lsevier.com/locate/ydbio

Developmental Biology 2

Genomes & Developmental Control

Tcf- and Vent-binding sites regulate neural-specific geminin

expression in the gastrula embryo

Jennifer J. Taylor a, Ting Wang b, Kristen L. Kroll a,*

a Department of Molecular Biology and Pharmacology, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO 63110, USAb Department of Genetics, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO 63110, USA

Received for publication 22 August 2005, revised 12 October 2005, accepted 14 October 2005

Available online 9 December 2005

Abstract

Vertebrate neural development has been extensively investigated. However, it is unknown for any vertebrate gene how the onset of neural-

specific expression in early gastrula embryos is transcriptionally regulated. geminin expression is among the earliest markers of dorsal, prospective

neurectoderm at early gastrulation in Xenopus laevis. Here, we identified two 5V sequence domains that are necessary and sufficient to drive

neural-specific expression during gastrulation in transgenic Xenopus embryos. Each domain contained putative binding sites for the transcription

factor Tcf, which can mediate Wnt signaling and for Vent homeodomain proteins, transcriptional repressors that mediate BMP signaling. Results

from embryos transgenic for constructs with mutated Tcf or Vent sites demonstrated that signaling through the Tcf sites was required for dorsal-

specific expression at early gastrulation, while signaling through the Vent sites restricted geminin expression to the prospective neurectoderm at

mid-gastrulation. Consistent with these results, geminin 5V regulatory sequences and endogenous Xgem responded positively to Wnt signaling and

negatively to BMP signaling. The two 5V sequence domains were also conserved among geminin orthologs. Together, these results demonstrate

that signaling through Tcf and Vent binding sites regulates transcription of geminin in prospective neurectoderm during gastrulation.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Transcription; Geminin; Xenopus; Transgenic; Wnt; BMP; Tcf; Vent; Neural

Introduction

Formation of the vertebrate neural plate is initiated by the

process of neural induction, which subdivides ectoderm into

prospective neural and non-neural domains. Studies of neural

induction began with the demonstration that transplantation of

the dorsal blastopore lip (dorsal mesendoderm or organizer)

of a gastrula stage newt embryo to the ventral side of another

embryo elicited formation of a secondary body axis, with

secondary neural tissue induced from the host embryo

(Spemann and Mangold, 1924). Structures roughly equivalent

to the amphibian organizer in other vertebrates (the shield in

zebrafish and the node in chick and mouse) have similar

inductive properties (Beddington, 1994; Shih and Fraser,

1996; Waddington, 1930). Numerous studies have since

sought to identify the molecular signals that induce neural

fate.

0012-1606/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.ydbio.2005.10.047

* Corresponding author. Fax: +1 314 362 7058.

E-mail address: kkroll@wustl.edu (K.L. Kroll).

Bone morphogenetic protein (BMP) signaling is a prominent

candidate regulatory pathway for determining ectodermal fate.

Over-expression of dominant negative receptors in Xenopus

ectoderm results in neural differentiation (Hawley et al., 1995;

Hemmati-Brivanlou and Melton, 1992; Xu et al., 1995). bmp4

mRNA is also downregulated in the future neural ectoderm as the

organizer forms (Fainsod et al., 1997; Wilson and Hemmati-

Brivanlou, 1995). Additionally, BMP antagonists including

Follistatin, Noggin and Chordin, are expressed in the organizer

and induce neural tissue (Hemmati-Brivanlou et al., 1994; Lamb

et al., 1993; Sasai et al., 1995). Despite the apparent importance

of organizer-derived BMP antagonists in specifying neural cell

fate, in some cases neural tissue can formwhen organizer tissue is

genetically or surgically ablated (Davidson et al., 1999; Klingen-

smith et al., 1999; Shih and Fraser, 1996; Smith and Schoenwolf,

1989) and mice mutant for BMP antagonists form neural plates

(Bachiller et al., 2000; Matzuk et al., 1995; McMahon et al.,

1998). In the early chick embryo, BMP antagonists also appear

insufficient to induce neural tissue (Connolly et al., 1997; Levin,

1998; Storey et al., 1992; Streit et al., 1998; Wilson et al., 2000).

89 (2006) 494 – 506

www.e

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 495

In contrast, recent studies have re-addressed and confirmed the

importance of BMP antagonism for dorsal and neural develop-

ment in Xenopus (Delaune et al., 2005; Khokha et al., 2005;

Reversade et al., 2005; Wawersik et al., 2005).

Additional signaling pathways also regulate neural cell fate.

In avian and ascidian embryos, FGF (Fibroblast Growth

Factor) signaling induces neural markers, and blocking FGF

signaling inhibits neural development (Darras and Nishida,

2001; Hudson et al., 2003; Lemaire et al., 2002; Miya and

Nishida, 2003; Streit et al., 2000; Wilson et al., 2000). Work in

Xenopus also suggests that FGF signaling may facilitate neural

induction (Delaune et al., 2005; Hongo et al., 1999; Kessel and

Pera, 1998; Mason, 1996; Wawersik et al., 2005), but its exact

role remains unclear. Deficiencies in neural tissue following

inhibition of FGF signaling may reflect a failure of underlying

mesoderm to be properly formed, maintained, or patterned

during gastrulation. Furthermore, in Xenopus and zebrafish,

FGF signaling appears more critical for formation of posterior

than anterior neural tissue (Haremaki et al., 2003; Koshida et

al., 2002; Lamb and Harland, 1995; Launay et al., 1996;

Rentzsch et al., 2004; Ribisi et al., 2000).

Studies have also implicated Wnt signaling in defining

presumptive neurectoderm, in both positive and negative

regulatory roles. In Xenopus, Wnt ligands (xWnt8, xWnt3a,

andmWnt8), the receptor xFrz8, or intracellular components that

potentiate Wnt signaling (xDsh, dnGSK3, or a constitutively

active beta-catenin) can induce neural tissue and suppress bmp4

expression (Baker et al., 1999). Early Wnt signaling also

correlates with expression of the BMP antagonists noggin and

chordin in the dorsal animal and marginal zone tissues of

blastulae (De Robertis and Kuroda, 2004; Wessely et al., 2001).

In zebrafish, Beta-catenin suppresses bmp2b through activation

of bozozok, and bozozok mutants have defects in the anterior

neurectoderm (Fekany-Lee et al., 2000; Leung et al., 2003).

Therefore, suppression of BMP signaling by the Wnt pathway

may be a conserved mechanism mediating dorsal and/or neural

development. Conversely, in avian embryos Wnt signaling

inhibits neural fates (Wilson et al., 2001). These apparently

conflicting roles ofWnt signaling inXenopus and avian embryos

may reflect temporal differences in the roles of Wnt signaling

during development. InXenopus,Wnt signaling prior to themid-

blastula transition (MBT) directs dorsal development, but post-

MBT activity promotes ventral development (Hamilton et al.,

2001; Roel et al., 2002; Schneider et al., 1996; Stern, 2005).

Alternatively, Xenopus and avian embryos may specify neural

fate using differential mechanisms.

Understanding the transcriptional regulation of neural gene

expression may clarify the roles of the BMP, FGF, or Wnt

signaling pathways in neural development. The development of

transgenic approaches for manipulating the Xenopus embryo

and identification of genes that mark the early formation of

prospective neurectoderm now enable us to examine this issue.

One such gene, whose expression marks presumptive neural

tissue at early gastrulation, is geminin. geminin encodes a novel

coiled-coil protein with orthologs in vertebrates, Drosophila,

and C. elegans. The roles of Geminin in regulating both neural

cell fate and the fidelity of DNA replication were originally

defined in Xenopus (Kroll et al., 1998; McGarry and Kirschner,

1998). Functional studies in Xenopus and Drosophila have

shown that Geminin is required for neural and/or neuronal cell

fate specification (Kroll et al., 1998; Quinn et al., 2001).

Recently, we also demonstrated that Geminin is essential for

maintaining the neuronal progenitor state and controlling the

timing of neuronal differentiation (Seo et al., 2005).

Expression of geminin is among the earliest molecular

markers of prospective neural tissue in Xenopus. During

cleavage and blastula stages, maternal geminin RNA and

protein are distributed throughout the animal hemisphere (Kroll

et al., 1998). At late blastula through early gastrula stages,

geminin is upregulated in dorsal ectoderm, while ventral

transcripts are rapidly lost, generating a domain of expression

by the onset of gastrulation (st. 10� to 10+) that marks the

future neural plate. How the onset of neural-specific gene

expression at gastrulation is transcriptionally regulated was

previously unknown for any vertebrate gene. Here, we used

transgenic and computational approaches to determine

sequences required for regulating geminin transcription in the

prospective neurectoderm during gastrulation.

Materials and methods

Plasmid construction

Sequences 5V of the transcription start site of the human geminin (hgem)

locus were amplified by polymerase chain reaction (PCR) with Advantage 2

Polymerase (Clontech). PAC Clone RP3-369A17 containing the hgem genomic

locus, was used as template (Sanger Centre) to amplify 5.1 kb of upstream

sequence. The PCR product was cloned into pRAREeGFP (Davis et al., 2001).

Using [�4694 to +436] as template, plasmids containing 5V sequencesubdomain combinations or deletions were created by PCR with high fidelity

polymerase. PCR products were digested and ligated into pRAREeGFP,

pGL3basic (Promega), or pRAREsCMV. The pRARE vectors (Davis et al.,

2001) are derived from pCS2 (Rupp et al., 1994), by removal of the sCMV IE94

promoter and replacement with a polylinker, followed by eGFP (Clontech) and

the SV40 virus late polyadenylation signal. pRAREsCMVadditionally contains

the simian cytomegalovirus minimal promoter (sCMV) and was used for

constructs lacking the hgem sequences adjacent to the transcription start site.

pGL3basic is a promoterless vector driving expression of firefly luciferase.

Numerous pRAREeGFP- and pGL3basic-based constructs were tested and the

same pattern of reporter expression was obtained regardless of vector. Each

empty vector yielded few embryos showing any spatially-restricted reporter

expression upon introduction into transgenic embryos (pRAREeGFP shown in

Figs. 2J, K). The few false positives may have resulted from non-specific

trapping of in situ probe in the archenteron or may reflect effects of a particular

chromosomal integration site on the expression pattern in an individual

transgenic embryo. Subcloning products were verified by DNA sequencing.

Mutagenesis primers changed CAAAG to TGCCT (for Tcf), AGA-

CAAATTCC to GTGAGGGCGAA (for XVent1), or CATTAGTAAT to

TGCGGACCCG (for XVent2). Using construct [�3783 to �2920] + [�2425

to �1580] as template, PCR was performed with mutagenesis primers

containing the above changes. Mutagenesis was performed using the Quick

Change Site-directed Mutagenesis kit protocol (Stratagene). Mutations were

verified by DNA sequencing.

All primers are available upon request.

Transgenic X. laevis embryos

Transgenic embryos were generated as described (Kroll and Amaya, 1996)

with modifications. For each reaction, approximately 6.25� 105 X. laevis sperm

nuclei were incubated with 1 Ag linearized DNA in a total volume of 10 Al for 5

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506496

min at room temperature (approximately 23-C). To this reaction, 2 Al egg extract,23 Al sperm dilution buffer (SDB), 2 Al 100 mMMgCl2, and 0.5 Al NotI (1 unit)were added. This reaction was incubated at room temperature for 10 min. The

reaction mixture was then diluted 1:40 in SDB and injected using a Harvard 11

infusion pump. Embryos were collected at the stages indicated (Nieuwkoop and

Faber, 1967), fixed inMEMFA (0.1MMOPS/2 mMEGTA/1 mMMgSO4/3.7%

Formaldehyde) for 2 h, and stored in 100% EtOH at �20-C. To prepare DNA

constructs for transgenesis, pRAREeGFP or pRAREsCMV-based constructs and

pCMVeGFP were linearized with NotI; pGL3basic-based constructs were

linearized with SalI. The linearized DNA was placed over a PCR purification

column (Qiagen) and diluted to 250 ng/Al with water.

In situ hybridization

In situ hybridizations were performed as described (Harland, 1991) with

modifications. Vitelline envelopes were not removed from embryos and

archenterons were not punctured prior to fixation in MEMFA. Embryos were

rehydrated through ethanol washes. A single wash with acetic anhydride was

performed. Following hybridization, embryos were washed with 2� SSC three

times and then 0.2� SSC without CHAPS twice. RNase treatment was omitted.

Probes were generated by linearization of plasmid templates (CS2PeGFP

with NcoI, CS2-FLAGhgem with EcoRI, pCS2-XgemL with EcoRI and pCS2-

Luc with BamHI) and transcription with the appropriate RNA polymerase (T3

or T7) in the presence of digoxigenin-11-UTP (Roche). Hybridization signals

were detected using alkaline phosphatase (AP)-conjugated anti-digoxigenin

antibodies (Roche) with Nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl-

phosphate (NBT/BCIP; Roche) as the substrate.

Microinjection

Capped mRNA was created by linearization of pCS2 + beta-catenin with

NotI , pCS2 + noggin with EcoRI, pCS-BMP4 with NotI, pCS2P-XVent1 with

NotI, pCSFAPS-Xvent2 with PmeI, VPXvent-1 with SfiI, and VPXvent-2 with

NotI and transcription with T3 (VPXvent-1 and VPXvent-2) or SP6 RNA

polymerase using the mMessage Machine kit (Ambion) according to the

manufacturer’s instructions.

Transgenic embryos were cultured in 0.2� Marc’s Modified Ringers

(MMR) + 4% Ficoll + gentamycin (100 Ag/ml) to the 2-cell stage. 125 pg beta-

catenin mRNA, 125 pg noggin mRNA, or 250 pg bmp4 mRNAwere injected

in a volume of 10 nl into one cell. Embryos were cultured in 0.2� MMR +

gentamycin (100 Ag/ml) to late gastrula (st. 12), then fixed in MEMFA for

2 h and stored in ethanol at �20-C.

Embryos produced by in vitro fertilization were injected at the 2-cell stage

with 20 ng beta-catenin morpholino oligonucleotide, 25 pg Xvent1 RNA, or

125 pg Xvent2 mRNA and 62.5 pg beta-galactosidase mRNA into one cell in a

volume of 10 nl. Embryos raised to the 8-cell stage were injected ventrally with

100 pg VPXvent-1 mRNA or 400 pg VPXvent-2 mRNA, with 50 pg beta-

galactosidase mRNA into one cell in a volume of 8 nl. Embryos raised to the

32-cell stage were injected in a single A-tier blastomere (Dale and Slack, 1987)

with 5 pg Xvent1 mRNA, 25 pg Xvent2 mRNA, 25 pg VPXvent-1 mRNA or

100 pg VPXvent-2 mRNA with 12.5 pg beta-galactosidase mRNA in a total

volume of 2 nl. All embryos were cultured in 0.2� MMR + gentamycin (100

Ag/ml) to mid-gastrula (st. 11.5). They were fixed in MEMFA for 1 h, stained

with 5-bromo-4chloro-3-indolyl-h-d-galactopyrandoside (X-gal) for approxi-

mately 3 h at 37-C and fixed in MEMFA for an additional hour. They were then

stored in ethanol at �20-C.

Luciferase assays

Luciferase assays were performed as described using the Dual Luciferase

Reporter Assay System (Promega) (Osada et al., 2000). Embryos were injected

with 100 pg firefly luciferase reporter (pGL3basic constructs) and 2 pg control

Renilla luciferase plasmid (pRL-TK) into one cell at the 2-cell stage. Some samples

were additionally coinjected with 12.5 pg beta-catenin mRNA or 12.5 pg noggin

mRNA. Embryos were cultured to mid-gastrula (st. 11). Pools of 3 embryos were

collected in triplicate for each reporter construct. Each pool was homogenized in

100 Al 1� passive lysis buffer (Promega). Following centrifugation for one min, 5

Al of supernatantwas assayed using 50Al LAR II (Promega) followedby50Al Stop

and Glo Reagent (Promega) using a Berhold luminometer. Firefly luciferase

activity was normalized to Renilla luciferase activity.

Results

5.1 kb of 5V regulatory sequence recapitulates the endogenous

Xgem expression pattern

To identify 5V sequence elements that regulate the dorsal,

neural-specific expression pattern of geminin, we used PCR to

isolate 5.1 kb of sequence from human geminin (hgem),

including 4694 bp upstream from the transcription start site

and 436 bp of the 5V UTR and cloned these sequences into

promoterless eGFP and luciferase reporter vectors (seeMaterials

and Methods). We generated transgenic X. laevis embryos for

these constructs, [�4694 to +436], and collected gastrula (st.

10.5) through tadpole (st. 35) stage embryos. The pattern of

reporter expression driven by these 5V sequences was analyzedby in situ hybridization. In situ hybridization for eGFP mRNA is

usually a more sensitive method than fluorescence for detecting

transgene expression in early stage Xenopus embryos because

yolk proteins produce interfering auto-fluorescence. In addition,

eGFP protein is often not accumulated to levels detectable by

fluorescence by early gastrulation in transgenic embryos. In

gastrulae, reporter expression was enriched in dorsal ectoderm

and the reporter pattern was similar to that of endogenous

Xenopus laevis geminin (Xgem) (Fig. 1, compare A, D and

Supplemental Fig. 1). 49.0% of embryos showed dorsal-specific

expression (N = 463). To determine a baseline percentage of

transgenic embryos produced under our experimental condi-

tions, we generated transgenic embryos carrying pCMVeGFP, a

plasmid with eGFP driven by the simian CMV enhancer/

promoter. 37.7% of gastrulae carrying pCMVeGFP expressed

eGFP as determined by in situ hybridization (N = 61).

Comparison with this baseline suggests that nearly all embryos

transgenic for [�4694 to +436] drove dorsal-specific reporter

expression. In contrast, few embryos transgenic for empty

reporter vectors expressed reporter with any spatial specificity

(6%, N = 157) (see Materials and Methods). At tailbud and

tadpole stages, transgenic embryos carrying [�4694 to +436]

expressed the reporter in neural structures including the brain,

spinal cord, eye, otic vesicle, and branchial arches, resembling

endogenous Xgem expression (Figs. 1B, C versus E, F).

To determine if sequences outside of the 5.1 kb tested

contributed to the early expression pattern of geminin, we also

generated embryos transgenic for the entire PAC clone

containing hgem. The pattern of hgem expression in these

embryos did not differ from reporter expression driven by

[�4694 to +436] in gastrula or in tailbud stage embryos (data

not shown). Thus, 5.1 kb of upstream sequence from hgem

recapitulates the early expression pattern of Xgem.

Two 5V sequence domains are necessary and sufficient to

recapitulate the endogenous Xgem expression pattern

To determine which sequences between �4694 and +436

are necessary to recapitulate the endogenous geminin

Fig. 1. Embryos transgenic for [�4694 to +436] express eGFP in a pattern that recapitulates endogenous geminin expression. Embryos were analyzed by in situ

hybridization for eGFP (A–C) or Xgem (D–F) at st. 11.5 (A, D), st. 23 (B, E) and st. 35 (C, F). (A, D) Lateral views, dorsal up and anterior facing right. (B, C, E, F)

Lateral views, dorsal up. (B, E) Embryos are curved with anterior structures facing out.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 497

expression pattern, 5V and internal deletion series were

constructed (Figs. 2A and 3A). Transgenic embryos carrying

each construct were created and analyzed for reporter

expression by in situ hybridization. For each construct,

the expression pattern obtained and the frequency of

Fig. 2. 5V deletion series identifies two sequence domains that regulate dorsal-specific

analyzed by in situ hybridization at st. 11.5 for the pattern of eGFP expression. (B, D,

views with dorsal up. (C, E, G, I, K) Numbers represent the fraction of total embryo

gastrulae showing dorsal-specific reporter expression were

compared to data for transgenic embryos carrying the entire

5.1 kb of 5V sequence (Figs. 2B, C). Embryos carrying the

reporter vector alone were generated as negative controls

(Figs. 2J, K).

expression. Transgenic embryos carrying the constructs schematized in Awere

F, H, J) Side views, posterior facing right, and dorsal up. (C, E, G, I, K) Posterior

s carrying the indicated construct that displayed the expression pattern shown.

Fig. 3. Internal deletion series confirms the necessity of two 5V regulatory domains. Transgenic embryos carrying the constructs schematized in Awere analyzed by in

situ hybridization at st. 11.5 for the pattern of reporter expression. (B, D, F, H, J) Side views with posterior to the right and dorsal up. (C, E, G, I, K) Posterior views

with dorsal up. (B–G, J, K) In situ for eGFP. (H, I) In situ for firefly luciferase. (C, E, G, I, K) Numbers represent the fraction of total embryos carrying the indicated

construct that displayed the expression pattern shown.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506498

We initially analyzed transgenic embryos carrying 5Vdeletionseries constructs (Fig. 2A). Embryos transgenic for either

construct [�4175 to +436] or [�3783 to +436] expressed the

reporter in a dorsal-specific pattern similar to the positive

control [�4694 to +436] (Fig. 2A, data not shown, and D, E

versus B, C). In contrast, the construct [�3274 to +436] drove

eGFP expression in both ventral and dorsal ectoderm (Figs.

2A, F, G). Thus, sequences between �3783 and �3274 contain

a domain that regulates the dorsal-specific expression pattern.

Embryos transgenic for [�1611 to +436] had weak expression

with no spatially restricted pattern (Figs. 2A, H, I). This result

suggested that sequences between �3274 and �1611 contain a

second regulatory domain. The smallest constructs in the 5Vdeletion series, [�1135 to +436] and [�49 to +436], drove

little eGFP expression with no spatially restricted pattern (data

not shown). Results of the 5V deletion series identified two

sequence domains necessary for regulating geminin expression.

Internal deletions confirmed the necessity of these two

regulatory domains. Deletion of �3783 to �3274 (D[�3783 to

�3274]) or of �2404 to �1611 (D[�2404 to �1611]) resulted

in ubiquitous reporter expression (Figs. 3A–C, H, I). Although

the internal deletion construct D[�2404 to �1611] drove

ubiquitous expression, the 5V deletion construct [�1611 to

+436] resulted in weak or absent reporter expression (Figs. 3H,

I versus Figs. 2H, I), suggesting that the two regulatory domains

act additively to regulate expression. In contrast, other internal

deletions, D[�3286 to�2921] and D[�2905 to�2425], did not

disrupt the dorsal-specific reporter expression, demonstrating

that sequences between �3286 to �2425 are not necessary for

regulating the early expression pattern (Figs. 3A, D–G). We

also deleted �1591 to +436 and cloned the remaining 5Vsequence into a reporter vector containing a heterologous

minimal promoter (pRAREsCMV). This construct drove

dorsal-specific expression, indicating that sequences adjacent

to the transcription start site do not control the spatial pattern of

reporter expression (Figs. 3A, J, K). Of the 5.1 kb of sequence

originally assayed, �3783 to �3274 and �2404 to �1611

contain domains necessary for regulating the early dorsal-

specific expression of geminin.

To determine whether these two 5V sequence domains were

sufficient to drive dorsal-specific reporter expression, we

subcloned these regions in combination into pRAREsCMV.

Construct [�3783 to �2920] + [�2425 to �1580] drove

reporter expression specifically in the dorsal ectoderm (Figs.

4A–C and Supplemental Fig. 1). This construct also drove

neural-specific expression at tailbud stages (st. 23) (data not

Fig. 4. Sufficiency of sequence domains to drive or restore dorsal-specific expression. Transgenic embryos carrying the constructs schematized in A were analyzed

by in situ hybridization at st. 11.5 for the pattern of reporter expression. [�3783 to �2920] + [�2425 to �1580] and [�3783 to �2920] + [�2289 to �2020] were

subcloned into pRAREsCMV. D[�2404 to �1611] + [�2289 to �2020] and D[�2404 to �1611] + [�2025 to �1775] were subcloned into pGL3basic. (B, D, F, H)

Side views with posterior to the right and dorsal up. (C, E, G, I) Posterior views with dorsal up. (B–E) In situ for eGFP. (F– I) In situ for firefly luciferase. (C, E, G,

I) Numbers represent the fraction of total embryos carrying the indicated construct that displayed the expression pattern shown.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 499

shown). Thus, sequences �3783 to �2920 and �2425 to

�1580 contain elements that are necessary (Figs. 3B, C and H,

I) and sufficient (Figs. 4B, C) to recapitulate the early dorsal-

specific expression of endogenous Xgem.

Analysis of forty constructs in transgenic embryos defines two

regulatory domains: �3783 to �3336 and �2289 to �2020

To determine which sequences within these two domains are

critical for regulating the dorsal-specific pattern, we tested a

number of constructs containing subdomains of sequence

between �3783 to �2920 and �2425 to �1580. For example,

construct [�3783 to �2920] + [�2289 to �2020] drove dorsal-

specific expression (Figs. 4A, D, E). Additionally, when �2289

to �2020 was added back within D[�2404 to �1611], the

dorsal pattern of expression was restored (compare Figs. 3A, H,

I and Figs. 4A, F, G). Conversely, when added back to the same

deletion, the subdomain �2025 to �1775 did not restore the

dorsal-specific expression pattern (Figs. 4A, H, I). Thus,�2289

to �2020 was sufficient to restore dorsal-specific expression to

D[�2404 to�1611] and, in combination with�3783 to�2920,

directed dorsal-specific expression. Several similar constructs

were tested (Supplemental Fig. 2). Smaller (approximately 90 to

160 bp) subdomains between �3783 and �3274 or �2289 and

�2020 did not restore dorsal specific expression when added

back to either D[�3783 to �3274] or D[�2404 to �1611]. We

also tested small sequence domains in combination to determine

if they were sufficient to drive dorsal-specific expression. For

example, construct [�3783 to �3443] + [�2281 to �2204],

drove dorsal-specific expression in very few embryos (7.7%

N = 130; data not shown). Our inability to further restrict

subdomain size suggests that the regulatory sequences span

multiple necessary transcription factor binding sites. Further-

more, the sequence context surrounding a site may contain

binding sites for additional regulatory proteins or cofactors, may

provide appropriate spacing to enable transcription factor

protein–DNA interactions to occur, or may in some other

sequence-specific manner affect the function of a transcription

factor binding to a neighboring site.

A total of forty constructs were tested in transgenic Xenopus

embryos for the ability to drive dorsal-specific reporter

expression (Supplemental Fig. 2). Comparing the results for

all forty constructs narrowed the two functional domains to

�3783 to �3336 and �2289 to �2020. When both domains

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506500

were present, reporter expression was dorsal-specific in an

average of 41.7 T 4.2 percent of embryos, similar to results

obtained for 5.1 kb of sequence, indicating that most transgenic

embryos drove reporter expression in a dorsal-specific pattern.

Identification of putative binding sites within the geminin 5Vregulatory domains

The transgenic analyses described above identified two 5Vsequence domains that are necessary and sufficient for

regulating neural-specific geminin expression. To identify

potential transcription factor binding sites within these

domains, we used the PATSER program (G. Hertz and G.

Stormo, unpublished; http://ural.wustl.edu) to compare gemi-

nin 5V sequences to binding site matrices from the TRANSFAC

database (Hertz and Stormo, 1999; Matys et al., 2003).

PATSER assigns each potential site a score proportional to

the calculated free energy of a protein–DNA interaction (Hu et

al., 2004). We searched for sequences with high scores,

focusing on transcription factors expressed prior to or during

gastrulation in Xenopus and factors that could potentially

mediate BMP, FGF, or Wnt signaling. Applying these criteria,

we found that each regulatory domain contains a canonical

binding site for a Vent homeodomain protein. Vent proteins are

transcriptional repressors, which promote ventral and epider-

mal cell fates in response to BMP signaling (Gawantka et al.,

1995; Imai et al., 2001; Melby et al., 1999; Onichtchouk et al.,

1996; Onichtchouk et al., 1998; Schmidt et al., 1996). Each

domain also contains canonical binding sites for Tcf/LEF

factors, HMG-family transcription factors that mediate Wnt

signaling (Cadigan and Nusse, 1997; Moon and Kimelman,

1998). Four additional Tcf sites are present outside of the two

functional 5V domains according to PATSER, but deletions

removing these sites did not alter the dorsal-specific expression

pattern. Therefore, testing sub-domains of 5V sequence for the

ability to drive dorsal-specific expression in transgenic

Xenopus embryos excluded putative transcription factor

binding sites that were dispensable for regulating geminin

expression in prospective neurectoderm.

Vent sites are necessary to suppress ventral geminin expression

at mid-gastrulation

A canonical XVent1 site (GACAAATTCC) was identified at

�3699 (Friedle et al., 1998), and a canonical XVent2 site

(CATTAGTAAT) was identified at �2264 (Martynova et al.,

2004; Trindade et al., 1999). To determine if these sites are

necessary to regulate geminin expression, they were mutated

individually and as a double mutant construct in the context of

[�3783 to �2920] + [�2425 to �1580] (Fig. 5A). The non-

mutated construct contains domains that are necessary and

sufficient to regulate the dorsal-specific expression pattern of

geminin (Figs. 3B, C and H, I, Figs. 4B, C and Supplemental

Fig. 1). The three mutated constructs and the non-mutated

construct were used to create transgenic embryos, which were

analyzed for the pattern of reporter expression by in situ

hybridization at early (st. 10.5) and mid-gastrulation (st. 11.5).

At stage 10.5, the pattern obtained for all three mutated

constructs was similar to the expression pattern seen for

endogenous Xgem and for transgenic embryos carrying the

non-mutated construct (Figs. 5B, C versus D, E, and data not

shown). In contrast, at stage 11.5, each of the three mutated

constructs drove ubiquitous expression, which differed from

the dorsal-specific expression pattern of endogenous Xgem and

from the pattern of expression obtained for transgenic embryos

carrying the non-mutated construct (Figs. 5H, I versus J, K, and

data not shown). Therefore, signaling through these Vent sites

is not required to establish neural-specific reporter expression

at early gastrulation, but both sites are necessary to repress non-

neural geminin expression at mid- to late-gastrula stages.

Tcf binding sites are necessary for dorsal-specific expression of

geminin at early gastrulation

Three canonical Tcf binding sites ((T/A)CAAAG) (Haremaki

et al., 2003; Hilton et al., 2003; Letamendia et al., 2001;McGrew

et al., 1999) were identified at �3506, �3460 and �2214. Each

site was mutated individually and as a triple mutant construct in

the context of [�3783 to�2920] + [�2425 to�1580] (Fig. 5A).

These site-mutated constructs were used to generate transgenic

gastrulae, which were analyzed by in situ hybridization.

Constructs containing mutations of individual Tcf sites drove

dorsal-specific reporter expression in the same or a slightly lower

percent of embryos than the non-mutant construct (data not

shown). However, mutation of all three Tcf sites resulted in weak

reporter expression, without any spatial localization in early- (st.

10.5) and mid-gastrulae (st 11.5), compared to embryos

transgenic for the non-mutated construct (Figs. 5F, G versus B,

C and L, M versus H, I). These data indicate that signaling

through these Tcf sites is additive and required for the initial

dorsal-specific expression of geminin at early gastrulation.

5V regulatory sequence and Xgem respond positively to Wnt

signaling and negatively to BMP signaling

To determine if the identified Vent and Tcf sites are likely to

mediate BMP and Wnt signaling, respectively, we tested

responsiveness of the geminin 5V sequences and of endogenous

Xgem to these signaling pathways. Embryos transgenic for

construct [�4694 to +436] were injected at the 2-cell stage with

mRNA for bmp4, beta-catenin, or noggin. At late gastrula (st.

12), reporter expression was suppressed in embryos injected

with bmp4 compared to uninjected transgenic embryos (Fig. 6A

versus B). Conversely, injection of beta-catenin or noggin

increased reporter expression compared to uninjected transgenic

embryos (Fig. 6A versus C, D). Luciferase reporter activity

driven by [�4694 to +436] was increased by coinjection with

mRNA for either beta-catenin or noggin; luciferase activity

driven by the pGL3basic vector was unaffected by beta-catenin

or noggin coinjection (Fig. 6E). Consistent with the findings

above, endogenous Xgem expression was decreased by inject-

ing a beta-catenin morpholino oligonucleotide (MO) (Heasman

et al., 2000) (Fig. 6F versus G). Injecting mRNA for Xvent1 or

Xvent2 at the 2-cell or 32-cell stage also decreased Xgem

Fig. 5. Mutation of putative Tcf- and Vent-binding sites in the geminin 5V regulatory region alters reporter expression pattern in transgenic gastrulae. Using the

construct schematized in A, consensus binding sites for Vent and Tcf at the positions shown were mutated (see Materials and methods). Sites are numbered relative to

the transcription start site. (B–M) Transgenic embryos carrying the non-mutated or mutated constructs were analyzed at early (st. 10.5; B–G) and mid-gastrulation

(st. 11.5; H–M) by in situ hybridization for eGFP expression. (B, C, H, I) Construct with non-mutated sites. (D, E, J, K) Construct with mutated XVent1 and XVent2

sites. (F, G, L, M) Construct with all three Tcf sites mutated. (B, D, F, H, J, L) Side views with posterior to right and dorsal up. (C, E, G, I, K, M) Posterior views with

dorsal up. (C, E, G, I, K, M) Numbers represent the fraction of total embryos carrying the indicated construct that displayed the expression pattern shown.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 501

expression in injected cells (data not shown and Fig. 6F versus

H, I). Conversely, ventral injection of mRNA for dominant

negative versions of Xvent1 or Xvent2 (VPXvent-1 or VPXvent-

2) (Onichtchouk et al., 1998) at the 8-cell or 32-cell stage

resulted in expanded or ectopic Xgem expression (data not

shown and Fig. 6F versus J, K). Together, these data

demonstrate that both endogenous Xgem and the geminin 5Vregulatory sequences can respond positively to Wnt signaling

and negatively to BMP signaling. These data are consistent with

our identification of putative binding sites for downstream

effectors of these pathways that are necessary for dorsal-specific

expression of geminin at gastrula stages (Fig. 5).

Fig. 6. geminin regulatory sequences respond to the Wnt and BMP signaling pathways in the gastrula embryo. (A–D) In situ hybridization for eGFP in embryos

transgenic for [�4694 to +436] plasmid. Animal hemisphere views of late gastrula embryos (st. 12), dorsal up. Embryos were injected in one cell at the two cell stage

(left bilateral half in the images shown) with bmp4 (B), beta-catenin (C), or noggin (D) mRNA. (A) shows an uninjected sibling embryo. (E) Luciferase assays of

embryos injected with pGL3basic or this vector containing 5.1 kb of geminin 5V regulatory sequences. Samples were co-injected with mRNAs as indicated.

Normalized values (Firefly to Renilla) for three pools of embryos per sample were averaged. Representative results are shown. Error bars represent standard errors.

(F–K) In situ hybridization for Xgem at st. 11.5. (F–J) Dorsal views with posterior down. (K) Side view with posterior to the right and dorsal up. (F) uninjected

embryo (G) Injected with beta-catenin morpholino oligonucleotide into 1 cell at the 2-cell stage. (H–K) Injected with mRNA for Xvent1 (H), Xvent2 (I), VPXvent-1

(J), or VPXvent-2 (K) into one A-tier blastomere at the 32-cell stage. (G–K) Blue X-gal staining indicates site of injection. (B–D and G) mRNA delivery was not

restricted to a small ectodermal domain, and axial patterning of the injected embryos was broadly affected.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506502

Evolutionary conservation of geminin regulatory sequences

We compared 5.1 kb of human geminin 5V sequence to 5Vsequences for several other geminin orthologs using WU-

BLAST to generate pair-wise alignments. Domain �2289 to

�2020 contained the most highly conserved sequence (data not

shown). In contrast, domain �3783 to �3336 was not more

highly conserved than other 5V sequences that were unnecessaryto regulate the dorsal-specific expression pattern (data not

shown). Narrowing the 5V sequence by transgenic analysis priorto computational analyses was therefore critical to identify the

domains necessary for regulating dorsal-specific expression.

Human geminin 5V sequence was analyzed in this study

because it was a vertebrate ortholog whose sequence had been

determined when the project was initiated. We found that the 5Vsequence from hgem recapitulated the expression pattern of

Xgem in transgenic Xenopus embryos (Fig. 1 and Supplemen-

tal Fig. 1). This suggests that the mechanisms regulating

geminin expression in these two species may be similar.

Alignment of Xgem with hgem 5V sequences is shown in Fig.

7A. The two regulatory domains of hgem isolated in our

transgenic analysis (dark grey) are within the sequences that

aligned with Xgem (medium grey). Putative Tcf and Vent sites

are also contained in the aligned Xgem sequence, suggesting

that regulation of geminin via Tcf and Vent sites may also be

conserved. Similar pair-wise alignments are shown for human

versus Pan troglodytes, Mus musculus, or Xenopus tropicalis

geminin orthologs in Supplemental Fig. 3. In all cases, the

regulatory domains from hgem are within the aligned

sequence, and contain putative Tcf and Vent binding sites.

Furthermore, 5 kb of 5V sequence from Xgem was sufficient to

drive dorsal-specific expression at mid-gastrulation in trans-

genic Xenopus (Figs. 7B, C). The functional domains from

hgem and the putative Tcf and Vent sites within them appear to

be conserved among other geminin orthologs.

Discussion

These results provide new information regarding transcrip-

tional regulation of neural-specific gene expression in the

vertebrate embryo during gastrulation. Analysis of forty 5Vsequence-reporter constructs in transgenic embryos identified

Fig. 8. Model of regulation of early geminin expression. (A) Early gastrula:.

Nuclear-localized Beta-catenin is active dorsally. Our data demonstrates that

Beta-catenin can activate Xgem and the geminin 5V sequences and that Tcf sites

in the geminin regulatory sequences are necessary for geminin expression in

dorsal ectoderm at early gastrulation. (B) Mid gastrula: geminin can activate its

own transcription. BMP4 activates the transcriptional repressors XVent1 and

XVent2. Our data demonstrates that XVent1 and XVent2 sites in the geminin

regulatory sequences are necessary to restrict geminin expression to dorsal

tissues at mid-gastrulation. We also found that Xgem or geminin regulatory

sequences responded negatively to BMP signaling. Schematics are side views

of st. 10.5 (A) and st. 11.5 (B) Xenopus gastrulae with posterior facing right and

dorsal up.

Fig. 7. Conservation of Regulatory Domains. (A) The 5.1 kb of hgem 5V sequence is aligned to 6300 bp of Xgem 5V sequence using WU-BLAST (http://

blast.wustl.edu) (parameters: W = 5, M = 5, N = �3, Q = 10, R = 10). Sequences are represented by gray bars. In hgem, the two regulatory domains are represented

by dark gray bars. Aligned sequences between hgem and Xgem are represented by medium-gray bars and linked by green lines marking the ends of the alignments.

Putative binding sites for Tcf, XVent1 and XVent2 are labeled by red, green and blue vertical lines, respectively. These putative sites are determined by scanning both

sequences against scoring matrices of these transcription factors using the program PATSER (http://ural.wustl.edu) and using cutoff values determined by the

program. Locations of lines (above or below the sequences) indicate the orientation of the sites (on forward or reverse strand respectively). Lengths of lines are

proportional to the scores of the sites which indicate the strength of the sites. (B, C) Transgenic embryos carrying 5 kb of upstream sequence from Xgem cloned into

pRAREeGFP were analyzed at st 11.5 by in situ hybridization for eGFP. (B) side view with posterior to the right and dorsal up. (C) posterior view with dorsal up.

Numbers indicate sequence location relative to the transcription start site.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 503

two domains that are necessary and sufficient for directing

neural-specific expression. These sequence elements are

conserved among geminin orthologs and contain canonical

binding sites for Tcf, XVent1, and XVent2. Mutation of these

sites demonstrated that signaling through Tcf sites is required

for dorsal-specific expression at early gastrulation. Signaling

through XVent1 and XVent2 sites is necessary to restrict

geminin expression to prospective neurectoderm at mid-

gastrulation. Furthermore, both Xgem and the geminin 5Vsequences responded positively to Wnt signaling and nega-

tively to BMP signaling, suggesting that the identified,

canonical Tcf and Vent binding sites may mediate regulation

of geminin by the Wnt and BMP pathways, respectively.

Below, we propose a model for the regulation of neural-specific

geminin expression based upon previous work and our results

reported here (Fig. 8).

geminin expression is upregulated in dorsal ectoderm at

early gastrula stages (Kroll et al., 1998). Our results show that

Tcf sites in the geminin 5V regulatory sequences act additively

and are necessary for this early dorsal geminin expression. In

Xenopus blastulae, maternal Beta-catenin is nuclear-localized

dorsally (Larabell et al., 1997; Schneider et al., 1996). This

dorsally-localized Beta-catenin may act in Tcf containing

regulatory complexes to activate geminin expression in dorsal

ectoderm at early gastrulation (Fig. 8A).

Baker et al. (1999) found that Wnt signaling downregulated

bmp4 expression during gastrulation. Wnt signaling may

downregulate bmp4 expression by activating geminin, which

can suppress bmp4 (Kroll et al., 1998). In addition, maternal

Beta-catenin can activate expression of the BMP antagonists

noggin and chordin at blastula stages (Kuroda et al., 2004;

Wessely et al., 2001). This highlights two possible roles for

Wnt signaling in early neural development: direct activation of

early neural markers such as geminin and inhibition of BMP

activity. These mechanisms may act cooperatively or indepen-

dently to facilitate neural gene expression downstream of

dorsally-localized maternal Beta-catenin.

By mid-gastrulation, geminin mRNA is restricted to dorsal

ectoderm (Kroll et al., 1998). geminin expression may be

excluded from ventral ectoderm by BMP signaling. XVent1

and XVent2 are transcriptional repressors downstream of BMP

signaling (Gawantka et al., 1995; Onichtchouk et al., 1996;

Schmidt et al., 1996). We have shown that XVent1 and XVent2

binding sites each suppress ventral reporter expression driven

by geminin 5V regulatory sequences at mid-gastrula stages. We

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506504

propose that BMP signaling activates XVent1 and XVent2,

which repress expression of geminin in ventral ectoderm (Fig.

8B). Meanwhile, Geminin may maintain its own transcription

in the dorsal ectoderm, perhaps partially through its ability to

negatively regulate bmp4 expression (Kroll et al., 1998).

Whereas Tcf sites are required to mediate dorsal activation of

geminin expression at early gastrula, at mid-gastrulation the

XVent1 and XVent2 sites mediate suppression of geminin

expression in ventral ectoderm.

Our mutagenesis results demonstrate that repression of

geminin expression through the XVent1 and XVent2 sites does

not affect the pattern of geminin expression at early gastrulation,

when expression of geminin first prefigures the prospective

neural plate. Signaling through these sites appears to play a later

role in restricting geminin expression to the dorsal ectoderm.

Although our results suggest that BMP signaling regulates

geminin expression through these Vent sites, our results do not

exclude the possibility that BMP signaling could also regulate

geminin expression through alternate mechanisms.

Ets transcription factors are downstream effectors of FGF

signaling (Bertrand et al., 2003; Brent and Tabin, 2004;

Janknecht et al., 1996; Munchberg and Steinbeisser, 1999;

O’Hagan et al., 1996). The 5.1 kb of upstream sequence from

hgem contains six potential Ets binding sites according to

PATSER, but deletion of domains containing these sites did not

affect the dorsal-specific expression pattern. Thus, FGF

signaling through these Ets sites does not appear to directly

regulate geminin transcription at gastrulation. FGF signaling

may regulate geminin expression by other mechanisms such as

by influencing BMP or Wnt signaling. For example, FGF

signaling can antagonize BMP signaling by phosphorylating

the linker region of Smad1 (De Robertis and Kuroda, 2004;

Kuroda et al., 2005; Pera et al., 2003).

In this study, we used transgenic Xenopus embryos to

isolate two 5V sequence domains in a systematic manner

independent of sequence conservation or the presence of

canonical binding sites. An alternate method to identify

regulatory sequences assumes that, among orthologous genes,

regulatory sequences will be more highly conserved than

neighboring sequences. This approach has several potential

difficulties. Relative to coding regions, regulatory sequences

are difficult to align computationally because they frequently

contain few functional binding sites interspersed with

sequences that do not regulate gene expression and are not

conserved across species (Rodriguez-Trelles et al., 2003; Wray

et al., 2003). In addition, sequence motifs matching canonical

transcription factor binding sites appear by chance; the

presence of a particular motif does not indicate whether that

site is functionally relevant. For example, as discussed above,

the PATSER program identified four Tcf sites, which when

deleted were not necessary for dorsal-specific expression of

geminin at gastrula stages. Conversely, sequences that do not

obviously match canonical transcription factor binding motifs

may impact gene expression by facilitating transcription factor

or cofactor binding at neighboring sites. We assayed sequences

for the ability to recapitulate the endogenous Xgem expression

pattern. These analyses eliminated sequences conserved by

chance as well as nonfunctional putative binding sites, and

identified regulatory sequences that were relatively less

conserved than surrounding sequences that were unnecessary

to direct dorsal-specific expression at gastrula stages.

Through this work, we defined sequences required for

regulating the onset of expression of a neural marker in early

gastrulae, information not previously known for any vertebrate

gene. We identified two geminin 5V sequence elements, which

are conserved among geminin orthologs. Our data reveal a

requirement for signaling through Tcf binding sites for neural-

specific geminin expression at the early gastrula stage.We found

that signaling through XVent1 and XVent2 sites does not affect

initiation of the dorsal-specific pattern, but is needed to restrict

geminin expression at mid-gastrulation. In agreement with these

data, Wnt signaling upregulated and BMP signaling down-

regulated the expression of geminin 5V sequence and endoge-

nous Xgem. Together, these data demonstrate that signaling

through Tcf and Vent binding sites regulates the transcription of

geminin. Future studies will identify protein complexes that

bind the Tcf and Vent sites identified here. We will also further

examine whether the transcriptional regulatory sequence ele-

ments defined here are necessary to regulate other geminin

orthologs and other early markers of presumptive neurectoderm,

such as Sox3 (Penzel et al., 1997), using the PhyloCon program

(Wang and Stormo, 2003) and transgenic analyses.

Acknowledgments

The authors thank Gary Stormo for assistance with

computational analyses, Janet Heasman for the beta-catenin

morpholino, Christof Niehrs for the VPXvent-1 and VPXvent-

2 plasmids, and Diane Redmond for graphic arts assistance.

This work was funded by grants from the NIH (R01

GM66815-01), the March of Dimes, and the Whitehall

Foundation to K.L.K. J.J.T. is supported by HHMI predoctoral

grant 59003447. T.W. is supported partly by NIH training

grant in genomic science 2T32HG00045 and a Kauffman

fellowship.

Appendix A. Supplementary data

Supplementary data associated with this article can be found

in the online version at doi:10.1016/j.ydbio.2005.10.047.

References

Bachiller, D., Klingensmith, J., Kemp, C., Belo, J.A., Anderson, R.M., May,

S.R., McMahon, J.A., McMahon, A.P., Harland, R.M., Rossant, J., De

Robertis, E.M., 2000. The organizer factors Chordin and Noggin are

required for mouse forebrain development. Nature 403, 658–661.

Baker, J.C., Beddington, R.S., Harland, R.M., 1999. Wnt signaling in Xenopus

embryos inhibits bmp4 expression and activates neural development. Genes

Dev. 13, 3149–3159.

Beddington, R.S., 1994. Induction of a second neural axis by the mouse node.

Development 120, 613–620.

Bertrand, V., Hudson, C., Caillol, D., Popovici, C., Lemaire, P., 2003. Neural

tissue in ascidian embryos is induced by FGF9/16/20, acting via a

combination of maternal GATA and Ets transcription factors. Cell 115,

615–627.

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506 505

Brent, A.E., Tabin, C.J., 2004. FGF acts directly on the somitic tendon

progenitors through the Ets transcription factors Pea3 and Erm to regulate

scleraxis expression. Development 131, 3885–3896.

Cadigan, K.M., Nusse, R., 1997. Wnt signaling: a common theme in animal

development. Genes Dev. 11, 3286–3305.

Connolly, D.J., Patel, K., Cooke, J., 1997. Chick noggin is expressed in the

organizer and neural plate during axial development, but offers no evidence

of involvement in primary axis formation. Int. J. Dev. Biol. 41, 389–396.

Dale, L., Slack, J.M., 1987. Fate map for the 32-cell stage of Xenopus laevis.

Development 99, 527–551.

Darras, S., Nishida, H., 2001. The BMP signaling pathway is required together

with the FGF pathway for notochord induction in the ascidian embryo.

Development 128, 2629–2638.

Davidson, B.P., Kinder, S.J., Steiner, K., Schoenwolf, G.C., Tam, P.P., 1999.

Impact of node ablation on the morphogenesis of the body axis and the

lateral asymmetry of the mouse embryo during early organogenesis. Dev.

Biol. 211, 11–26.

Davis, R.L., Turner, D.L., Evans, L.M., Kirschner, M.W., 2001. Molecular

targets of vertebrate segmentation: two mechanisms control segmental

expression of Xenopus hairy2 during somite formation. Dev. Cell 1,

553–565.

De Robertis, E.M., Kuroda, H., 2004. Dorsal–ventral patterning and neural

induction in Xenopus embryos. Annu. Rev. Cell Dev. Biol. 20, 285–308.

Delaune, E., Lemaire, P., Kodjabachian, L., 2005. Neural induction in Xenopus

requires early FGF signalling in addition to BMP inhibition. Development

132, 299–310.

Fainsod, A., Deissler, K., Yelin, R., Marom, K., Epstein, M., Pillemer, G.,

Steinbeisser, H., Blum, M., 1997. The dorsalizing and neural inducing gene

follistatin is an antagonist of BMP-4. Mech. Dev. 63, 39–50.

Fekany-Lee, K., Gonzalez, E., Miller-Bertoglio, V., Solnica-Krezel, L., 2000.

The homeobox gene bozozok promotes anterior neuroectoderm formation

in zebrafish through negative regulation of BMP2/4 and Wnt pathways.

Development 127, 2333–2345.

Friedle, H., Rastegar, S., Paul, H., Kaufmann, E., Knochel, W., 1998. Xvent-1

mediates BMP-4-induced suppression of the dorsal-lip-specific early

response gene XFD-1V in Xenopus embryos. EMBO J. 17, 2298–2307.

Gawantka, V., Delius, H., Hirschfeld, K., Blumenstock, C., Niehrs, C., 1995.

Antagonizing the Spemann organizer: role of the homeobox gene Xvent-1.

EMBO J. 14, 6268–6279.

Hamilton, F.S., Wheeler, G.N., Hoppler, S., 2001. Difference in XTcf-3

dependency accounts for change in response to beta-catenin-mediated Wnt

signalling in Xenopus blastula. Development 128, 2063–2073.

Haremaki, T., Tanaka, Y., Hongo, I., Yuge, M., Okamoto, H., 2003. Integration

of multiple signal transducing pathways on Fgf response elements of the

Xenopus caudal homologue Xcad3. Development 130, 4907–4917.

Harland, R.M., 1991. In situ hybridization: an improved whole-mount method

for Xenopus embryos. Methods Cell Biol. 36, 685–695.

Hawley, S.H., Wunnenberg-Stapleton, K., Hashimoto, C., Laurent, M.N.,

Watabe, T., Blumberg, B.W., Cho, K.W., 1995. Disruption of BMP signals

in embryonic Xenopus ectoderm leads to direct neural induction. Genes

Dev. 9, 2923–2935.

Heasman, J., Kofron, M., Wylie, C., 2000. Beta-catenin signaling activity

dissected in the early Xenopus embryo: a novel antisense approach. Dev.

Biol. 222, 124–134.

Hemmati-Brivanlou, A., Melton, D.A., 1992. A truncated activin receptor

inhibits mesoderm induction and formation of axial structures in Xenopus

embryos. Nature 359, 609–614.

Hemmati-Brivanlou, A., Kelly, O.G., Melton, D.A., 1994. Follistatin, an

antagonist of activin, is expressed in the Spemann organizer and displays

direct neuralizing activity. Cell 77, 283–295.

Hertz, G.Z., Stormo, G.D., 1999. Identifying DNA and protein patterns with

statistically significant alignments of multiple sequences. Bioinformatics

15, 563–577.

Hilton, E., Rex, M., Old, R., 2003. VegT activation of the early zygotic gene

Xnr5 requires lifting of Tcf-mediated repression in the Xenopus blastula.

Mech. Dev. 120, 1127–1138.

Hongo, I., Kengaku, M., Okamoto, H., 1999. FGF signaling and the anterior

neural induction in Xenopus. Dev. Biol. 216, 561–581.

Hu, Y., Wang, T., Stormo, G.D., Gordon, J.I., 2004. RNA interference of

achaete– scute homolog 1 in mouse prostate neuroendocrine cells reveals its

gene targets and DNA binding sites. Proc. Natl. Acad. Sci. U. S. A. 101,

5559–5564.

Hudson, C., Darras, S., Caillol, D., Yasuo, H., Lemaire, P., 2003. A conserved

role for the MEK signalling pathway in neural tissue specification and

posteriorisation in the invertebrate chordate, the ascidian Ciona intestinalis.

Development 130, 147–159.

Imai, Y., Gates, M.A., Melby, A.E., Kimelman, D., Schier, A.F., Talbot, W.S.,

2001. The homeobox genes vox and vent are redundant repressors of dorsal

fates in zebrafish. Development 128, 2407–2420.

Janknecht, R., Monte, D., Baert, J.L., de Launoit, Y., 1996. The ETS-related

transcription factor ERM is a nuclear target of signaling cascades involving

MAPK and PKA. Oncogene 13, 1745–1754.

Kessel, M., Pera, E., 1998. Unexpected requirements for neural induction in the

avian embryo. Trends Genet. 14, 169–171.

Khokha, M.K., Yeh, J., Grammer, T.C., Harland, R.M., 2005. Depletion of

three BMP antagonists from Spemann’s organizer leads to a catastrophic

loss of dorsal structures. Dev. Cell 8, 401–411.

Klingensmith, J., Ang, S.L., Bachiller, D., Rossant, J., 1999. Neural induction

and patterning in the mouse in the absence of the node and its derivatives.

Dev. Biol. 216, 535–549.

Koshida, S., Shinya, M., Nikaido, M., Ueno, N., Schulte-Merker, S., Kuroiwa,

A., Takeda, H., 2002. Inhibition of BMP activity by the FGF signal

promotes posterior neural development in zebrafish. Dev. Biol. 244, 9–20.

Kroll, K.L., Amaya, E., 1996. Transgenic Xenopus embryos from sperm

nuclear transplantations reveal FGF signaling requirements during gastru-

lation. Development 122, 3173–3183.

Kroll, K.L., Salic, A.N., Evans, L.M., Kirschner, M.W., 1998. Geminin, a

neuralizing molecule that demarcates the future neural plate at the onset of

gastrulation. Development 125, 3247–3258.

Kuroda, H., Wessely, O., De Robertis, E.M., 2004. Neural induction in

Xenopus: requirement for ectodermal and endomesodermal signals via

Chordin, Noggin, beta-Catenin, and Cerberus. PLoS Biol. 2, E92.

Kuroda, H., Fuentealba, L., Ikeda, A., Reversade, B., De Robertis, E.M., 2005.

Default neural induction: neuralization of dissociated Xenopus cells is

mediated by Ras/MAPK activation. Genes Dev. 19, 1022–1027.

Lamb, T.M., Harland, R.M., 1995. Fibroblast growth factor is a direct neural

inducer, which combined with noggin generates anterior–posterior neural

pattern. Development 121, 3627–3636.

Lamb, T.M., Knecht, A.K., Smith, W.C., Stachel, S.E., Economides, A.N.,

Stahl, N., Yancopolous, G.D., Harland, R.M., 1993. Neural induction by the

secreted polypeptide noggin. Science 262, 713–718.

Larabell, C.A., Torres, M., Rowning, B.A., Yost, C., Miller, J.R., Wu, M.,

Kimelman, D., Moon, R.T., 1997. Establishment of the dorso-ventral axis in

Xenopus embryos is presaged by early asymmetries in beta-catenin that are

modulated by the Wnt signaling pathway. J. Cell Biol. 136, 1123–1136.

Launay, C., Fromentoux, V., Shi, D.L., Boucaut, J.C., 1996. A truncated FGF

receptor blocks neural induction by endogenous Xenopus inducers.

Development 122, 869–880.

Lemaire, P., Bertrand, V., Hudson, C., 2002. Early steps in the formation of

neural tissue in ascidian embryos. Dev. Biol. 252, 151–169.

Letamendia, A., Labbe, E., Attisano, L., 2001. Transcriptional regulation by

Smads: crosstalk between the TGF-beta and Wnt pathways. J. Bone Jt.

Surg., Am. 83-A (Suppl. 1), S31–S39.

Leung, T., Bischof, J., Soll, I., Niessing, D., Zhang, D., Ma, J., Jackle, H.,

Driever, W., 2003. bozozok directly represses bmp2b transcription and

mediates the earliest dorsoventral asymmetry of bmp2b expression in

zebrafish. Development 130, 3639–3649.

Levin, M., 1998. The roles of activin and follistatin signaling in chick

gastrulation. Int. J. Dev. Biol. 42, 553–559.

Martynova, N., Eroshkin, F., Ermakova, G., Bayramov, A., Gray, J., Grainger,

R., Zaraisky, A., 2004. Patterning the forebrain: FoxA4a/Pintallavis and

Xvent2 determine the posterior limit of Xanf1 expression in the neural

plate. Development 131, 2329–2338.

Mason, I., 1996. Neural induction: do fibroblast growth factors strike a cord?

Curr. Biol. 6, 672–675.

Matys, V., Fricke, E., Geffers, R., Gossling, E., Haubrock, M., Hehl, R.,

J.J. Taylor et al. / Developmental Biology 289 (2006) 494–506506

Romano, L.A., Karas, D., Kel, A.E., Kel-Margoulis, O.V., Kloos, D.U.,

Land, S., Lewicki-Potapov, B., Michael, H., Munch, R., Reuter, I., Rotert,

S., Saxel, H., Scheer, M., Thiele, S., Wingender, E., 2003. TRANSFAC:

transcriptional regulation, from patterns to profiles. Nucleic Acids Res. 31,

374–378.

Matzuk, M.M., Lu, N., Vogel, H., Sellheyer, K., Roop, D.R., Bradley, A., 1995.

Multiple defects and perinatal death in mice deficient in follistatin. Nature

374, 360–363.

McGarry, T.J., Kirschner, M.W., 1998. Geminin, an inhibitor of DNA

replication, is degraded during mitosis. Cell 93, 1043–1053.

McGrew, L.L., Takemaru, K., Bates, R., Moon, R.T., 1999. Direct regulation of

the Xenopus engrailed-2 promoter by the Wnt signaling pathway, and a

molecular screen for Wnt-responsive genes, confirm a role for Wnt

signaling during neural patterning in Xenopus. Mech. Dev. 87, 21–32.

McMahon, J.A., Takada, S., Zimmerman, L.B., Fan, C.M., Harland, R.M.,

McMahon, A.P., 1998. Noggin-mediated antagonism of BMP signaling is

required for growth and patterning of the neural tube and somite. Genes

Dev. 12, 1438–1452.

Melby, A.E., Clements, W.K., Kimelman, D., 1999. Regulation of dorsal gene

expression in Xenopus by the ventralizing homeodomain gene Vox. Dev.

Biol. 211, 293–305.

Miya, T., Nishida, H., 2003. An Ets transcription factor, HrEts, is target of FGF

signaling and involved in induction of notochord, mesenchyme, and brain

in ascidian embryos. Dev. Biol. 261, 25–38.

Moon, R.T., Kimelman, D., 1998. From cortical rotation to organizer gene

expression: toward a molecular explanation of axis specification in

Xenopus. Bioessays 20, 536–545.

Munchberg, S.R., Steinbeisser, H., 1999. The Xenopus Ets transcription factor

XER81 is a target of the FGF signaling pathway. Mech. Dev. 80, 53–65.

Nieuwkoop, P.D., Faber, J., 1967. Normal Table of Xenopus laevis Daudin

Hubrecht Lab., Utrecht; North-Holland Publishing Co., Amsterdam.

O’Hagan, R.C., Tozer, R.G., Symons, M., McCormick, F., Hassell, J.A., 1996.

The activity of the Ets transcription factor PEA3 is regulated by two distinct

MAPK cascades. Oncogene 13, 1323–1333.

Onichtchouk, D., Gawantka, V., Dosch, R., Delius, H., Hirschfeld, K.,

Blumenstock, C., Niehrs, C., 1996. The Xvent-2 homeobox gene is part

of the BMP-4 signalling pathway controlling dorsoventral patterning of

Xenopus mesoderm. Development 122, 3045–3053.

Onichtchouk, D., Glinka, A., Niehrs, C., 1998. Requirement for Xvent-1 and

Xvent-2 gene function in dorsoventral patterning of Xenopus mesoderm.

Development 125, 1447–1456.

Osada, S.I., Saijoh, Y., Frisch, A., Yeo, C.Y., Adachi, H., Watanabe, M.,

Whitman, M., Hamada, H., Wright, C.V., 2000. Activin/nodal responsive-

ness and asymmetric expression of a Xenopus nodal-related gene converge

on a FAST-regulated module in intron 1. Development 127, 2503–2514.

Penzel, R., Oschwald, R., Chen, Y., Tacke, L., Grunz, H., 1997. Characteriza-

tion and early embryonic expression of a neural specific transcription factor

xSOX3 in Xenopus laevis. Int. J. Dev. Biol. 41, 667–677.

Pera, E.M., Ikeda, A., Eivers, E., De Robertis, E.M., 2003. Integration of IGF,

FGF, and anti-BMP signals via Smad1 phosphorylation in neural induction.

Genes Dev. 17, 3023–3028.

Quinn, L.M., Herr, A., McGarry, T.J., Richardson, H., 2001. The Drosophila

Geminin homolog: roles for Geminin in limiting DNA replication, in

anaphase and in neurogenesis. Genes Dev. 15, 2741–2754.

Rentzsch, F., Bakkers, J., Kramer, C., Hammerschmidt, M., 2004. Fgf signaling

induces posterior neuroectoderm independently of Bmp signaling inhibi-

tion. Dev. Dyn. 231, 750–757.

Reversade, B., Kuroda, H., Lee, H., Mays, A., De Robertis, E.M., 2005.

Depletion of Bmp2, Bmp4, Bmp7 and Spemann organizer signals

induces massive brain formation in Xenopus embryos. Development

132, 3381–3392.

Ribisi Jr., S., Mariani, F.V., Aamar, E., Lamb, T.M., Frank, D., Harland, R.M.,

2000. Ras-mediated FGF signaling is required for the formation of posterior

but not anterior neural tissue in Xenopus laevis. Dev. Biol. 227, 183–196.

Rodriguez-Trelles, F., Tarrio, R., Ayala, F.J., 2003. Evolution of cis-regulatory

regions versus codifying regions. Int. J. Dev. Biol. 47, 665–673.

Roel, G., Hamilton, F.S., Gent, Y., Bain, A.A., Destree, O., Hoppler, S., 2002.

Lef-1 and Tcf-3 transcription factors mediate tissue-specific Wnt signaling

during Xenopus development. Curr. Biol. 12, 1941–1945.

Rupp, R.A., Snider, L., Weintraub, H., 1994. Xenopus embryos regulate the

nuclear localization of XMyoD. Genes Dev. 8, 1311–1323.

Sasai, Y., Lu, B., Steinbeisser, H., De Robertis, E.M., 1995. Regulation of

neural induction by the Chd and Bmp-4 antagonistic patterning signals in

Xenopus. Nature 377, 757.

Schmidt, J.E., von Dassow, G., Kimelman, D., 1996. Regulation of dorsal–

ventral patterning: the ventralizing effects of the novel Xenopus homeobox

gene Vox. Development 122, 1711–1721.

Schneider, S., Steinbeisser, H., Warga, R.M., Hausen, P., 1996. Beta-catenin

translocation into nuclei demarcates the dorsalizing centers in frog and fish

embryos. Mech. Dev. 57, 191–198.

Seo, S., Herr, A., Lim, J.W., Richardson, G.A., Richardson, H., Kroll, K.L.,

2005. Geminin regulates neuronal differentiation by antagonizing Brg1

activity. Genes Dev. 19, 1723–1734.

Shih, J., Fraser, S.E., 1996. Characterizing the zebrafish organizer: microsur-

gical analysis at the early-shield stage. Development 122, 1313–1322.

Smith, J.L., Schoenwolf, G.C., 1989. Notochordal induction of cell wedging in

the chick neural plate and its role in neural tube formation. J. Exp. Zool.

250, 49–62.

Spemann, H., Mangold, H., 1924. Uber Induktion von Embryonanlagen durch

Implantation artfremder Organisatoren. Roux’ Arch. Entwicklungsmech.

Org. 100, 538–599.

Stern, C.D., 2005. Neural induction: old problem, new findings, yet more

questions. Development 132, 2007–2021.

Storey, K.G., Crossley, J.M., De Robertis, E.M., Norris, W.E., Stern, C.D.,

1992. Neural induction and regionalisation in the chick embryo. Develop-

ment 114, 729–741.

Streit, A., Lee, K.J., Woo, I., Roberts, C., Jessell, T.M., Stern, C.D., 1998.

Chordin regulates primitive streak development and the stability of induced

neural cells, but is not sufficient for neural induction in the chick embryo.

Development 125, 507–519.

Streit, A., Berliner, A.J., Papanayotou, C., Sirulnik, A., Stern, C.D., 2000.

Initiation of neural induction by FGF signalling before gastrulation. Nature

406, 74–78.

Trindade, M., Tada, M., Smith, J.C., 1999. DNA-binding specificity and

embryological function of Xom (Xvent-2. Dev. Biol. 216, 442–456.

Waddington, C., 1930. Developmental mechanics of chick and duck embryos.

Nature 125, 924–925.

Wang, T., Stormo, G.D., 2003. Combining phylogenetic data with co-regulated

genes to identify regulatory motifs. Bioinformatics 19, 2369–2380.

Wawersik, S., Evola, C., Whitman, M., 2005. Conditional BMP inhibition in

Xenopus reveals stage-specific roles for BMPs in neural and neural crest

induction. Dev. Biol. 277, 425–442.

Wessely, O., Agius, E., Oelgeschlager, M., Pera, E.M., De Robertis, E.M.,

2001. Neural induction in the absence of mesoderm: beta-catenin-

dependent expression of secreted BMP antagonists at the blastula stage in

Xenopus. Dev. Biol. 234, 161–173.

Wilson, P.A., Hemmati-Brivanlou, A., 1995. Induction of epidermis and

inhibition of neural fate by Bmp-4. Nature 376, 331–333.

Wilson, S.I., Graziano, E., Harland, R., Jessell, T.M., Edlund, T., 2000. An

early requirement for FGF signalling in the acquisition of neural cell fate in

the chick embryo. Curr. Biol. 10, 421–429.

Wilson, S.I., Rydstrom, A., Trimborn, T., Willert, K., Nusse, R.,

Jessell, T.M., Edlund, T., 2001. The status of Wnt signalling

regulates neural and epidermal fates in the chick embryo. Nature

411, 325–330.

Wray, G.A., Hahn, M.W., Abouheif, E., Balhoff, J.P., Pizer, M., Rockman,

M.V., Romano, L.A., 2003. The evolution of transcriptional regulation in

eukaryotes. Mol. Biol. Evol. 20, 1377–1419.

Xu, R.H., Kim, J., Taira, M., Zhan, S., Sredni, D., Kung, H.F., 1995. A

dominant negative bone morphogenetic protein 4 receptor causes

neuralization in Xenopus ectoderm. Biochem. Biophys. Res. Commun.

212, 212–219.