J

s

ia

Developmental Biology 217, 230–243 (2000)doi:10.1006/dbio.1999.9551, available online at http://www.idealibrary.com on

TCF Is the Nuclear Effector of the b-Catenin SignalThat Patterns the Sea Urchin Animal–Vegetal Axis

Alin Vonica,*,†,1 Wei Weng,1 Barry M. Gumbiner,* andudith M. Venuti†,2

Department of Anatomy and Cell Biology, College of Physicians and Surgeons, ColumbiaUniversity, New York, New York 10032; *Cellular Biochemistry and Biophysics Program,Memorial Sloan–Kettering Cancer Center, New York, New York 10021; and†Marine Biological Laboratory, Woods Hole, Massachusetts 02543

The mechanism of animal–vegetal (AV) axis formation in the sea urchin embryo is incompletely understood. Specificationof the axis is thought to involve a combination of cell–cell signals and as yet unidentified maternal determinants. InXenopus the Wnt pathway plays a crucial role in defining the embryonic axes. Recent experiments in sea urchins havehown that at least two components of the Wnt signaling pathway, GSK3b and b-catenin, are involved in embryonic AV

axis patterning. These results support the notion that the developmental network that regulates axial patterning indeuterostomes is evolutionarily conserved. To further test this hypothesis, we have examined the role of b-catenin nuclearbinding partners, members of the TCF family of transcriptional regulators, in sea urchin AV axis patterning. To test the roleof TCFs in mediating b-catenin signals in sea urchin AV axis development we examined the consequences of microinjectingRNAs encoding altered forms of TCF on sea urchin development. We show that expression of a dominant negative TCFresults in a classic “animalized” embryo. In contrast, microinjected RNA encoding an activated TCF produces a highly“vegetalized” embryo. We show that the transactivational activity of endogenous sea urchin TCF is potentiated by LiCltreatment, which vegetalizes embryos by inhibiting GSK3, consistent with an in vivo interaction between endogenousb-catenin and TCF. We also provide evidence indicating that all of b-catenin’s activity in patterning the sea urchin AV axiss mediated by TCF. Using a glucocorticoid-responsive TCF, we show that TCF transcriptional activity affects specificationlong the AV axis between fertilization and the 60-cell stage. © 2000 Academic Press

Key Words: sea urchin embryos; TCF/LEF; Xtcf-3; dominant negative; VP-16; b-catenin; animal–vegetal axis.

ts(

dun

INTRODUCTION

In the sea urchin embryo the initial animal–vegetal (AV)axis is specified during oogenesis (reviewed in Davidson etal., 1998). During the early cleavages the AV axis is furtherpatterned to produce five distinct territories (Davidson,1989). Territorial specification is thought to occur by acombination of maternally localized determinants thatspecify the vegetal micromeres and cell–cell signals thatinitially emanate from them. The large micromeres arethought to function as an organizer since transplantation to

1 Both authors have contributed equally to this work.2 To whom correspondence should be addressed at the Depart-

ment of Anatomy and Cell Biology, College of Physicians andSurgeons of Columbia University, 630 W. 168th Street, New York,

bNY 10032. Fax: (212) 305-3970. E-mail: jmv14@columbia.edu.

230

he animal pole of a host embryo will induce vegetaltructures and in some instances a complete second gutHorstadius, 1973; Ransick and Davidson, 1993).

A large body of evidence has identified b-catenin as a keycomponent in axis patterning in both vertebrate and inver-tebrate embryos. b-Catenin is a multifunctional proteinimportant in both cell adhesion and the Wnt signalingcascade (Moon et al., 1997; Miller and Moon, 1996). In theabsence of Wnt signals b-catenin is found in two distinctprotein complexes: at the plasma membrane, whereb-catenin complexes with cadherin and links it to the actincytoskeleton, and in the cytoplasm, where a complex whichincludes GSK3b, APC, and axin targets b-catenin for degra-ation. In the presence of Wnt signals and through an as yetnidentified mechanism, the activity of GSK3b is antago-ized, resulting in an increased stability of b-catenin. Sta-

ilized b-catenin translocates to the nucleus, binds mem-

0012-1606/00 $35.00Copyright © 2000 by Academic Press

All rights of reproduction in any form reserved.

1

G

witditih

Wp

1h

t1fU

Ni1

sq(gi

fP(TcCE

231Patterning of Sea Urchin Animal–Vegetal Axis

bers of the TCF/LEF HMG domain family (TCFs) oftranscriptional regulators, and activates Wnt target genes(Clevers and van der Wetering, 1997; Bienz, 1998).

Recent data from several sources have implicated indi-vidual components of the Wnt signaling pathway in thespecification of the sea urchin AV axis (Fig. 1). It has longbeen known that LiCl alters the pattern of tissues along thesea urchin AV axis (Herbst, 1892; Lallier, 1975; Horstadius,1973). Tissue-specific markers suggest this occurs by ex-panding the vegetal derivatives at the expense of the animalderivatives (Nocente-McGrath et al., 1991). One target ofLiCl is the cytoplasmic kinase, GSK3b (Klein and Melton,996), which is a negative regulator of b-catenin stability

(Yost et al., 1996). In the sea urchin, b-catenin is localizedto the nuclei of the vegetal-most blastomeres (Logan et al.,1999) and any treatment that alters the expression ofb-catenin also affects the pattern of tissue specificationalong the AV axis. Overexpression of a stabilized b-catenin(Wikramanayake et al., 1998) or inactive GSK3b (Emily-Fenouil et al., 1998) results in a vegetalized embryo resem-bling that obtained from LiCl treatment. Similarly, block-ing b-catenin entry into the nucleus by overexpressingcadherin (Wikramanayake et al., 1998; Logan et al., 1999) or

SK3b (Emily-Fenouil et al., 1998) animalizes the embryo.While the identification of components of the Wnt path-ay in AV axis patterning in the sea urchin has provided

nsight into the mechanism of specification and patterning,he nature of upstream components of this cascade and theownstream effectors of the b-catenin signal remains to bedentified. In Xenopus and other embryos, TCFs are media-ors of the nuclear b-catenin signal. Nuclear b-cateninnteracts with TCFs to activate Wnt-responsive genes;owever, in the absence of Wnt signaling and nuclear

b-catenin, TCF proteins repress transcription (Brannon etal., 1997; Cavallo et al., 1998; Roose, et al., 1998; Fisher andCaudy, 1998; Levanon, et al., 1998).

TCFs were originally identified as lymphoid-specificDNA-binding proteins in mice and humans (Oosterwegel,et al., 1991; van de Wetering et al., 1991; Travis et al., 1991;

aterman et al., 1991). Two hybrid screens revealed unex-ected interactions between Lef-1 and b-catenin (Behrens et

al., 1996; Molenaar et al., 1996) and provided a link betweenWnt signaling and gene regulation. Further evidence thatTCFs are downstream effectors of Wnt signals came fromexperiments in Xenopus demonstrating that dominantnegative TCFs that bind DNA but not b-catenin block Wntand b-catenin signaling (Molenaar et al., 1996; van deWetering et al., 1997; Kengaku et al., 1998; Brannon et al.,997; Dorsky et al., 1998). Loss of function of the TCFomologue in Drosophila, dTCF/pangolin, results in seg-

ment polarity defects which cannot be rescued by overex-pression of the b-catenin homologue Armadillo, indicatinghat dTCF is required for Wingless signaling (Brunner et al.,997). In Xenopus and Drosophila, genes involved in axisormation and segmental identity, siamois, Xnr-3, twin, and

bx, are regulated by TCFs in association with b-catenin

Copyright © 2000 by Academic Press. All right

(Brannon et al., 1997); mutations in the TCF binding sites ofthese genes reduce their response to Wnt signals.

The established link between TCFs and Wnt signaling ledto the identification of TCF homologues in a wide variety ofspecies. All members of the TCF family share anN-terminal b-catenin binding domain and possess a virtu-ally identical DNA-binding domain, a high-mobility-groupdomain, which induces a sharp bend in the DNA helix(Love et al., 1995). b-Catenin binds directly to the

-termini of TCF/LEF factors (Huber et al., 1996) andncreases their transactivational activity (Molenaar et al.,996).To test the role of TCFs in mediating b-catenin signals in

ea urchin AV axis development we examined the conse-uences of expressing altered forms of Xenopus TCF3Xtcf-3) on sea urchin development. We assayed, by reporterene microinjections, for endogenous sea urchin TCF activ-ty and examined the effect of altering b-catenin levels on

this activity. Using a glucocorticoid-responsive Xtcf-3 weexamined the competence of sea urchin embryos to respondto exogenous TCF and the timing of TCF’s effect onspecification along the AV axis.

MATERIALS AND METHODS

Embryos

Adult Lytechinus variegatus were obtained from Beaufort Bio-logicals, Duke University Marine Laboratory (Beaufort, NC) andfrom Susan Decker Services (Davie, FL). Eggs were harvested afterintracoelomic injection of 0.5 M KCl and fertilized with a dilutedsuspension of sperm. Embryos were cultured in Millipore filteredartificial seawater (MFSW). For LiCl treatment, cultures weretreated with 30 mM LiCl in MFSW from immediately afterfertilization until the hatched blastula stage (;16 h at 17°C).

Plasmid Construction

RNA expression vectors were constructed in pCS21 (Turner andWeintraub, 1994). The full-length open reading frame of Xtcf-3 wasobtained by RT-PCR from stage 8 Xenopus embryo RNA, using 59(CGCCGAATTCCGGCATGCCTCAACTAAACAGCGGCGGG-GGGG) and 39 (ACGTCTAGAGCTCAGTCACTGGATTTG-GTCACC) primers derived from the published sequence of Xtcf-3(Molenaar et al., 1996). The PCR product was inserted as an EcoRI/XbaI fragment into pCS21 and confirmed by sequencing. To con-struct the ‚b-Xtcf-3 (DN Xtcf-3; Fig. 2C) expression vector, theragment corresponding to amino acids 88 to 553 was generated byCR using the wt Xtcf-3 plasmid as template with 59

CGCGGAATTCATGAGCAAAGCTCATTTCTGAAGAGGACT-GAATGAACAGGATGCAGCGTTCTTCAAGGG, which in-ludes an N-terminal 30-bp Myc tag) and 39 (CGCGTCTAGAT-AGTCACTGGATTTGGTCACC) primers and cloned as ancoRI/XbaI insert into pCS21. pCS2-VP16‚bXtcf-3 (Activated

Xtcf-3; Fig. 2D) was constructed by inserting in frame at the EcoRIsite of pCS2-‚bXtcf-3 a PCR fragment encoding amino acids 411 to490 of VP16, cloned with 59 (CGCCGATTCATGTCGACGGC-CCCCCCGACCG) and 39 (CCGCGAATTCCCCACCG-

TACTCGTCAATTCC) primers and digested with EcoRI. pCS2-

s of reproduction in any form reserved.

l

Xag

232 Vonica et al.

FIG. 1. Components of the Wnt pathway are implicated in specification of cell fates along the sea urchin AV axis. Overexpression studieshave shown that components of the Wnt signaling pathway are involved in the specification of fates along the AV axis. Increasedstabilization of b-catenin, by inhibiting GSK3b with LiCl treatment or by injecting a kinase-dead GSK3b (A) or stabilized b-catenin (B),eads to vegetalization of the embryo. Decreased nuclear b-catenin resulting from GSK3b overexpression (C) or titrating cytoplasmic

b-catenin with cadherin (D) results in the animalized phenotype.FIG. 2. Constructs used in microinjections. The structure of dominant negative Xtcf-3 (DN Xtcf-3; C), Activated Xtcf-3 (D), and inducible

tcf-3 (GR Xtcf-3; E), are shown in comparison with wild-type Lef1 (A) and Xtcf-3 (B). The respective domains are indicated (after Cleversnd van de Wetering, 1997); CAD, context-sensitive activation domain; HMG, high-mobility group; NLS, nuclear localization signal; GR,

lucocorticoid ligand binding domain; GR AD, glucocorticoid activation domain).

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

233Patterning of Sea Urchin Animal–Vegetal Axis

FIG. 3. DN Xtcf-3 animalizes and Activated Xtcf-3 vegetalizes sea urchin embryos. Micrographs of 1-week-old living L. variegatusembryos injected with DN Xtcf-3 RNA at fertilization appear animalized (B and C) compared to controls at the same stage (A). Embryosinjected with Activated Xtcf-3 at fertilization demonstrate features of vegetalization after 1 week of culture (D and E). This includes

incomplete archenterons and supernumerary pigments cells.

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

ap

d(b

a

mArcw(i

(o

iLeesA

234 Vonica et al.

GR‚bXtcf-3 (GR Xtcf-3; Fig. 2E) was a gift from Paul Wilson(Cornell Medical School). The pTOP-FLASH luciferase reporterplasmid (Korinek, et al., 1997) was a gift from Hans Clevers (Uni-versity Hospital, Utrecht, The Netherlands).

In Vitro Transcription and Microinjection

Capped RNAs were in vitro transcribed with SP6 polymerase(Promega Corp., Madison, WI) from vectors linearized with NotI.Before injection, the appropriate RNA and DNA dilutions in 40%glycerol were filtered by centrifugation through a Millipore0.45-mm Low Protein Binding-Millex filter (Millipore, Bedford,MA). RNA microinjections were performed as in Mao et al. (1996)nd DNA injections as described in McMahon et al. (1985).TOP-FLASH reporter plasmid was linearized with BamHI and 20

fg/embryo was co-injected with a 4:1 excess of BamHI-digested seaurchin sperm DNA.

The embryos injected with GR Xtcf-3 were treated with 10 mMexamethasone beginning at fertilization and at intervals thereafterKolm and Sive, 1995). Control embryos included uninjected em-ryos treated with dexamethasone and GR‚bXtcf-3-injected em-

bryos without dexamethasone treatment. For Xenopus reporterassays, 100 pg of pTOP-FLASH was injected into Xenopus embryosat the 2-cell stage and animal caps were assayed at a later time. Todetermine the effect of LiCl on endogenous Xenopus TCF activity,embryos injected with reporter DNA were incubated for 5 minwith 0.3 M LiCl at the 32-cell stage. RNA encoding ActivatedXtcf-3 (10 pg) was injected at the 2- to 4-cell stage, in theventromarginal zone, and axes were scored at the tadpole stage.

Reporter Assays

Embryos injected with the pTOP-FLASH reporter were har-vested shortly after hatching and deposited in a round-bottom96-well plate and excess MFSW was aspirated. Embryos were thenlysed in 50 ml Lysis buffer (Luciferase Assay; Promega Corp.,Madison, WI) and the assay was performed on a Berthold luminom-eter with 10 ml of lysate. Approximately 3 3 300 embryos wereassayed in each experiment and each experiment was duplicated.Xenopus embryos injected with the reporter plasmid were recov-ered in Lysis buffer at stage 9, using 50 ml for four embryos. Eachssay was done in triplicate.

Immunocytochemistry

Embryos were fixed in 100% ice-cold methanol for 20 min andthen transferred to 13 PBS with 0.1% Tween (PBST). Monoclonalantibodies Endo1, 6A3, and EctoV were provided by D. McClay(Duke University) and 6A9 by C. Ettensohn (Carnegie–MellonUniversity). Polyclonal antibody against myosin heavy chain

FIG. 4. Tissue-restricted markers further demonstrate alterationmages of control embryos (A, B, G, H), DN Xtcf-3 RNA-injected em) cultured to the late gastrula stage and examined by indirect imxpression is restricted to the outer epithelium in the animal halfxpression is more globally distributed (C, D). In contrast, Activatetaining (E, F). Conversely, endoderm marker (Endo1) expression

ctivated Xtcf-3 RNA-injected embryos (K, L) compared to controls (G

Copyright © 2000 by Academic Press. All right

(MHC) was from G. Wessel (Brown University). Primary antibodieswere diluted in PBST as follows: 1:2 for EctoV; 1:10 for Endo1, 6A3,and 6A9; and 1:200 for anti-MHC. The myc antibody was mono-clonal anti-myc 9E10.2 (as ascites fluid) used at a 1:50 dilution.Monoclonal antibodies were detected with FITC-conjugated goatanti-mouse antibody at 1:100 dilution in PBST (Cappel OrganonTeknika). TRITC-conjugated goat anti-rabbit secondary antibodydiluted 1:200 in PBST was used to detect anti-MHC. Embryos werewashed with PBST and then blocked with 5% normal goat serum(Sigma, St. Louis, MO) in PBST for 30 min, incubated with dilutedprimary antibody for 1 h at room temperature, and washed threetimes with PBST. A 30-min incubation with secondary antibodywas followed by 3 3 5-min washes of PBST before embryos were

ounted in 1:1 PBST:glycerol. Embryos were viewed on a Zeissxiovert 100 inverted microscope equipped with DIC and epifluo-

escence optics. Images were captured with a Dage DC330 videoamera (Dage-MTI, Inc., Michigan City, MI). Confocal microscopyas performed with a Zeiss Laser Scanning Confocal Microscope

Axiovert; Carl Zeiss, Inc., Thornwood, NY), using a 403 watermmersion objective.

RESULTS

Dominant Negative and Activated Xtcf-3 HaveOpposing Effects on Cell Fate Specificationalong the AV Axis

One-week-old embryos were examined after injectionwith RNA encoding a dominant negative Xtcf-3 (DN Xtcf-3)protein that lacks the N-terminal b-catenin binding domainFig. 2C). These embryos resemble those observed withther treatments that reduce b-catenin stability (Emily-

Fenouil et al., 1998) or the ability of b-catenin to enter thenucleus (Wikramanayake et al., 1998; Logan et al., 1999). Atthe highest concentration used (2 pg/embryo) DN Xtcf-3-injected embryos resemble Dauer larvae (Figs. 3B and 3C)that develop from animal halves obtained from bisectingeggs or early embryos perpendicular to the AV axis (Hors-tadius, 1973; Wikramanayake and Klein, 1995). As theconcentration of RNA is reduced (2–0.2 pg/embryo) thedegree of animalization is also reduced (data not shown).Embryos animalized with high concentrations of DN Xtcf-3rarely contained morphological features of mesodermal orendodermal differentiation at early stages of development.However, when embryos were allowed to develop for ex-tended times as shown in Fig. 3, spicules were observed.Since the injected RNA may have decayed after this lengthof time and the protein product from this injected RNA is

ate specifications along the AV axis. Bright-field and fluorescentos (C, D, I, J), and Activated Xtcf-3 RNA-injected embryos (E, F, K,nofluorescence for lineage marker expression. In controls, EctoVe embryo (A, B), but in DN Xtcf-3 RNA-injected embryos EctoVf-3 RNA-injected embryos are almost completely devoid of EctoVuced in DN Xtcf-3 RNA-injected embryos (I, J) and expanded in

s in fbrymuof th

d Xtcis red

, H).

s of reproduction in any form reserved.

maLacvtpsacVitf1

Mcrct1

enE

fMD(X(etspheaelecsaus

tafgeprti6XhcTbmgaeu

235Patterning of Sea Urchin Animal–Vegetal Axis

unlikely to be present 4 days after injection, the embryosmay be able to recover from the dominant negative effectsof DN Xtcf-3. Also, since sea urchin embryos are highlyregulative, the appearance of spicules in these animalizedembryos may be due to regulative events.

In contrast, 1-week-old living embryos injected withRNA encoding Activated Xtcf-3 (Figs. 3D and 3E) have amorphological phenotype opposite to that observed withDN Xtcf-3 injections. Activated Xtcf-3 RNA encodes aprotein that uses the potent activation domain of the viraltranscription factor VP16 (Triezenberg et al., 1988) in placeof the b-catenin binding domain (Fig. 2D) to activate targetgenes. The activity of this construct was first tested inXenopus. Ventral injections of 10 pg RNA produced com-plete secondary axes in 100% of the injected embryos (datanot shown). Sea urchin embryos injected with the ActivatedXtcf-3 were severely vegetalized. They resembled embryosinjected with RNA encoding a stabilized b-catenin (Wikra-

anayake et al.,1998) or inactive GSK3b (Emily-Fenouil etl., 1998) or embryos treated with high concentrations ofiCl (Wikramanayake et al., 1998; Nocente-McGrath, etl., 1991), all treatments thought to increase the nuclearoncentration of b-catenin. The Activated Xtcf-3-egetalized embryos show very little morphological varia-ion and develop as epithelial balls containing numerousigment and other cells internally (Figs. 3D and 3E). Occa-ionally invaginated portions of the outer wall are evidents is observed in severe exogastrulae formed from highoncentrations of LiCl (Nocente-McGrath et al., 1991; J.enuti, unpublished observations). The Activated Xtcf-3-

njected embryos hatch later than controls, presumably dueo reduced animal territory development which is requiredor hatching enzyme production (Emily-Fenouil et al.,998).

Alteration of Tissue-Specific Marker Expression byMutant Forms of Xtcf-3

To ascertain the degree of animalization/vegetalizationattained with microinjection of altered forms of Xtcf-3 weexamined injected embryos with lineage-specific markersand compared their patterns of expression to those observedin control embryos of the same stage. The antibodies usedwere Endo1 as an endodermal marker (Wessel and McClay,1985), EctoV as an oral ectoderm marker (Coffman andMcClay, 1990), and 6A9 (Ettensohn and McClay, 1988) andanti-MHC (Wessel et al., 1990) as mesodermal markers.

onoclonal antibody 6A9 recognizes primary mesenchymeells (Ettensohn and McClay, 1988). Polyclonal anti-MHCecognizes muscle cells derived from secondary mesen-hyme and a subset of myoepithelial cells in the endodermhat also label with Endo1 (Wessel et al., 1990; Venuti et al.,993).Control and DN Xtcf-3- and Activated Xtcf-3-injected

mbryos were examined at the late gastrula stage by immu-ofluorescence. Control embryos (Figs. 4A and 4B) express

ctoV in a portion of the outer ectoderm cells destined to

Copyright © 2000 by Academic Press. All right

orm the oral epithelium of the pluteus larva (Coffman andcClay, 1990). Embryos of the same stage injected withN Xtcf-3 RNA express this ectodermal marker globally

Figs. 4C and 4D), whereas embryos injected with Activatedtcf-3 express this marker in only a few cells of the embryo

Figs. 4E and 4F). DN Xtcf-3-injected embryos developctoderm, an animal-half derivative, at the expense of otherissues, whereas the Activated Xtcf-3-injected embryos lackignificant ectoderm differentiation. Control embryos ex-ress the endodermal marker Endo1 in the midgut andindgut (Figs. 4G and 4H) but DN Xtcf-3-injected embryosxpress virtually none of this endodermal marker (Figs. 4Ind 4J). In contrast, embryos injected with Activated Xtcf-3xpress Endo1 at high levels throughout most of the epithe-ial layer (Figs. 4K and 4L). Reduced endodermal markerxpression in DN Xtcf-3-injected embryos and the in-reased expression in Activated Xtcf-3-injected embryosupport our morphological observations on living embryost later stages. Overexpression of DN Xtcf-3 in the searchin leads to an animalized phenotype and overexpres-ion of Activated Xtcf-3 leads to a vegetalized one.To further examine the conversion of the embryonic

issues along the AV axis by mutant forms of Xtcf-3 wenalyzed injected embryos by indirect immunofluorescenceor mesoderm formation. In control embryos at the lateastrula stage (Fig. 5B) muscle cells are just beginning toxpress MHC at the tip of the archenteron. Similarly,rimary mesenchyme cells are visible as a well-organizeding of cells (Fig. 5C) that express the marker 6A9 aroundhe base of the archenteron. In contrast, DN Xtcf-3 RNA-njected embryos do not express MHC and only a fewA9-positive cells are observed (Figs. 5D–5F). Activatedtcf-3 RNA-injected embryos at this stage (Figs. 5G–5I),owever, contain numerous cells expressing 6A9 and in-reased staining of cells in the outer epithelium for MHC.his pattern of staining with both Endo1 and MHC anti-odies in the outer epithelium resembles that seen inyoepithelial cells normally found in the sphincters of the

ut of unperturbed embryos (Wessel et al., 1990; Venuti etl., 1993) and is also observed in the expanded endoderm ofmbryos treated with high concentrations of LiCl (J. Venuti,npublished observations).

Endogenous TCF Activity in the Sea UrchinEmbryo Can Be Potentiated with LiCl Treatment

To demonstrate that an endogenous TCF activity ispresent in the sea urchin and can account for the regulationof b-catenin target genes in the unperturbed embryo, weexamined the activity of an injected luciferase reporter genethat contained multiple TCF/LEF binding sites upstream ofa c-fos minimal reporter (pTOP-FLASH; Korinek et al.,1997). The pTOP-FLASH reporter gene was injected into L.variegatus at fertilization and luciferase activity measuredat the hatched blastula stage. Luciferase activity in injectedembryos averaged an approximate 10-fold increase in activ-

ity over background (Fig. 6A). Since TCF/LEFs are thought

s of reproduction in any form reserved.

Fgwred(IeFtwbl

IG. 5. Comparison of the effects of microinjecting DN Xtcf-3 and Activated Xtcf-3 on mesodermal marker expression. In lateastrula-stage control embryos (A–C) MHC expression is detectable in a few developing muscle cells at the tip of the archenteron andeakly in myoepithelial cells within the developing sphincters of the endoderm (B). PMC marker expression delineates a well-developed

ing of skeletogenic mesenchyme cells by this stage (C). In contrast, DN Xtcf-3-injected embryos (D–F) have little to no MHC (E) or PMCxpression (F); however, embryos injected with Activated Xtcf-3 (G–I) contain numerous skeletogenic cells, and MHC staining isramatically increased in cells of the outer epithelium. Previous examination of LiCl-treated embryos double labeled with MHC and Endo1

data not shown) show similar epithelial staining which corresponds to increased MHC expression in myoepithelial cells of the endoderm.n general, mesodermal cell differentiation is reduced in DN Xtcf-3-injected embryos compared to control and Activated Xtcf-3-injectedmbryos.IG. 6. Endogenous TCF activity is detectable by activation of reporter gene expression and is potentiated by LiCl treatment. (A) Whenhe pTOP-FLASH reporter gene was injected into L. variegatus embryos and assayed at the hatched blastula stage for luciferase activity,e found greater than 10-fold increase in enzyme activity over background. This activity was further increased to greater than 25-fold overackground if embryos injected with the pTOP-FLASH reporter were also treated with LiCl (30 mM) to increase endogenous b-catenin

evels. (B) A proportionate increase in reporter activity was also observed in injected Xenopus embryos after LiCl treatment.

FIG. 7. Embryos animalized by DN Xtcf-3 are insensitive to LiCl treatment but DN Xtcf-3 can reverse the effects of LiCl. Controlembryos (A–C), embryos injected with DN Xtcf-3 (D–F), embryos vegetalized with 30 mM LiCl (G–I), or embryos injected with DN Xtcf-3and treated with 30 mM LiCl (J–L) were compared for tissue-specific marker expression at the gastrula stage. EctoV expression is restrictedto the animal-half of control embryos (A), is reduced in LiCl-treated embryos (G), but is expanded in DN Xtcf-3 with (J) or without (D) LiCltreatment. In contrast, Endo1 expression is restricted to the mid- and hindgut of controls (B), expanded in LiCl-treated embryos (H), but isreduced in embryos injected with DN Xtcf-3 both with (K) and without (E) LiCl treatment. Mesoderm expression is severely reduced in DNXtcf-3-injected embryos (F), even with LiCl treatment (L) compared with control (C) and LiCl-treated (I) embryos. Our data support thenotion that the effect of LiCl is completely reversed by DN Xtcf-3 microinjection and there is no effect of LiCl that cannot be counteredby the DN Xtcf-3.FIG. 8. Xenopus C-cadherin has no effect on embryos vegetalized by Activated Xtcf-3. Embryos injected with Xenopus C-cadherin (D–F)display a phenotype resembling that observed with DN Xtcf-3; embryos are animalized with expanded ectodermal marker expression (D)and endoderm (E) and mesoderm (F) marker expression reduced compared to controls (A–C). Embryos injected with Activated Xtcf-3 alone(G–I) or together with C-cadherin (J–L) are vegetalized; endoderm is expanded (H and K) at the expense of ectoderm (G and J) whilemesoderm is not severely affected (I and L). No effect of increased adhesion due to C-cadherin overexpression was observed, nor did reduced

nuclear b-catenin influence the phenotype of Activated Xtcf-3-injected embryos.

XtiaTnott

C1

wmtlr

awjecn7awtad

fipC(R

gw

WXdtdDbgctn

d

238 Vonica et al.

to require b-catenin for activation of target genes we testedthe effect of elevating endogenous b-catenin levels onendogenous TCF activity by treating embryos injected withthe pTOP-FLASH reporter with LiCl (30 mM) from fertili-zation through hatching. LiCl treatment increased lucif-erase activity to greater than 25-fold over background (Fig.6A). LiCl treatment potentiated the endogenous TCF activ-ity, suggesting that endogenous TCF is a mediator ofendogenous b-catenin signaling. In a similar experiment in

enopus laevis, 100 pg of pTOP-FLASH was injected intohe animal pole of two-cell-stage embryos. Luciferase activ-ty also increased approximately 2-fold over the endogenousctivity in Xenopus embryos after LiCl treatment (Fig. 6B).herefore, LiCl increased endogenous TCF activity in Xe-opus embryos to the same extent as in sea urchins. Inther experiments (data not shown) the reporter with mu-ated TCF sites showed no change in activity with LiClreatment.

The Effect of DN Xtcf-3 Is Insensitive to Elevatingb-Catenin Levels with LiCl

Since TCF/LEFs are thought to act as repressors of tran-scription in the absence of b-catenin (Brannon et al., 1997;

avallo et al., 1998; Roose et al., 1998; Fisher and Caudy,998; Levanon et al., 1998), we tested whether the effects of

b-catenin elevation on sea urchin AV axis patterning couldbe accounted for solely through b-catenin’s interaction

ith TCF. Control embryos, embryos treated with LiCl (30M), embryos injected with DN Xtcf-3, and embryos

reated with LiCl and injected with DN Xtcf-3 were col-ected at the late gastrula stage and examined for tissue-estricted marker expression.

LiCl-treated embryos show elevated expression of Endo1nd reduced expression of EctoV (Figs. 7G–7I) comparedith controls (Figs. 7A–7C), whereas embryos microin-

ected with DN Xtcf-3 (Figs. 7D–7F) show increased EctoVxpression and reduced Endo1 expression compared toontrols. When DN Xtcf-3-injected embryos were simulta-eously treated with LiCl and then examined (Figs. 7J andK) no differences were observed between these embryosnd those injected with DN Xtcf-3 alone. The effect of LiClas completely reversed by DN Xtcf-3 injection, suggesting

hat the effects of b-catenin are mediated through TCF/LEFnd that most, if not all, effects of LiCl on early embryos areependent on TCF/LEF-responsive promoters.

Xenopus C-Cadherin Has No Effect on EmbryosVegetalized by Activated Xtcf-3

Previous studies in sea urchins interfered with b-cateninunction either by reducing b-catenin stability by increas-ng levels of GSK3b (Emily-Fenouil et al., 1998) or byreventing b-catenin nuclear entry by titrating with-cadherin (Wikramanayake et al., 1998) or Lv-cadherin

Logan et al., 1999). We have shown that Activated Xtcf-3

NA microinjection bypasses the need for b-catenin to

Copyright © 2000 by Academic Press. All right

enerate a vegetalized phenotype. We therefore testedhether overexpression of C-cadherin to reduce nuclear

b-catenin levels would interfere with the vegetalized phe-notype of embryos injected with Activated Xtcf-3. Weexamined the phenotype of embryos injected with XenopusC-cadherin RNA, Activated Xtcf-3 RNA, or both RNAsinjected simultaneously.

Embryos injected with C-cadherin (Figs. 8D–8F) showpatterns of tissue-restricted expression indicative of ani-malized embryos. EctoV expression is global in theC-cadherin-injected embryo while vegetal marker expres-sion (EndoV and 6A9) is virtually absent. In contrast,embryos injected with Activated Xtcf-3 demonstrate re-duced EctoV-, increased Endo1-, and disorganized 6A9-labeled cells, indicative of a vegetalized phenotype (Figs.8G–8I). When C-cadherin and Activated Xtcf-3 RNAs wereco-injected, the phenotype resembled that seen with Acti-vated Xtcf-3 RNA alone (Figs. 8J–8L). No apparent effect onadhesion due to C-cadherin overexpression could be ob-served in the vegetalized embryos that result from theco-injection. Primary mesenchyme ingression occurs nor-mally despite the observed requirement for internalizationof cadherin during these events (Miller and McClay, 1997).

Exogenous TCF Affects Specification along the AVAxis during Cleavage

To examine the timing of TCF/LEF function in the seaurchin embryo we asked when during development exog-enously introduced TCF could activate b-catenin targets.

e utilized a hormone-inducible glucocorticoid receptortcf-3 fusion protein (GR Xtcf-3) that was dependent onexamethasone binding for nuclear translocation and func-ional activity and contains a transcriptional activationomain within the GR ligand binding region (Fig. 2E;ahlman-Wright et al., 1994). This construct was designedased on the observation that in the absence of ligand thelucocorticoid receptor is retained in the cytoplasm in aomplex with hsp90, but upon ligand binding, hsp90 reten-ion is repressed and the fusion protein can enter theucleus (Cadepond et al., 1991).Embryos microinjected with GR Xtcf-3 were treated with

examethasone (10 mM) at various times after fertilizationand examined at a later stage for phenotypic effects. GRXtcf-3-injected embryos became vegetalized only whentreated with dexamethasone between fertilization and the60-cell stage. Treatment of uninjected embryos with dexa-methasone did not cause any morphological abnormalitiesnor did microinjection of the GR Xtcf-3 RNA withoutdexamethasone treatment.

The degree of vegetalization and proportion of embryosaffected were dependent on the stage that dexamethasonewas added (Fig. 9). The severity of vegetalization andpercentage of embryos affected decreased from fertilizationthrough the 60-cell stage. Up to the 16-cell stage themajority of embryos demonstrate a classic vegetalized phe-

notype with increased Endo1 and reduced EctoV staining.

s of reproduction in any form reserved.

iiCLi1

t

b

s

ara

a

g

sdnmaioat

239Patterning of Sea Urchin Animal–Vegetal Axis

However, if treated at the 60-cell stage or later, greater than90% of the treated embryos appear phenotypically normal.

To demonstrate that the GR Xtcf-3 protein was stillpresent after the 60-cell stage and that dexamethasone wasaltering the subcellular localization of the protein in in-jected embryos, we examined embryos microinjected withGR Xtcf-3 RNA after culture with or without dexametha-sone treatment for subcellular protein localization. Em-bryos were fixed and examined at the hatched blastula stagewith antibody to myc, which recognizes the myc epitopetag on the GR Xtcf-3 protein (Fig. 2E). The antibody de-tected diffuse protein staining in embryos without dexa-methasone treatment, but after treatment with dexameth-asone from fertilization through hatching there wasincreased nuclear localization of GR Xtcf-3 (data notshown). These results demonstrate that the GR Xtcf-3protein is retained in the cytoplasm in the absence ofdexamethasone, but induction by the addition of dexameth-asone results in increased nuclear localization of GR Xtcf-3protein.

DISCUSSION

b-Catenin is thought to function in cell fate specificationthrough direct association with TCF/LEF transcription fac-

FIG. 9. Inducible Xtcf-3 reveals an early window of competenceto respond to the b-catenin signal. To determine when duringdevelopment embryos are competent to respond to the b-cateninignal, we examined GR Xtcf-3-injected embryos induced withexamethasone at various times after injection. The percentage oformal vs vegetalized embryos observed for each stage that dexa-ethasone was added is shown. An average of 120 embryos was

ssessed for each induction in three separate experiments. Vegetal-zation occurs only between fertilization and the 60-cell stage andnly when GR Xtcf-3-injected embryos are treated with dexameth-sone. Untreated GR Xtcf-3-injected embryos and dexamethasone-reated uninjected embryos appeared morphologically normal.

tors and the accumulation of the resulting protein complex s

Copyright © 2000 by Academic Press. All right

n the nucleus. TCF/LEFs act as transcriptional repressorsn the absence of nuclear b-catenin (Brannon et al., 1997;avallo et al., 1998; Roose et al., 1998; Fisher et al., 1998;evanon, et al., 1998), but transactivational activity isncreased by cotransfection with b-catenin (Molenaar et al.,996). It is thought that b-catenin provides the transactiva-

tion component and TCF/LEFs the DNA binding compo-nent of the TCF/b–catenin complex that forms in thenucleus. We therefore hypothesized that in the sea urchin,overexpression of a dominant negative form of Xtcf-3 thatcannot interact with b-catenin would repress b-cateninargets and produce the opposite effect of overexpressing

b-catenin animalizing the embryo. Conversely, a constitu-tively active Xtcf-3 should activate b-catenin targets andvegetalize the embryo even in the absence of b-catenininding.Using modified forms of Xtcf-3, we show that overexpres-

ion of a dominant negative form of Xtcf-3 which lacks theb-catenin binding domain animalizes the sea urchin em-bryo and leads to broadening of the ectodermal territory atthe expense of endoderm. Others have reported that loss ofb-catenin function by overexpression of cadherins alsointerferes with the specification of the micromere lineage(Logan et al., 1999). However, when we block b-cateninfunction by interfering with its ability to interact with TCFand activate target genes, we observe a small number ofcells that stain with antibodies that recognize the progenyof micromeres, the PMCs. In DN Xtcf-3-animalized em-bryos 4 days after injection we observed spicules, theterminal differentiation product of PMCs. Interestingly,Huang et al. (1999) using RT-PCR analysis of PMC geneexpression also report that the PMC lineage is not asseverely affected by DN Sp Tcf/Lef overexpression as wasobserved with cadherin overexpression (Wikramanayake etal., 1998; Logan et al., 1999). These results can be inter-preted to mean that the micromere specification function ofb-catenin may not be mediated by TCF. However, they canalso be explained if the dose of DN Xtcf-3 used in theseexperiments is not sufficient to block micromere specifica-tion by b-catenin. Alternatively, decay of the injected RNAnd its resulting protein product may occur early enough foregulation to occur. We are currently investigating theselternatives.In contrast, the activated form of Xtcf-3, which containsVP-16 activation domain in place of the b-catenin-binding

domain, vegetalizes the embryo and expands the endoder-mal territory at the expense of the ectoderm. In furthersupport of a reduced animal territory in these embryos theyhatch later than controls of the same age, presumablybecause the embryos are producing less hatching enzyme asobserved when GSK3b is inhibited (Emily-Fenouil et al.,1998). The use of a fusion between Xtcf-3 and the transcrip-tional activation domain of VP-16 is a unique approach tobypass the necessity for a b-catenin TCF/LEF interactionand allow transactivation of potential b-catenin targetenes in the embryo. This approach also has been used

uccessfully in Xenopus (A. Vonica and B. Gumbiner,

s of reproduction in any form reserved.

evK

tna

c1

e

se

t

abs

pbnarbn

t

gtTT

sem

t

tpt

240 Vonica et al.

manuscript in preparation) and in tissue culture cells toactivate b-catenin targets (Aoki et al., 1999). It also has beenmployed in the activation of other embryonic genes inivo (Rusch and Levine, 1997; Cambridge et al., 1997;essler et al., 1997; Lemaire et al., 1998; Ferreiro et al.,

1998; Rundle et al., 1998; Mariani and Harland, 1998). Weshow that VP-16 fusion proteins are useful tools to analyzegene activation during sea urchin embryogenesis.

These combined results provide evidence that the AVaxis patterning effects of b-catenin, and perhaps otherupstream components of the Wnt pathway in the seaurchin, are mediated through TCF/LEF transcriptionalregulators. Since TCF/LEFs function as transcriptional re-pressors in the absence of nuclear b-catenin it is possiblethat in animal hemisphere blastomeres TCF normally func-tions to repress vegetal-territory-specific genes. Only invegetal hemisphere cells where b-catenin translocates tohe nucleus are vegetal-specific genes activated. Whenormal embryos or animal-half embryos are treated withgents that result in ectopic nuclear b-catenin, blastomeres

that are normally fated to differentiate as ectoderm areconverted to more vegetal fates. This supports the hypoth-esis that a b-catenin gradient functions in the sea urchinembryo to establish cell fates along the AV axis (Logan etal., 1999) as has been shown for the DV avis in Xenopus(Larabell et al., 1997). However, it does not address whetherTCF/LEFs repress vegetal target genes in animal blas-tomeres independent of corepressors such as groucho orwhether these corepressors are normally displaced fromtargets by a b-catenin/TCF complex. Since potential core-pressors have not been identified in the sea urchin, thisremains to be determined.

Since b-catenin is a multifunctional protein with a role inell adhesion as well as gene regulation (Miller and Moon,997; Moon et al., 1997), altering b-catenin levels by

overexpressing cadherins may have effects secondary tothose involved in vegetal cell fate specification (Logan etal., 1999). Similarly, blocking GSK3b function to elevateb-catenin may have consequences secondary to b-cateninffects. If b-catenin functions through other regulators to

specify cell fate, then the effect of overexpressing b-cateninhould be to alter the DN Xtcf-3 phenotype. Since elevatingndogenous b-catenin with LiCl in DN Xtcf-3-injected

embryos has no effect on the animalized phenotype, TCFappears to be the principal effector of the b-catenin signal inhe sea urchin embryo. Similarly, if eliminating nuclear

b-catenin by co-injecting C-cadherin has secondary effectsinvolving adhesion, then co-injection of C-cadherin andActivated Xtcf-3 should result in a phenotype distinct fromthat observed with Activated Xtcf-3 alone. Co-injection ofC-cadherin, however, does not alter the vegetalized pheno-type induced by overexpressing Activated Xtcf-3. All theeffects of b-catenin on the initial specification of cell fateslong the AV axis in the early sea urchin embryo appear toe mediated through TCF and TCF appears to be down-tream of b-catenin.

Since we used altered forms of Xenopus Xtcf-3 as tools in n

Copyright © 2000 by Academic Press. All right

these studies we wanted to demonstrate that an endoge-nous TCF/LEF activity is present in the sea urchin embryoand mediates the b-catenin signal. Use of TCF-responsiveromoters in Xenopus and zebrafish has shown that TCFsy themselves are repressors of transcription but in combi-ation with b-catenin they activate target genes (Brannon etl., 1997; Sumoy et al., 1999). Expression of a TCF/LEF-esponsive reporter containing multimerized TCF/LEFinding sites in the sea urchin is indicative of an endoge-ous TCF and b-catenin activity in the early embryo. In

further support of the activity of an endogenous Xtcf-3/b-catenin complex in the early embryo, we elevated endoge-nous b-catenin levels in reporter-injected embryos by LiClreatment (Logan et al., 1999). Reporter activity was en-

hanced to greater than 25-fold over background in LiCl-treated embryos. A proportionate increase in reporter activ-ity was also observed with LiCl treatment when similarexperiments were performed in Xenopus. These data sug-est that elevated endogenous b-catenin resulting from LiClreatment provides increased ectopic nuclear partners forCF and that an interaction occurs between endogenousCF/LEF and b-catenin in the early sea urchin embryo. In

further support of these conclusions we have learned that aTCF/LEF homologue has been cloned from the sea urchinStrongylocentrotus purpuratus (Sp Tcf/Lef; Huang et al.,1999). Sp Tcf/Lef appears to represent the single TCF/LEFfamily member present in the S. purpuratus genome. Con-sistent with an early role in specification along the AV axis,Northern analysis revealed that a single Sp Tcf/Lef tran-script is expressed maternally and decreases beginning atthe early blastula stage (Huang et al., 1999). Interestingly,we have microinjected DN and Activated Lef-1 RNAs intosea urchin embryos at the same concentrations used withaltered forms of Xtcf-3 and observed only weak effects (datanot shown). These results suggest that the two moleculesdiffer in their ability to bind targets in sea urchins and thatthe sea urchin targets involved in early specification alongthe AV axis are responding to a homologue that moreclosely resembles TCF than LEF. In addition, expression ofthe Sp Tcf/Lef RNA is ubiquitous (Huang et al., 1999),supporting the notion that TCF may function as a repressorof vegetal-specific genes in nonvegetal cells.

Initial specification of the vegetal territory is thought tooccur at the 16-cell stage when b-catenin nuclear expres-ion is first apparent (Logan et al., 1999). In support for anarly role for b-catenin in vegetal fate specification, micro-eres in which b-catenin nuclear localization has been

blocked by overexpressing Lv-cadherin fail to induce asecond axis (Logan et al., 1999). If TCF is the partner forb-catenin in the earliest specification of vegetal fates, thenthe TCF/b-catenin complex must function at this time. Weested this by microinjecting a regulatable form of Xtcf-3.

Generally microinjected RNAs are translated soon afterhey are introduced (Tada et al., 1997) so that the encodedrotein products are often present prematurely and/or ec-opically. One way the function of exogenously introduced

uclear proteins can be regulated is to control their nuclear

s of reproduction in any form reserved.

atrt

p

ouodh(

sm

T

X

o

g

B

241Patterning of Sea Urchin Animal–Vegetal Axis

entry. Glucocorticoid receptor fusion proteins have beenused successfully in tissue culture (Hollenberg et al., 1993)and in the Xenopus embryo (Kolm and Sive, 1995; Tada etal., 1997) to exert control over the timing of transcriptionfactor activity by regulating nuclear entry. We injected anRNA, GR Xtcf-3, which encodes a fusion protein thatcombines the glucocorticoid receptor ligand binding do-main with the Xtcf-3 DNA binding domain (Fig. 1E). TheGR Xtcf-3 protein is held in the cytoplasm until the GRligand, dexamethasone, is added to embryo cultures. Theactivation domain present in the glucocorticoid ligandbinding domain (Dahlman-Wright et al., 1994) was used inplace of the b-catenin binding domain. Embryos injectedwith GR Xtcf-3 RNA at fertilization and induced withdexamethasone at various times after injection showed avegetalized phenotype (Fig. 9), but only if induced before the60-cell stage. These experiments demonstrate that nuclearXtcf-3 must be present prior to the 60-cell stage to influencethe differentiation of cells along the vegetal pathway. In-duction of nuclear TCF at later stages did not result inincreased vegetal tissue despite the presence of TCF proteinin the nucleus at the early blastula stage (data not shown).This suggests that the complex between TCF and b-cateninctivates target genes involved in specifying the vegetalerritory during early cleavage stages. Our data support aole for TCFs in the early establishment of the vegetalerritory of the sea urchin embryo.

If b-catenin/TCF interactions are important for vegetalfate specification early in development what is the signifi-cance of nuclear b-catenin observed in later stages? Oneossibility is that b-catenin acts initially through a TCF/

LEF factor to specify general vegetal identity, whereas latermore specific vegetal identities are specified by other local-ized factors or signaling molecules. The identification andcharacterization of a single TCF/LEF homologue in S.purpuratus whose expression is restricted to the earlyembryo supports this notion. Alternatively, other factorsmay repress the activity of the b-catenin/TCF complex inther cells of the embryo. Somewhat paradoxically, whenninjected micromeres are transplanted to the animal polef normal embryos and induce a second archenteron, theyo not induce nuclear b-catenin expression in the animalemisphere cells that will form the ectopic vegetal tissues

Logan et al., 1999). This is surprising because in unper-turbed embryos nuclear b-catenin appears first in the mi-cromeres, later spreading from vegetal tier to vegetal tier.b-Catenin is essential for the ability of micromeres toinduce a second axis in host embryos (Logan et al., 1999),uggesting that the initial signal emanating from micro-eres is dependent on nuclear b-catenin. Conversion of

host animal-half tissue to vegetal fates, however, does notrequire activation of nuclear b-catenin in the host cells.

his can be explained if b-catenin is required to activateinitial vegetal-territory-specific genes in responding cellsbut once these genes are activated the responding cells donot require nuclear b-catenin. Our data using an inducible

tcf-3 support the hypothesis that the initial specification

Copyright © 2000 by Academic Press. All right

f the vegetal territory requires TCF/b-catenin but subse-quent events do not. Nuclear b-catenin in vegetal cells atlater stages may have a function distinct from the specifi-cation of initial vegetal territory genes and may involvecomplexes with other nuclear proteins or displacement ofcorepressors to refine the derivatives of the vegetal blas-tomeres.

The roles of the Wnt pathway and b-catenin signaling inestablishing cell fates along embryonic axes appear to bewell conserved across species (Moon and Kimelman, 1998).What has not been established is how conserved are thedownstream targets of these signals. Since the axes in-volved are distinct in vertebrates and invertebrates, thespecific gene targets may also diverge. b-Catenin/TCF tar-et genes identified in Xenopus may not be conserved

across all phyla (Sumoy et al., 1999). Therefore a majorquestion which remains is: what are the targets ofb-catenin/TCF in establishing embryonic axes? These asyet unidentified targets will play an important role inestablishing early vegetal cell fates in the sea urchin andshould provide important clues for understanding endoder-mal and mesodermal specification not only in sea urchinsbut in all embryos.

ACKNOWLEDGMENTS

We thank P. Cserjesi (Columbia University) and W. Klein (M. D.Anderson Cancer Center) for comments on the manuscript. Wealso express our appreciation to P. Wilson (Cornell Medical School),who generously supplied the GR-Xtcf-3 construct, and to D.McClay (Duke University), C. Ettensohn (Carnegie–Mellon), andG. Wessel (Brown University), for providing antibodies. Additionalthanks to J. Massague (Memorial Sloan–Kettering) for access to theluminometer and to M. I. Arnone (Zoological Station Naples) forassistance with the initial sea urchin microinjections. This workwas supported by NSF Grant IBN9506346 and ACS GrantIRG8800610PJ8 to J.M.V, NIH Grants GM37432 and NCI-P30-CA-08784 to B.M.G., and NIH Training Grant HD07506-02NICHD tothe MBL.

REFERENCES

Aoki, M., Hecht, A., Kruse, U., Kemler, R., and Vogt, P. K. (1999).Nuclear endpoint of Wnt signaling: Neoplastic transformationinduced by transactivating lymphoid-enhancing factor 1. Proc.Natl. Acad. Sci. USA 96, 139–144.

Behrens, J., von Kries, J. P., Kuhl, M., Bruhn, L., Wedlich, D.,Grosschedl, R., and Birchmeier, W. (1996). Functional interac-tion of b-catenin with the transcription factor LEF-1. Nature 382,638–642.

Bienz, M. (1998). TCF: Transcriptional activator or repressor? Curr.Opin. Cell Biol. 10, 366–372.

Brannon, M., Gomperts, M., Sumoy, L., Moon, R., and Kimelman,D. (1997). A b-catenin/XTcf-3 complex binds to the siamoispromoter to regulate dorsal axis specification in Xenopus. GenesDev. 11, 2359–2370.

runner, E., Peter, O., Schweizer, L., and Basler, K. (1997). pangolin

encodes a LEF-1 homolog that acts downstream of Armadillo to

s of reproduction in any form reserved.

C

C

C

C

D

D

D

D

E

F

F

H

H

H

H

H

K

K

K

K

K

L

L

L

M

M

M

M

M

M

M

242 Vonica et al.

transduce the Wingless signal in Drosophila. Nature 385, 829–833.

Cadepond, F., Schweizer-Groyer, G., Segard-Maurel, I., Jibard, N.,Hollenberg, S. M., Giguere, V., Evans, R. M., and Baulieu, E. E.(1991). Heat shock protein 90 as a critical factor in maintainingglucocorticosteroid receptor in a nonfunctional state. J. Biol.Chem. 266, 5834–5841.ambridge, S. B., Davis, R. L., and Minden, J. S. (1997). Drosophilamitotic domain boundaries as cell fate boundaries. Science 277,825–828.avallo, R., Cox, R., Moline, M., Roose, J., Polevoy, G., Clevers, H.,Peifer, M., and Bejsovec, A. (1998). Drosophila tcf and grouchointeract to repress wingless signalling activity. Nature 395,604–608.levers, H., and van de Wetering, M. (1997). TCF/LEF factors earntheir wings. Trends Genet. Sci. 13, 485–489.offman, J. A., and McClay, D. R. (1990). A hyaline layer proteinthat becomes localized to the oral ectoderm and foregut of seaurchin embryos. Dev. Biol. 140, 93–104.ahlman-Wright, K., Almlof, T., McEwan, I. J., Gustafsson, J. A.,and Wright, A. P. (1994). Delineation of a small region within themajor transactivation domain of the human glucocorticoid re-ceptor that mediates transactivation of gene expression. Proc.Natl. Acad. Sci. USA 91, 1619–1623.avidson, E. H. (1989). Lineage-specific gene expression and theregulative capacities of the sea urchin embryo: A proposedmechanism. Development 105, 421–445.avidson, E. H., Cameron, R. A., and Ransick, A. (1998). Specifi-cation of cell fate in the sea urchin embryo: Summary and someproposed mechanisms. Development 125, 3269–3290.orsky, R., Moon, R., and Raible, D. (1998). Control of neural crestcell fate by the Wnt signaling pathway. Nature 396, 370–373.

mily-Fenouil, F., Ghilione, C., Llhomond, G., Lepage, T., andGache, C. (1998). GSK3b/shaggy mediates patterning along theanimal–vegetal axis of the sea urchin embryo. Development 125,2489–2498.

Ettensohn, C. A., and McClay, D. R. (1988). Cell lineage conversionin the sea urchin embryo. Dev. Biol. 125, 396–409.

erreiro, B., Artinger, M., Cho, K., and Niehrs, C. (1998). Antimor-phic goosecoids. Development 125, 1347–1359.

isher, A. L., and Caudy, M. (1998). Groucho proteins: Transcrip-tional corepressors for specific subsets of DNA-binding transcrip-tion factors in vertebrates and invertebrates. Genes Dev. 12,1931–1940.erbst, C. (1892). Experimentalle untersuchungen uber den ein-fluss der veran derten chemischen zusammenstzung des umge-benden mediums auf die entwicklung der thiere I teil. Versuchean seeigeleiren. Z. Wiss. Zool. 55, 446–518.ollenberg, S., Cheng, P., and Weintraub, H. (1993). Use of aconditional MyoD transcription factor in studies of MyoD trans-activation and muscle determination. Proc. Natl. Acad. Sci. USA90, 8028–8032.orstadius, S. (1973). “Experimental Embryology of Echinoderms.”Clarendon, Oxford.uang, L., Li, X., El-Hodiri, H. M., Wkramanayake, A. H., andKlein, W. H. (1999). Involvement of Tcf/Lef in establishing celltypes along the animal–vegetal axis of sea urchins. Dev. GenesEvol., in press.uber, O., Korn, R., Mclaughlin, J., Ohsugi, M., Herrmann, B. G.,and Kemler, R. (1996). Nuclear localization of b-catenin by

interaction with transcription factor LEF-1. Mech. Dev. 59, 3–10.

Copyright © 2000 by Academic Press. All right

engaku, M., Capdevila, J., Rodgriguez-Esteban, C., De la Pena, J.,Johnson, R. L., Belmonte, J. C. I., and Tabin, C. J. (1998). DistinctWNT pathways regulating AER formation and dorsoventralpolarity in the chick limb bud. Science 280, 1274–1277.essler, D. S. (1997). Siamois is required for formation of Spe-mann’s organizer. Proc. Natl. Acad. Sci. USA 94, 13017–13022.lein, P. S., and Melton, D. A. (1996). A molecular mechanism forthe effect of lithium on development. Proc. Natl. Acad. Sci. USA93, 8455–8459.olm, P. J., and Sive, H. L. (1995). Efficient hormone-inducibleprotein function in Xenopus laevis. Dev. Biol. 171, 267–272.

orinek, V., Barker, N., Morin, P. J., van Wichen, D., de Weger, R.,Kinzler, K. W., Vogelstein, B., and Clevers, H. (1997). Constitu-tive transcription activation by a b-catenin-Tcf complex in APC2/2 colon carcinoma. Science 275, 1784–1787.

Lallier, R. (1975). In “The Sea Urchin Embryo: Biochemistry andMorphogenesis” (G. Czihak, Ed.), pp. 473–507. Springer, NewYork.

Larabell, C. A, Torres, M., Rowning, B. A., Yost, C., Miller, J. R.,Wu, M., Kimelman, D., and Moon, R. T. (1997). Establishment ofthe dorso-ventral axis in Xenopus embryos is presaged by earlyasymmetries in beta-catenin that are modulated by the Wntsignaling pathway. J. Cell Biol. 136, 1123–1136.

Lemaire, P., Darras, S., Caillol, D., and Kodjabachian, L. (1998). Arole for the vegetally expressed Xenopus gene Mix.1 in endodermformation and in the restriction of mesoderm to the marginalzone. Development 125, 2371–2380.

evanon, D., Goldstein, R., Bernstein, Y., Tang, H., Goldenberg, D.,Stifani, S., Paroush, Z., and Groner, Y. (1998). Transcriptionalrepression by AML1 and LEF-1 is mediated by the TLE/Grouchocorepressors. Proc. Natl. Acad. Sci. USA 95, 11590–11595.

ogan, C. Y., Miller, J. R., Ferkowicz, M. J., and McClay, D. R.(1999). Nuclear bcatenin is required to specify vegetal cell fatesin the sea urchin embryo. Development 126, 345–357.

ove, J. J., Li, X., Case, D. A., Giese, K., Grosschedl, R., and Wright,P. E. (1995). Structural basis for DNA bending by the architec-tural transcription factor LEF-1. Nature 376, 791–795.ao, C. A., Wikramanayake, A. H., Gan, L., Chuang, C. K.,Summers, R. G., and Klein, W. H. (1996). Altering cell fates in seaurchin embryos by overexpressing SpOtx, an orthodenticle-related protein. Development 122, 1489–1498.ariani, F. V., and Harland, R. M. (1998). XBF-2 is a transcriptionalrepressor that converts ectoderm into neural tissue. Develop-ment 125, 5019–5031.cMahon, A. P., Flytzsanis, C. N., Hough-Evans, B. R., Katula,K. S., Britten, R. J., and Davidson, E. H. (1985). Introduction ofcloned DNA into sea urchin egg cytoplasm, replication andpersistence during embryogenesis. Dev. Biol. 108, 420–430.iller, J. R., and McClay, D. R. (1997). Characterization of the roleof cadherin in regulating cell adhesion during sea urchin devel-opment. Dev. Biol. 192, 323–339.iller, J. R., and Moon, R. T. (1996). Signal transduction throughb-catenin and specification of cell fate during embryogenesis.Genes Dev. 10, 2527–2539.olenaar, M., van de Wetering, M., Oosterwegel, M., Peterson-Maduro, J., Godsave, S., Korinek, V., Roose, J., Destree, O., andClevers, H. (1996). Xtcf-3 transcription factor mediatesb-catenin-induced axis formation in Xenopus embryos. Cell 86,391–399.oon, R. T., Brown, J. D., and Torres, M. (1997). WNTs modulatecell fate and behavior during vertebrate development. Trends

Genet. 13, 157–162.

s of reproduction in any form reserved.

N

O

R

R

R

R

S

T

T

T

T

v

v

V

W

W

W

W

W

243Patterning of Sea Urchin Animal–Vegetal Axis

Moon, R. T., and Kimelman, D. (1998). From cortical rotation toorganizer gene expression, toward a molecular explanation ofaxis specification in Xenopus. BioEssays 20, 536–545.ocente-McGrath, C., McIssac, R., and Ernst, S. G. (1991). Alteredcell fate in LiCl-treated sea urchin embryos. Dev. Biol. 147,445–450.osterwegel, M., van de Wetering, M., Dooijes, D., Klomp, L.,Winoto, A., Georgopoulos, K., Meijlink, F., and Clevers, H.(1991). Cloning of murine TCF-1, a T cell-specific transcrip-tion factor interacting with functional motifs in the CD3-epsilon and T cell receptor alpha enhancers. J. Exp. Med. 173,1133–1142.ansick, A., and Davidson, E. H. (1993). A complete second gutinduced by transplanted micromeres in the sea urchin embryo.Science 259, 1134–1138.

oose, J., Molenaar, M., Peterson, J., Hurenkamp, J., Brantjes, H.,Moerer, P., van der Wetering, M., Destree, O., and Clevers, H.(1998). The Xenopus Wnt effector of Xtcf-3 interacts with Grou-cho related transcriptional repressors. Nature 395, 608–612.undle, C. H., Macias, M. P., Yueh, Y. G., Gardner, D. P., andKappen, C. (1998). Transactivation of Hox gene expression in aVP16-dependent binary transgenic mouse system. Biochim. Bio-phys. Acta 1398, 164–178.usch, J., and Levine, M. (1997). Regulation of a dpp target gene inthe Drosophila embryo. Development 124, 303–311.

umoy, L., Kiefer, J., and Kimelman, D. (1999). Conservation ofintracellular Wnt signaling components in dorsal–ventral axisformation in zebrafish. Dev. Genes Evol. 209, 48–58.ada, M., O’Reilly, M. A., and Smith, J. C. (1997). Analysis ofcompetence and of Brachyury autoinduction by use of hormone-inducible Xbra. Development 124, 2225–2234.

ravis, A., Amsterdam, A., Belanger, C., and Grossschedl, R.(1991). LEF-1, a gene encoding a lymphoid-specific protein withan HMG domain, regulates T-cell receptor alpha enhancer func-tion. Genes Dev. 5, 880–894.riezenberg, S. J., Kingsbury, R. C., and McKnight,S. L. (1988).Functional dissection of VP16, the trans-activator of herpessimplex virus immediate early gene expression. Genes Dev. 2,

718–729.

Copyright © 2000 by Academic Press. All right

urner, D. L., and Weintraub, H. (1994). Expression of achaete-scute homolog 3 in Xenopus embryos converts ectodermal cellsto neural fate. Genes Dev. 8, 1434–1447.

an de Wetering, M., Oosterwegel, M., Dooijes, D., and Clevers, H.(1991). Identification and cloning of TCF-1, a T lymphocyte-specific transcription factor containing a sequence-specific HMGbox. EMBO J. 10, 123–132.

an de Wetering, M., Cavallo, R., Dooijes, D., van Beest, M., van Ex,J., Ypma, A., Jursh, D., Jones, T., and Bejsovec, A. (1997).Armadillo coactivates transcription driven by the product of theDrosophila segment polarity gene dTCF. Cell 88, 789–799.

enuti, J. M., Kozlowski, M. T., Gan, L., and Klein, W. H. (1993).Developmental potential of muscle cell progenitors and the myo-genic factor SUM-1 in the sea urchin embryo. Mech. Dev. 41, 3–14.aterman, M. L., Fischer, W. H., and Jones, K. A. (1991). Athymus-specific member of the HMG protein family regulatesthe human T cell receptor C alpha enhancer. Genes Dev. 5,656–669.essel, G. M., and McClay, D. R. (1985). Sequential expression ofgerm-layer specific molecules in the sea urchin embryo. Dev.Biol. 111, 451–463.essel, G. M., Zhang, W., and Klein, W. H. (1990). Myosin heavychain accumulates in dissimilar cell types of the macromerelineage in the sea urchin embryo. Dev. Biol. 140, 447–454.ikramanayake, A. H., and Klein, W. H. (1995). Autonomous andnon-autonomous differentiation of ectoderm in different seaurchin species. Development 121, 1497–1505.ikramanayake, A. H., Huang, L., and Klein, W. H. (1998).b-Catenin is essential for patterning the maternally specifiedanimal–vegetal axis in the sea urchin embryo. Proc. Natl. Acad.Sci. USA 95, 9343–9348.

Yost, C., Torres, M., Miller, J. R., Huang, E., Kimelman, D., and Moon,R. T. (1996). The axis-inducing activity, stability, and subcellulardistribution of beta-catenin is regulated in Xenopus embryos byglycogen synthase kinase-3. Genes Dev. 10, 1443–1454.

Received for publication August 2, 1999Revised October 27, 1999

Accepted October 27, 1999

s of reproduction in any form reserved.