateninwasmostlydetectedinthecyto- solfraction(Fig.5B).Nochangeintotalß-catenin

levelswasobservedinthenucleusorcytoplasm

aftertransfectionwithWnt1(Fig5B,lane2),or

treatment with RNAi (lanes 3 and 4). When we

stainedforactiveß-catenin,aclearincreasewas

evidentafterstimulationwithWnt1(Fig.5B,lane

2).Interestingly,whencellsweretransfectedwith

RNAi against RanBP3, active ß-catenin signifi- cantlyincreasedinthenuclearfractionandde- creased in the cytosolic fraction (Fig. 5B, lane

4), suggesting that RanBP3 relocates active ß- catenin from the nucleus to the cytoplasm. As

controls for fractionation, TCF4 was used as a

nuclear marker andα-Tubulinasacytoplasmic

marker.Bothproteinswerestronglyenrichedin

thepropercompartments.

Nuclear/cytoplasmicfractionationdatadoesnot

allways reflect the subcellular localization in liv- ingcells,aspoolsofproteinsthatarenottightly

boundtonuclearorcytoplasmicstructuresand

are relatively small may leak through NPCs of

permeabilized cells. We therefore assayed the

effect of RanBP3 overexpression on active ß- catenininsituusingtheanti-activeß-cateninan- tibody.Inourhands,thisantibodydidnotvisual- ize endogenous dephosphorylated ß-catenin in

Wnt1transfectedHEK293cells(datanotshown).

We therefore tested two colon carcinoma cell

lines, SW480 and DLD1 that have a constitu- tively activated ß-catenin due to a mutation in

APC (Rosin-Arbesfeld et al., 2003). In SW480,

but not in DLD1, the anti-dephospho-ß-catenin

antibodyrecognizesaclearnuclearsignalabove

background(Fig.6AandC).Thepresenceofthis

signal correlates with the exceptionally high ß- cateninactivityasmeasuredinluciferaseassays

(Fig. 6D), that is approximately 30 fold higher

thaninDLD1.Importantly,RanBP3overexpres- Figure 5. Depletion of RanBP3 results in nuclear

accumulation of active ß-Catenin. A. Depletion of

RanBP3doesnotalterthelevelsofbothtotalandac- tivedephosphorylatedß-Catenin.HEK293cellswere

transfected with or without Wnt1 and shRNA con- structsagainstGFPorRanBP3.72hoursaftertrans- fection,wholecelllysateswereanalyzedbyWestern

blotwithindicatedantibodies.B.RNAiagainstRan- BP3 results in increased levels of active ß-Catenin

in the nucleus. HEK293 cells were transfected with

indicated constructs and 72 hours post transfection

nuclearandcytoplasmicextractswerepreparedand

analyzedbyWesternblot.TCF4andTubulinstaining

areshownasmarkersforpurityofthenuclearandcy- toplasmicfractions.Asaloadingcontrolinthenuclear

fractions TCF4 and a non-specific reaction of the anti- bodyrecognizingactiveß-Cateninareshown.

Hendriksen et al., Figure 5

Nuclear extracts

total β-Catenin

α-Tubulin nonspecific band total β-Catenin active β-Catenin

α-Tubulin Cytoplasmic

extracts

TCF4

β-Actin active β-Catenin

TCF4

+ + +

-

1 2 3 4

92 - 92 - 66 - 47 - 42 - 92 - 92 - 66 - 47 - 125 - kDa

B

total β-Catenin β-Actin Whole cell

extracts

active β-Catenin Wnt

FPG BP3 RNAi

+

- + +

1 2 3 4 92 -

92 - 42 - kDa

A

FPG BP3 RNAi

2

sionleadstoaclearreductionofactiveß-catenin

signalfromtheSW480nuclei(Fig.6A),buthas

no influence on total ß-catenin localization (Fig.

6B).Thisindicatesthat,evenintheextremelyac- tiveSW480cellline,onlyaverysmallproportion

of total ß-catenin is properly dephosphorylated

andactive,andthatthisisthepoolRanBP3acts

upon.

RanBP3 enhances nuclear export of active ß- catenin independently of CRM1

Reduction of active nuclear ß-catenin by Ran- BP3inSW480cellswasnotaccompaniedbyan

increaseincytoplasmicsignal,raisingtheques- tionwhetherRanBP3inducesenhancednuclear

exportofactiveß-cateninoritsincreasedphos- pohorylation.However,enhancednuclearexport

wouldresultindilutioninacytoplasmicvolume

thatisroughlytenfoldlargerthanthatofthenu- cleus,precludingdetectionbytheanti-dephos-

pho-ß-cateninantibody.Todiscriminatebetween

thetwopossibilities,wemimickedtheactivestate

of ß-catenin using a monomeric RFP (mRFP)

tagged, constitutively active form of ß-catenin,

the previously employed ß-cateninΔGSK3ß. To testwhetherthisfusionproteinwasbiologically

active,weperformedaTCF-reporterassayinthe

malignantmesotheliomacelllineNCI-H28,which

carriesahomozygousdeletionoftheß-catenin

gene(Calvoetal.,2000).Thispreventedpossible

activatingeffectsofthismutantonendogenous

ß-catenin. mRFP-ß-cateninΔGSK3ß activated the verylowendogenousTCFactivityofthesecells

toagreatextent(Figure7B).Wenextcompared

thesubcellularlocalizationofthisproteininthe

presenceorabsenceofexogenousRanBP3(Fig.

7A).Carewastakentorecordcellsofsimilarlow

expressionlevels(Fig.7C).Incontrolcells,more

mRFP-ß-cateninΔGSK3ß was present in the nu- cleicomparedtothecytoplasm(mediannuclear

Figure 6. RanBP3 induces specific depletion of endogenous nuclear active ß-Catenin. SW480(A,B)orDLD1

(C)coloncarcinomacellsweretransfectedwithRanBP3expressionplasmidsandstainedafter45hfordephospho- ß-Catenin (A and C) or total ß-Catenin (B). RanBP3 expression was visualized in the same cells using a RanBP3

polyclonal(AandC)ormonoclonalantibody.D.LuciferasereporterassayasinFig.2-4measuringrelativeß-Catenin

activity.CellsweretransfectedasinAandC;Errorbarsrepresentstandarddeviationsoftechnicalreplicates.

A

Hendriksen et al., Figure 6

D C

merge

RanBP3 active β-Cat RanBP3 total β-Cat merge

merge RanBP3 active β-Cat

DLD-1

SW480 SW480

191

76

6.7 2.4

0 50 100 150 200

SW480 SW480

+ BP3 DLD1 DLD1

+ BP3

Relativeluciferaseactivity

B

Figure 7. RanBP3 enhances nuclear export of active ß-Catenin independently of CRM1. A and C. Effect of

RanBP3 on mRFP-ΔGSK ß-Catenin nucleocytoplasmic distribution in HEK293 cells in presence or absence of 50 nM LMB for 3h. A. Box plot showing the distribution of nuclear-cytoplasmic ratios of mRFP-ΔGSK ß-Catenin of two independent experiments. P values are according to Mann-Whitney tests. Representive mRFP fluorescence images areshowninC.Highlightednuclearbordersaredrawnonthebasisofaccompanyingphasecontastimages.B.Func- tionality of mRFP-ΔGSK3-ß-Catenin. NCI-H28 cells (lacking endogenous ß-Catenin) were transfected with indicated constructsand48hoursaftertransfectionluciferaseactivitywasmeasured.Shownarerelativeluciferaselevelsas

corrected for transfection efficiency (Renilla luciferase activity). Bars represent standard deviations. D. Representive fluorescence images of HEK293 cells expressing GFP-Rev(1.4)-NES in the presence or absence of 50 nM LMB for 3h.EandF.Endogenousactivatedß-Cateninrelocalizesfromthenucleustothecytoplasmuponover-expressionof

RanBP3.HEK293cellsweretransfectedwithWntandRanBP3asindicatedtogetherwithTOP-TK-lucandRenilla

transcriptionreporterplasmidsandfractionatedafter48hasinFig.5.Localizationofactiveß-Cateninwasmonitored

using anti-active ß-Catenin antibody. Amounts of protein loaded were normalized on transfection efficiency (Renilla luciferaseactivity).Normalizedß-Catenin/TCFdependentluciferaseactivityisdepictedinF.

A

C

D

Control RanBP3

LMB Rev(1.4)-NES

Hendriksen et al., Figure 6

Control RanBP3 RanBP3 + LMB

0.51.01.5Nuclear/Cytoplasmicratio

p = 1e-11 p = 0.7

mRFP∆GSKβ-Cat

mRFP ∆GSK β-Cat

Control

B

1.0

1

1.5 2.8 257

0100200300 TOP

FOP TOP

FOP

mRFP ∆GSK β-Cat

_ +

Relativeluciferaseactivity

E

active β-catenin nuclear

cytoplasmic

+ +

- ^{BP3 Wnt}

1 2 3

endogenous

0 2 4

F

Rel.luc.activity

1 2 3

RanBP3 + LMB

2

tocytoplasmicratioof1.38,n=37).

In contrast, cells expressing exogenous Ran- BP3showedahighercytoplasmicthannuclear

mRFP-ß-cateninΔGSK3ß levels (median nuclear tocytoplasmicratioof0.77,n=41).Importantly,

additionof50mMoftheCRM1inhibitorlepto- mycin B (Wolff et al., 1997) did not significantly change the effect of RanBP3 (median nuclear

tocytoplasmicratioof0.80,n=52),eventhough

photobleaching experiments show that mRFP- ß-cateninΔGSK3ß rapidly shuttles between the nucleusandcytoplasm(datanotshown).Iden- ticalLMBtreatmentdramaticallyrelocalizedthe

NES-containing reporter protein Rev(1.4)-NES- GFP (Henderson and Eleftheriou, 2000) to the

nucleus (Fig. 6D). We conclude that RanBP3

enhances nuclear export of active ß-catenin,

andthatthisexportisindependentofCRM1.To

confirm that endogenous activated ß-catenin re- localisesfromthenucleustothecytoplasmupon

overexpressionofRanBP3inHEK293cells,we

transfectedthesecellswithWnt1andRanBP3.

Indeed,weobservedincreasedactiveß-catenin

levelsinbothnuclearandcytoplasmicfractions,

ofwhichthenuclearpoolwasmoresensitiveto

RanBP3 overexpression than the cytoplasmic

pool(Fig7E).Thedecreaseincytoplasmicactive

ß-catenin is consistent with increased nuclear

exportofß-cateninandsubsequentdegradation

inthecytoplasm.

RanBP3 suppresses dorsal-ventral axis for- mation in X. embryos

TostudytheroleofRanBP3inWntsignalingin

aphysiologicalcontext,weusedaX. laevisaxis

duplication assay. During X. laevis embryonic

development,Wntsignalingdeterminespattern- ingalongthedorsalventralaxis.Ectopicventral

injection of ß-catenin mRNA in 4-cell embryos

resulted in clear axis duplication (Fig. 8A,B).

The majority of the embryos (75%) showed a

complete duplication of the dorsoventral axis.

22% of the embryos showed a partial duplica- tion i.e. secondary axis without duplicated ce- ment gland. However, co-injection of ß-catenin

mRNAwithRanBP3mRNAresultedinastrong

suppressionofthedoubleaxisphenotypeinthe

majorityoftheembryos(63%).Onlyfewpartial

orverypartialsecondaryaxis(24%and13%re- spectively)wereobservedintheseembryos(Fig.

6B). We also co-injected ß-catenin mRNA with

mRNAoftheRanBP3“wv”mutantthatisdefec- tiveinRanGTPbinding.Thismutantsuppressed

the double axis phenotype but was not such a

potentinhibitorasthewildtypeRanBP3(Fig.8A,

B; p=4e-8). This data correlates with our findings

that this RanBP3 mutant binds ß-catenin with

less affinity (Fig. 1) and that it is less active in re- pressingtranscriptionalactivityofaTCFreporter

geneinhumancelllines(Fig.2and4).IfRanBP3

isaninhibitorofnuclearß-cateninfunction,dor- salinjectionofRanBP3mRNAisexpectedtore- sultinventralizationoftheembryo.Wetherefore

injected4-cellembryosdorsallywitheitherRan- BP3orcontrolmRNAandscoredventralization

afterthreedaysofdevelopmentbydorso-anterio

index(DAI).Mildtosevereventrilizationwasob- served(DIA1-4)in80%ofRanBP3injectedem- bryos(Fig8C),whileonlylessthen10%ofcon- trolinjectedembryosshowedthesephenotypes.

Completeventrilization(DIA0)wasnotobserved.

Animportantdirectdownstreamtargetofdorsal

nuclearß-cateninacitvityistheearlyWnt-induc- ible homeobox gene Siamois (Brannon et al.,

1997). We therefore tested whether expression

levelsofthisgenewerereducedintheRanBP3

injectedembryosbyRT-PCR.Infourindependent

experiments,wedetecteda~2folddecreasein

Siamoislevelsinlatestage9embryos(Figure8D

andE).Thisdecreaseisrathermild,consistent

with the incomplete ventralization phenotypes

observed.Together,weconcludethatRanBP3is

notonlyarepressorofWntsignalinginhuman

celllinesbutitalsofunctionsasanantagonistof

WntsignalinginX. laevisembryos.

Loss of function of ranbp3 results in a naked cuticle phenotype in D. melanogaster Wnt signaling is highly conserved between dif- ferent species. We identified the Drosophila RanBP3homologueandusedRNAitostudyits

role in Drosophila development. At the end of

embryogenesis,theventralepidermisiscovered

byacuticlethatisbuiltupbyarepeatingpattern

of naked cuticle and denticles (Fig. 9A). Wing- less (Wg, Drosophila Wnt) signaling increases

levels of Armadillo (ß-catenin) that specifies the fateofepidermalcellsresponsibleforsecreting

nakedcuticle.Therefore,lossofwgexpression

results in an embryo that is covered with den- ticles lacking naked cuticle (Nusslein-Volhard

andWieschaus,1980)andoverexpressionofwg

resultsinanakedcuticleembryo(Noordermeer

etal.,1992).Likewise,lossofaninhibitorofWnt

signaling also results in naked cuticle embryos

asshownbye.g.RNAiagainstDaxin(Willertet

al.,1999).Asacontrol,weinjectedembryoswith

ß-galactosidase double stranded RNA (dsRNA)

andobservedthatthemajority(97%)developed

intolarvaethatwereindistinguishablefromnon- injected wild-type larvae (Fig. 9A). 3% of these

controlembryosshowedsomeveryweakeffects

DAI 1 DAI 2 DAI 3 DAI 4 DAI 5

0.00.20.40.60.81.0 Dorso anterior index (DAI)

BP3 β-gal Inj. RNA:

Frequency p=0.03 p=0.9

Siamois ODC

BP3 β-gal BP3 β-gal

Relative expression

Inj. RNA:

0.60.81.0

C D

n=63 n=43

E ^{Siamois ODC}

Inj. RNA: BP3 β-g BP3β-g uninj. β-Cat

β-gal Inj. RNA:

0.00.20.40.60.81.0Frequency

β-Cat BP3 wt β-Cat

BP3 mut

B

n=33

n=51 n=38 n=41

Complete Partial

Very partial None Axis duplication

A B

Hendriksen et al., Fig. 7

p=4e-12

p=4e-4 p=4e-12

Figure 8. RanBP3 rescues ß-Catenin-induced double axis formation in X. laevis embryos.A.X. laevisembryos

wereinjectedventrallyatthe4-cellstagewithß-CateninmRNA,inthepresenceorabsenceofcontrolß-galacto- sidaseorXenopusRanBP3-bmRNA.Intheupperpanelwildtypenon-injectedembryosareshown.Middlepanel

showsdoubleaxisphenotypeasinducedbytheinjectionofß-CateninmRNA.Lowerpanelshowsembryosthatare

rescued from the double axis phenotype by co-expression of RanBP3 and ß-Catenin mRNA. B. Quantification of thedifferentphenotypesoftwoindependentexperimentsinfourcategories:completesecondaryaxis(withcement

gland),partialsecondaryaxis(i.e.anysecondaryaxislackingthecementgland),vestigialaxis(verysmallposterior

protrusionorpigmentedline)andnormal(onlyoneaxis).PvaluesareaccordingtoPearson’sChi-squaredtestfor

countdata.C.DorsalinjectionofRanBP3resultsinventralizationofXenopusembryos.4-cellstageembryoswere

injected dorsally with RanBP3 or control (ß-galactosidase) mRNA and analyzed three days later for ventralization

usingthestandarizeddorso-anteriorindex(DAI)(KaoandElinson,1988).Thisscalerunsfrom0(completeventraliza- tion)to5(normaldevelopment).Frequenciesarederivedfromthreeindependentexperiments.P-valuesasinB.D.

The ß-Catenin downstream target siamois is significantly downregulated in RanBP3 injected embryos. Embryos were injected as in C and analyzed for siamois or ornithine decarboxylase (ODC) mRNA using RT-PCR. Amplified ethidium bromide stained DNA of four experiments was quantified and normalized to mean signals from ß-galactosidase in- jectedembryosandrepresentedinaboxplot.P-valuesareaccordingtoMann-Whitneytests.E.Representivesignals

fromRT-PCRreactionsvisualizedbyethidiumbromidestaining.

2

Figure 9. Loss of RanBP3 by RNAi results in a naked cuticle phenotype in Drosophila. Shown are dark field imagesofcuticlepreparationsofcontrol(ß-galatosidase)(A),DrosophilaDaxin(BandC)andDrosophilaRanBP3

dsRNAinjectedembryos(D,E,F).LossofDaxinandRanBP3resultsinincreasedWntsignalingandreplacementof

denticlesbynakedcuticle.Partiallynakedcuticles(BandD),nearlynaked(E)cuticlesandnakedcuticles(CandF)

are shown. All views are ventral, top is posterior. G. Quantification of two representative experiments showing the fre- quencyofthecuticlephenotype.P-valuesarecalculatedasinFigure7B.Notethatthecontributionofthecompletely

nakedphenotypeintheRanBP3RNAiembryosisrelativelyhigh(resultsnotshown).H.RT-PCRshowingreductionin

RanBP3mRNAlevelsinRanBP3dsRNAinjectedembryos.EmbryoswereinjectedasinA.andRNAwasextracted

after 15h of development. RT-PCRs specific for RanBP3 or control (ribosomal protein RP49) were performed using nothing(0)oraseriesof2folddilutionsofextractedRNA.I.LossofRanBP3functionbydsRNAinjectionresultsinin- creasedexpressionofthewgtargetgeneengrailed.ShownareEngrailedantibodystainingofbufferinjectedembryo

(left),DaxindsRNAinjectedembryo(middle)andRanBP3dsRNAinjectedembryo(right).Notethatthebufferinjected

embryodevelopeduntillatestage11whereastheDaxinandRanBP3RNAiembryosshownarestage10embryos,

explainingthelargercellsintheformerembryo.ThenumberofEngrailedpositivecellrowsbetweenstage10and11

isidentical.Ventral-lateralviewisshown,posteriorisleft.

3.6

11.6

27.3

1.8

15.6

20.4

0 5 10 15 20 25 30

A B C

mock

H

G

Daxin Daxin

RanBP3 RanBP3

RanBP3 β-Gal

A B C

D E F

Daxin RNAi RanBP3 RNAi

I

RanBP3 RT-PCR RP49 RT-PCR control

RanBP3 RNAi

0 0

RanBP3 β-Gal RNAi

RNAi Daxin

RNAi

%(partially)naked

Hendriksen et al., Figure 8

input RNA (log2) input RNA (log2)

p=0.001

n=222 n=176 n=214

ondenticlebeltformation(Fig.9G).

RNAi against the (Daxin) resulted in a significant increase in naked cuticle phenotype in 24% of

theDaxindsRNAinjectedembryos(Fig.9G)with

phenotypesvaryingfrompartiallossofdenticles

to completely naked embryos (Fig. 9B and C).

InjectionofdsRNAagainsttheD. melanogaster

RanBP3 caused a partial or complete transfor- mationofdenticlesintonakedcuticlein14%of

theembryos(Fig.9D-F).Themostseverephe- notypes of the RanBP3 RNAi embryos showed

deformationofboththeheadandspiracles(Fig.

9F), resembling Daxin RNAi (Fig 9C). In addi- tion,almostallRanBP3RNAiembryosshowing

a strong naked cuticle phenotype were shorter

thantheembryosinjectedwithDaxindsRNA.To

confirm that the RanBP3 dsRNA injections result- ed in decreased RanBP3 levels, we performed

RT-PCRsonbufferandRanBP3dsRNAinjected

embryos. Fig. 9H shows that RanBP3 mRNA

levelswereindeeddecreasedinRanBP3dsRNA

injectedembryoswhileRP49controlmRNAlev- els remained unaffected. We then assayed the

effects of RanBP3 dsRNA injection on wg tar- get gene induction. For this, stage 10 RanBP3

orDaxindsRNAinjectedembryoswerestained

with anti-Engrailed antibody. Normal engrailed

expression is present in segmental stripes that

aretwocellswide(Fig.9I,left).Removalofthe

WntsignalinginhibitorDaxinbydsRNAinjection

resulted in a broader Engrailed expression pat- ternthatextendedfrom2to4rowsofcells(Fig.

9I;middle).InRanBP3dsRNAinjectedembryos,

Engrailed expression expanded by one row of

cells(Fig.9I,right).Thesein vivodatashowthat

removalofRanBP3leadstoaphenotypethatis

associated with Wnt signaling activation, sug- gestingthatRanBP3alsoactsasnegativeregu- latorofWntsignalinginD. melanogaster.

Discussion

Inthisstudy,weidentifyRanBP3asanovelin- hibitor of Wnt signaling that acts on ß-catenin

directlybyenhancingnuclearexportofitsactive

form.WeshowthatRanBP3bindsdirectlytoß- catenin and that the interaction is increased in

thepresenceofRanGTP.ExpressionofRanBP3

represses Wnt signaling both in vitro and in X.

laevis embryonic development. Inhibition of

RanBP3byRNAicausesover-activationofWnt

signaling in tissue culture cells and in D. mela- nogaster embryos. In addition, expression of

RanBP3 in human cells specifically reduces ac- tiveß-cateninlevelsinthenucleusandrelocates

ΔGSK3-ß-catenin from the nucleus to the cyto-

plasm,independentlyofCRM1.

RanBP3 was originally identified as a nuclear proteinthatcontainsFGrepeatsandaRanGTP- bindingdomain(Muelleretal.,1998).RanBP3can

directlybindthenuclearexportreceptorCRM1,

stimulatingtheformationofnuclearexportcom- plexesandincreasingtheexportrateofcertain

CRM1substrates(Englmeieretal.,2001;Lindsay

etal.,2001).OnemechanismbywhichRanBP3

could influence ß-catenin activity would therefore beincreasednuclearexportviatheCRM1path- way.Althoughthenuclearexportmechanismsof

ß-cateninarenotfullyunderstood,twopathways

have been proposed (Henderson and Fagotto,

2002). In the first, ß-catenin exits the nucleus in- dependentlyofnuclearexportreceptorsbyinter- actingdirectlywithproteinsofthenuclearpore

complex(WiechensandFagotto,2001;Eleftheri- ouetal.,2001).Inthesecondpathway,ß-catenin

exitsthenucleusviatheCRM1pathway,butas

ß-catenin does not contain NESs of its own, it

uses binding to APC to exit the nucleus. The

APC tumor suppressor does contain functional

NESs and has been shown to be exported by

CRM1 (Henderson, 2000; Neufeld et al., 2000;

Rosin-Arbesfeldetal.,2000).Therefore,RanBP3

could inhibit ß-catenin by stimulating its export

viaAPCandCRM1.However,fourlinesofevi- denceargueagainstthis.First,inaCRM1export

complex, RanBP3 would bind to the complex

via CRM1. Instead, we find that RanBP3 inter- acts directly with ß-catenin. Second, ß-catenin

activityisRanBP3-sensitiveincoloncarcinoma

cell line COLO320 (Quinn et al., 1979) that ex- pressesashorttypeIAPCtruncationlackingall

ß-catenin interaction sites (Rosin-Arbesfeld et

al., 2003). We cannot formally exclude that the

neuronal APC-like protein APC2 (van Es et al.,

1999), which is expressed in certain colon car- cinomacelllinescompensatesforlossofAPC.

However, in luciferase reporter assays, CRM1

overexpression does not reverse stimulation of

ß-catenin activity caused by depletion of Ran- BP3. Finally, RanBP3 mediated relocalization

of active ß-catenin is insensitive to leptomycin

B, a potent CRM1 inhibitor (Wolff et al., 1997).

Therefore,weconcludethatthemechanismby

whichRanBP3inhibitsß-cateninisindependent

ofCRM1andAPC.

Recently, it has been suggested that nuclear

ß-catenin signaling is carried out mainly by ß- catenindephosphorylatedatserine37andthre- onine41,whicharemaintargetsitesofGSK3ß

(Staaletal.,2002;vanNoortetal.,2002).Deple- tion of RanBP3 by RNAi specifically increases the amount of dephosphorylated ß-catenin in

2

nuclearfractions,whileRanBP3overexpression

hastheoppositeeffect.Noconcomitantincrease

ofcytoplasmicendogenousactiveß-cateninwas

observedbyoverexpressionofRanBP3,rathera

smalldecrease.Weattributethistocytoplasmic

phosphorylationandsubsequentdegradationof

wild-typeß-catenin.

Endogenous active ß-catenin was visualized in

situ, using the anti-active ß-catenin antibody

recognizing desphosphorylated ß-catenin. This

was only possible in SW480 colon carcinoma

cellsthatcontainahighlevelofactiveß-catenin,

duetoseverelydefectiveAPCfunction(Korinek

et al., 1997). RanBP3 overexpression reduced

active ß-catenin levels in the nucleus, but had

no effect on total ß-catenin. This suggests that

onlyasmallproportionoftotalß-cateninisac- tive in SW480 cells and confirms the specificity ofRanBP3foractiveß-catenin.Apparently,ab- sence of proper ß-catenin phosphorylation and

degradation is not sufficient for ß-catenin to be in an active, dephosphorylated state. Also, we

inferthatthemodulationbyRanBP3ofß-catenin

activity as measured in our luciferase reporter

assaysactsonasmalldephosphorylatedpool,

explaining why RanBP3 modulates wild-type

and ΔGSK3 ß-catenin to a similar extend (Fig. 2 and4).

To discriminate whether RanBP3 enhances ß- catenin N-terminal phosphorylation or nuclear

export, we have visualized both nuclear and

cytoplasmic distribution of active ß-catenin.

For this, we used a fluorescently tagged ß- cateninΔGSK3 that is resistant to N-terminal phosphorylation and degradation. As shown

in Figure 7, RanBP3 causes a clear and highly

significant shift of ß-cateninΔGSK3 from the nu- cleus to the cytoplasm. We therefore conclude

that RanBP3 directly enhances nuclear export

ofactiveß-catenin.HowdoesRanBP3perform

this task? Recent studies have indicated that

theinteractionsofnuclearfactorswithchroma- tinorwitheachotherarehighlydynamic(Dundr

et al., 2002; Phair et al., 2004). This suggests

that RanBP3 does not need to actively remove

ß-catenin from the TCF/LEF-chromatin com- plexes. We therefore favor the possibility that

association with RanBP3 prevents association

ofactiveß-cateninwithchromatinandkeepsit

in a more soluble state. In itself, this would be

sufficient to allow CRM1-independent nuclear exit.WedonotknowwhetherRanBP3accom- paniesß-catenintothecytoplasmandactsasa

truenuclearexportfactor.Thestimulatoryeffect

ofRanGTPontheß-catenin/RanBP3interaction,

andtheconsistentlyweakerinhibitoryeffectson

ß-catenin of a RanBP3 mutant unable to bind

RanGTPwouldargueinfavorofthispossibility.

Hydrolysis of RanGTP in the cytoplasm would

increase the efficiency of release of ß-catenin forsubsequentinteractionswiththecytoplasmic

interacting proteins, such as E-cadherin or the

APC/Axin/GSK3ßcomplex.

WestudiedtheeffectofRanBP3inX. laevis and

D. melanogaster embryogenesis.Overexpression

oftheX. laevishomologueofRanBP3duringear- ly embryogenesis inhibits ß-catenin-dependent

dorsoventral axis formation. RNA interference

of the D. melanogasterhomologue of RanBP3

causesnakedcuticlephenotypesandabroader

Engrailedexpressiondomainduetooveractiva- tionoftheWntsignalingpathway.Therefore,the

results obtained in these two model organisms

support the results obtained in cultured human

celllinesandindicatethattheinhibitoryfunction

ofRanBP3ishighlyconservedinmetazoanevo- lution.

Wnt signaling plays an important role in tumor

initiationandprogressioninavarietyofhuman

solid tumors, including colon carcinomas, he- patocellular carcinomas and melanomas (Bienz

andClevers,2000;Polakis,2000).Asanegative

modulator of Wnt signaling, RanBP3 is a novel

candidate tumor suppressor protein. Interest- ingly,theRanBP3geneislocated19p13.3,are- gionthatiscommonlydeletedinseveraltypesof

cancerandinwhichmultipletumorsuppressor

genesarelikelytobepresent(Leeetal.,1998);

(Oesterreichetal.,2001);(Tuccietal.,2001);(Ya- naihara et al., 2003); (Miyai et al., 2004); (Kato

etal.,2004);(Yangetal.,2004).Furtherworkis

requiredtodetermineiflossoftheRanBP3gene

contributestotheseorothertypesofcancer.

In conclusion, we have identified an unexpected roleforRanBP3asanovelinhibitorofWntsig- naling that enhances nuclear export of active

ß-catenin.Thisfunctionisseparatefromitsrole

in CRM1-mediated nuclear export. The struc- tural similarities between CRM1 and ß-catenin

suggest that RanBP3 may be a more general

cofactor for nuclear export of Armadillo repeat

proteins.

Materials and Methods

Data analysis

StatisticalanalysiswasdoneusingtheRsoftware

package(RDevelopmentCoreTeam,2005).

Reagents

Antibodiesusedwereß-catenin(C19220)(Trans-