Developmental Biology 344 (2010) 158–171

Contents lists available at ScienceDirect

Developmental Biology

j ourna l homepage: www.e lsev ie r.com/deve lopmenta lb io logy

Conserved expression of mouse Six1 in the pre-placodal region (PPR) andidentification of an enhancer for the rostral PPR

Shigeru Sato a,⁎, Keiko Ikeda a, Go Shioi b, Haruki Ochi c, Hajime Ogino c, Hiroshi Yajima a, Kiyoshi Kawakami a

a Division of Biology, Center for Molecular Medicine, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke, Tochigi 329-0498, Japanb Laboratory for Animal Resources and Genetic Engineering, RIKEN Center for Developmental Biology (CDB), 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japanc Graduate School of Biological Sciences, Nara Institute of Science and Technology (NAIST), 8916-5 Takayama, Ikoma, Nara 630-0192, Japan

⁎ Corresponding author. Fax: +81 285 44 5476.E-mail address: ssato@jichi.ac.jp (S. Sato).

0012-1606/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.ydbio.2010.04.029

a b s t r a c t

a r t i c l e i n f o

Article history:Received for publication 2 April 2010Revised 24 April 2010Accepted 26 April 2010Available online 21 May 2010

Keywords:Sensory placodeEnhancerComparative genomicsTransgenic mouseHomeodomainVertebrate evolution

All cranial sensory organs and sensory neurons of vertebrates develop from cranial placodes. In chick,amphibians and zebrafish, all placodes originate from a common precursor domain, the pre-placodal region(PPR), marked by the expression of Six1/4 and Eya1/2. However, the PPR has never been described inmammals and the mechanism involved in the formation of PPR is poorly defined. Here, we report theexpression of Six1 in the horseshoe-shaped mouse ectoderm surrounding the anterior neural plate in apattern broadly similar to that of non-mammalian vertebrates. To elucidate the identity of Six1-positivemouse ectoderm, we searched for enhancers responsible for Six1 expression by in vivo enhancer assays. Oneconserved non-coding sequence, Six1-14, showed specific enhancer activity in the rostral PPR of chick andXenopus and in the mouse ectoderm. These results strongly suggest the presence of PPR in mouse and that itis conserved in vertebrates. Moreover, we show the importance of the homeodomain protein-binding sites ofSix1-14, the Six1 rostral PPR enhancer, for enhancer activity, and that Dlx5, Msx1 and Pax7 are candidatebinding factors that regulate the level and area of Six1 expression, and thereby the location of the PPR. Ourfindings provide critical information and tools to elucidate the molecular mechanism of early sensorydevelopment and have implications for the development of sensory precursor/stem cells.

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

Introduction

In vertebrates, all cranial sensory organs (except the retina) andsensory neurons originate from the cranial placodes (Baker andBronner-Fraser, 2001; Brugmann and Moody, 2005; Schlosser, 2006;Streit, 2004). Adenohypophyseal, olfactory, lens, trigeminal, otic andepibranchial placodes have been described in amniotes, which giverise to adenohypophysis, olfactory epithelium, lens, trigeminalganglion, inner ear and vestibulo-acoustic ganglion, and epibranchialsensory ganglia (geniculate, nodose and petrosal ganglia), respec-tively. All placodes share similar characteristics; they form columnarepithelia adjacent to the neural tube, some of the cells delaminate toform sensory ganglia (Graham et al., 2007), and they are neurogenicwith the exception of lens and adenohypophyseal placodes. More-over, there is growing evidence supporting the notion that allplacodes originate from a common precursor domain termed thepre-placodal region (PPR) or pan-placodal primordium (Baker andBronner-Fraser, 2001; McLarren et al., 2003; Schlosser and Ahrens,2004; Streit, 2002). The horseshoe-shaped domain encircling theanterior neural plate was initially reported to be competent in

forming multiple placodes in amphibians (Jacobson, 1963). Fatemapping in chick and zebrafish revealed overlapping distribution ofprecursor cells of different placodes in this region (Bhattacharyyaet al., 2004; Kozlowski et al., 1997; Streit, 2002, Xu et al., 2008).Identification of Six (Six1 and Six4) and Eya (Eya1 and Eya2) genes(Esteve and Bovolenta, 1999; Ishihara et al., 2008a; Litsiou et al., 2005;Sahly et al., 1999; Schlosser and Ahrens, 2004; Streit, 2002) as specificmarkers whose expression match the PPR defined by fate mappingprovided further evidence that PPR is a territory with a distinctmolecular signature. In addition, there is evidence that cells in the PPRshare the property of expressing Pax6 followed by additional lens-marker genes and form the lens when cultured in isolation (Baileyet al., 2006).

Analyses of mutants and morphants in the mouse and zebrafishhave confirmed the role of Six1 and Eya1 in sense organ development.Severe defects are found in multiple sensory organs or placodederivatives; the anterior pituitary (adenohypophysis), olfactoryepithelium, trigeminal ganglion, inner ear and epibranchial ganglia(Bricaud and Collazo, 2006; Chen et al., 2009; Ikeda et al., 2007;Konishi et al., 2006; Kozlowski et al., 2005; Laclef et al., 2003; Li et al.,2003; Nica et al., 2006; Ozaki et al., 2004; Xu et al., 1999; Zheng et al.,2003; Zou et al., 2004). In human, mutations of SIX1 and EYA1 cause asensory disorder called branchio-oto-renal syndrome (Ruf et al.,2004). Importantly, these genes have specific functions in the PPR

159S. Sato et al. / Developmental Biology 344 (2010) 158–171

prior to the specification of individual placodes and/or determinecharacteristics shared among placodes. Overexpression of Six1 andEya1 expanded the PPR at the expense of neural crest and epidermis(Brugmann et al., 2004; Christophorou et al., 2009). Both genessupport placodal neuronal progenitor proliferation and subsequentneuronal differentiation through their effects on SoxB1 expression(Schlosser et al., 2008). Finally, the naïve ectoderm only becomescompetent to respond to otic placode-inducing signals when it hasfirst adopted a pre-placodal identity characterized by the expressionof Eya2 (Martin and Groves, 2006). Thus, there is a general agreementthat the PPR seems to represent a pool of sensory precursor cells and aregulatory network consisting of various genes including Six1/4 andEya1/2 confers pre-placodal characteristics to the region (Baker andBronner-Fraser, 2001; Christophorou et al., 2009; Ohyama et al., 2007;Schlosser, 2006). Recent studies have identified several signalingmolecules involved in the induction of the PPR (Ahrens and Schlosser,2005; Brugmann et al., 2004; Esterberg and Fritz, 2009; Glavic et al.,2004; Hong and Saint-Jeannet, 2007; Litsiou et al., 2005; Sjodal et al.,2007; Solomon and Fritz, 2002; Woda et al., 2003). However, threeimportant questions remain unanswered: 1) Does the PPR exist inmammals? 2) Is there any specific enhancer that can be used to label apart of or the entire PPR? and 3) If such an enhancer really exists, whatis the regulatory mechanism activating such a PPR-specific enhancer?

Identification of major enhancers that control the expression of akey developmental gene in a given developmental processes is criticalfor elucidating its underlying molecular basis, as shown in the case ofSox2 regulation during neural induction (Papanayotou et al., 2008;Takemoto et al., 2006; Uchikawa et al., 2003). Elucidation of theregulatory mechanisms of Six1/4 and Eya1/2 expression couldprovide a clue to answer the above questions. Unfortunately, noneof the major conserved enhancers of Eya1 activates transcription inthe PPR (Ishihara et al., 2008b). Also, the lack of information as towhether Six4 and Eya2 play a conserved critical role during earlysensory development points to the importance of analyzing Six1regulation. In the mouse, Six1 expression is detected in all placodes(excluding the lens placode) and their derivatives (Gu et al., 2004;Laclef et al., 2003; Oliver et al., 1995; Ozaki et al., 2004), and appears atembryonic day 7.5 (E7.5) in the endoderm (Gu et al., 2004) and at E8.0in the rostral region of the embryo (Chen et al., 2009). However, itsexpression in the horseshoe-shaped PPR or even the presence of suchregion in mouse embryo remains elusive.

Here, we report mouse Six1 expression in the ectoderm surround-ing the anterior neural plate in a pattern essentially similar to that ofnon-mammalian vertebrates. To elucidate the identity of Six1-positiveectoderm, we found that one conserved sequence had specific en-hancer activity in the PPR of chick and Xenopus and in the ectoderm ofmouse. Together, the results suggest the presence of a PPR in mouse.We also analyzed the regulatorymechanismactivating the unique Six1PPR enhancer and identified how Six1-positive domain/PPR isestablished.

Materials and methods

Genomic sequence analysis

The genomic sequences covering Six1 were obtained fromEnsembl. Global pairwise alignment was carried out using shuffle-LAGAN (Brudno et al., 2003), and the results were visualized using theVISTA Browser (Frazer et al., 2004). Conserved transcription factorbinding sites were identified using rVISTA (Loots and Ovcharenko,2004), Mulan (Ovcharenko et al., 2005) or TESS (Schug, 2008).

Reporter plasmid and transgene construction

The conserved non-coding sequences (CNSs) were isolated by PCRfrom genomic DNAs or by digesting genomic subclones and ligated

into ptkEGFP (Uchikawa et al., 2003). mSix1-14 (mouse Six1-14,565 bp) and cSix1-14PCR (chick Six1-14, 784 bp) were isolated usingthe primers listed in Table S1. The ptkmRFP1ver2 (Inoue et al., 2007)was used for construction of mRFP1 and multimerized reporters.Mutated ptkmRFP1-mSix1-14 reporters were constructed using theprimers listed in Table S2. For mouse transgenesis, wild-type andmutated Six1-14 were each ligated into ASShsp68lacZpA (Sasaki andHogan, 1996) or ASStkintronlacZpA and transgene DNA fragmentswere excised and purified using QIAEX II (Qiagen, Hilden, Germany).For Xenopus transgenesis, wild-type mSix1-14 was ligated intoISpBSIISK+betaGFP (Ogino et al., 2006). All plasmids were verifiedby DNA sequencing and purified by QIAfilter or EndoFree Plasmid Kit(Qiagen). Table S3 shows a list of plasmids.

Animals

Mice were housed in an environmentally-controlled room in theCDB, RIKEN Kobe and in the Center for Experimental Medicine of JichiMedical University, under the guidelines for animal experiments.Fertilized eggs of chick were purchased from Shiroyama Poultry Farm(Kanagawa, Japan), and incubated at 38 °C in a humidified rockingincubator. The developmental stage of chick embryos was determinedaccording to Hamburger and Hamilton (1951). Xenopus were kept inthe animal facility at NAIST. All experimental protocols were approvedby the Ethics Review Committee for Animal Experimentation of JichiMedical University.

Electroporation into chick embryos and detection of enhancer activity

Electroporation into chick embryos (Fig. 3A) was performed asdescribed previously (Ishihara et al., 2008b). In the initial screening ofenhancers, electroporation was verified using pCAG-HcRed (Matsudaand Cepko, 2004) that drives ubiquitous expression of HcRed underthe control of the strong CAG promoter/enhancer (Figs. 3B,C). FormSix1-14 mutation analysis, the ratio of the amount of plasmid formRFP1 reporters to EGFP control was kept constant (mRFP:EGFP=1:2) to adjust fluorescence intensity, and mRFP1 and EGFPimages were taken at the same exposure time. Embryos wereexamined at 6, 12 and 24 h post-electroporation (h.p.e.). Weperformed electroporation until we obtained more than 5 embryoswith homogeneous DNA distribution and normal morphology. Thepatterns of enhancer activities were highly reproducible, andessentially the same results were obtained from those embryos. Forhistological analysis, embryos were fixed and cryosections (14 μm)were prepared.

Generation and analysis of transgenic Xenopus embryos

Transgenic Xenopus embryos were generated using the modifiedsperm nuclear transplantation method (Ogino et al., 2008). Expres-sion of EGFP mRNA was detected by in situ hybridization formaximum sensitivity. Stained embryos were fixed, embedded in 2%agarose and thick vibratome sections (100 μm) were prepared.

Generation and histological analysis of transgenic mice

Transgenic mice were generated by microinjection using fertilizedeggs of CD-1 (ICR) using a standard protocol (Nagy et al., 2003). Forgenotyping E10.5 embryos, yolk sac DNA was isolated and subjectedto PCR with primers mSix1-14-1/mhsp68R or mSix1-14-1/ptkEGFP-RP (Table S4). Embryos at E8.0 were first fixed and processed for X-galstaining. Whole-embryo DNAs from lacZ-negative embryos weregenotyped using the aforementioned primers and primers specific tolacZ (genotyping) and Six1 (to monitor DNA quality) (Table S4).Mouse embryos were fixed and processed for X-gal staining as

160 S. Sato et al. / Developmental Biology 344 (2010) 158–171

described previously (Kimura et al., 1997). Stained embryos wererefixed and cryosections (16 μm) were prepared.

In situ hybridization

Whole mount in situ hybridization was performed as describedpreviously (Ikeda et al., 2007; Ishihara et al., 2008a) using RNA probesspecific to mouse Six1 (Oliver et al., 1995), chick Dlx5 (Ferrari et al.,1995) and chick Dlx6 (Brown et al., 2005). Stained embryos wererefixed and cryosections (20 μm) were prepared.

Immunofluorescence

Immunofluorescence staining using anti-mouse Six1 antibody(1:2000 dilution, rabbit IgG) was performed as described previously(Ikeda et al., 2007; Konishi et al., 2006). Anti-human Sox2 (1:200dilution, goat IgG, Santa Cruz Biotechnology, Santa Cruz, CA) was usedto stain the neural plate in chick. Anti-Pax6 (1:500 dilution, ascites)and anti-Pax3 (1:50 dilution, hybridoma supernatant) mouse mono-clonal antibodies from Developmental Hybridoma Bank, University ofIowa, were used to stain the lens and trigeminal placodes, respec-tively. The secondary antibodies were Alexa Fluor 488 anti-rabbit IgG(1:1000 dilution), Alexa Fluor 546 anti-goat IgG (1:1000 dilution) andAlexa Fluor 633 anti-mouse IgG (1:2000 dilution) (Molecular Probes/Invitrogen, Carlsbad, CA). Confocal microscope was used to examinecolocalization of proteins.

Overexpression experiments

For overexpression, full-length chick Dlx5 (Ferrari et al., 1995),Msx1 (Yokouchi et al., 1991) and Pax7 (Matsunaga et al., 2001) cDNAswere amplified using the primers listed in Table S5 and ligated intopCAGIG containing the strong ubiquitous CAG promoter/enhancerand an IRES-EGFP cassette (Matsuda and Cepko, 2004). A fragment ofthe Dlx5 homeodomain (cDlx5HD)was also amplified and ligated intopCAGIG-VP16 to overexpress cDlx5HD fused to heterologous activa-tion domain (Woda et al., 2003). Empty and various pCAGIGconstructs (3.0 mg/ml) were co-electroporated into the anteriorepiblast (Fig. 6F) with ptkmRFP1-mSix1-14×4 (125 ng/ml).

Verification of siRNA against chick Dlx5 using cultured cells

To knockdown chick Dlx5 we chose Stealth siRNA (Invitrogen).According to BLOCK-iT RNAi Designer (Invitrogen), three different25mer double-stranded Stealth siRNAs, NM_204159_stealth_466 (5′-CCGCGGACUAUUUAUUCCAGCUUUC), NM_204159_stealth_614 (5′-GGUCCAAGAUCAAGAAGAUCAUGAA) and NM_204159_stealth_453(5′-GAAAGUGCGCAAACCGCGGACUAUU), were synthesized using areference sequence for the chick Dlx5 mRNA NM_204159. As acontrol, Stealth RNAi Negative Control Medium GC Duplex (Invitro-gen) that does not match with any known vertebrate transcript wasused. HEK293 cells were plated at a density of 1×105 cells/well in a24-well plate the day before and transfected with Stealth siRNAs(20 pmol/well) and the chick Dlx5 expression vector pHM6-cDlx5(40 ng/well) using Lipofectamine 2000 (Invitrogen). Twenty-fourhours after transfection, total RNAs were extracted using ISOGENreagent (Nippon Gene, Tokyo, Japan) and treated with DNase I (RocheDiagnostics, Basel, Switzerland). RT-PCR was performed with 100 ngof total RNAs for chick Dlx5 and human BETA-ACTIN with primerslisted in Table S5 using QIAGENOneStep RT-PCR kit (Qiagen). Aliquots(4 μl out of 25 μl reaction) of each PCR reaction were taken at 16, 20,24, and 28 cycles, run on a 5% polyacrylamide gel and stained withVistra Green (Amersham/GE Healthcare). Quantitation of Dlx5 mRNAlevels relative to BETA-ACTINmRNAwas carried out using the STORMsystem (Amersham/GE Healthcare) and ImageQuant software

(Amersham/GEHealthcare). pHM6-cDlx5was constructed by ligatingthe full-length chick Dlx5 cDNA into pHM6 (Roche Diagnostics).

Knockdown of chick Dlx5 in vivo

For in vivo knockdown, Stealth siRNA against chick Dlx5,NM_204159_stealth_614, and Stealth RNAi Negative Control MediumGCDuplex (45 pmol/μl) were electroporated into the anterior epiblast(Fig. 6F) with reporter ptkmRFP1-mSix1-14×4 (31.25 ng/μl) andcontrol pCAGIG (625 ng/μl, used to verify siRNA delivery). Embryoswerefixed at 6 h.p.e. andmRFP1 and EGFP imageswere taken from thedorsal side at the same exposure time. To quantify knockdown effects,the mean values (gray values) of mRFP1 and EGFP channels of a fixedarea (380×190 μm rectangle, Fig. 7E), which covers a large part ofmRFP1-positive rostral PPR on the left side of the embryo, weremeasured. After subtracting the background fluorescence obtainedfromnon-electroporated embryos, themeanmRFP1 levels normalizedto the EGFP levels were calculated, and shown relative to the valueobtained from the negative control siRNA.

Purification of GST-fusion proteins and gel mobility shift assay

The gel shift assay was performed as described previously (Satoet al., 2002). Full-length chick Dlx5 and Msx1 cDNAs amplified usingthe primers listed in Table S5 were ligated into pGEX6P1 (Amersham/GE Healthcare, Amersham, UK). GST-fusion proteins were expressedin Escherichia coli BL21 and purified on a glutathione-Sepharose 4Bcolumn (Amersham/GE Healthcare). Oligonucleotides encompassingHD protein-binding sites of mSix1-14 (14-HD1 to HD3, Table S6) wereend-labeled with [alpha-32P]dCTP and used as probes. In somebinding reactions, 50-fold molar excess of unlabeled wild-type (14-HD1 to HD3) and mutated (14-HD1m to HD3m, Table S6) oligonu-cleotides were added as competitors.

Results

Mouse Six1 expression in the non-neural ectoderm surrounding theneural plate

To analyze the expression pattern of mouse Six1 in early somitestages, we performed whole mount in situ hybridization usingembryos at E8.0. As shown in Figs. 1A,B, Six1 mRNA was expressedin embryos with 0–1 somite and a very shallow foregut pocket, whichbest corresponds to the beginning of Theiler Stage 12a (TS12a) (Bardet al., 1998). Sectioning of the embryo revealed expression of Six1 inthe non-neural ectoderm flanking the thickened neural plate as wellas in the endoderm and mesoderm (Fig. 1C). The expression of Six1 inthe non-neural ectoderm was still noted in embryos with 5 somites(TS12b, Figs. 1D–F) and later in all placodes except the lens placode(data not shown) as has been reported (Oliver et al., 1995; Ozaki et al.,2004). Six1 expression was also detected in the developing somites atTS12b (Figs. 1E,K).

Since the Six1 hybridization signal in the non-neural ectodermwas weak, particularly at early 0–2 somite stages, and was difficultto precisely map the expression domains, we also examined Six1protein distribution in complete serial sections using a specific anti-Six1 antibody (Ikeda et al., 2007) (Figs. 1G–J). The distribution ofSix1 in the non-neural ectoderm was continuous, extendingcaudally and surrounded a large part of the developing neuralplate (1–2 somites, Figs. 1G–J) as reported in non-mammalianvertebrates (Bessarab et al., 2004; Litsiou et al., 2005; Schlosser andAhrens, 2004). Sagittal sections at TS12b (5 somites, Figs. 1K,L)showed confinement of Six1 expression to the non-neural ectodermunderlying the rostral neural plate. However, whether localizationof Six1 was exclusive to the non-neural (=surface) ectoderm orextended to the outer (non-neural) slope of the anterior neural

Fig. 1. Expression patterns of Six1 mRNA and protein in the PPR of mouse embryo. (A–C) Whole mount in situ hybridization for Six1 at 0–1 somite stage (TS12a). Anterior (A)and lateral (B) views. Arrowheads indicate signals in the ectoderm surrounding the anterior neural plate. The approximate position of transverse section (C) is indicated (redarrowhead in B). Six1 expression in the ectoderm is weak but specific to non-neural domain (black dotted line, C). Expression is also detected in the mesoderm and endoderm.(D–F) Six1 in situ hybridization at 5 somite stage (TS12b). Anterior (D) and lateral (E) views. Six1 is expressed under the ANR (white arrowheads), and extends posteriorlyalong the lateral edge of the neural plate (black and red arrowheads). The approximate position of the transverse section (F) is indicated (red arrowhead in E). Thehybridization signal in the ectoderm (black dotted line, F) becomes clear at this stage. Also note Six1 expression in somites (asterisks, E). (G–I) Immunohistochemical detectionof Six1 at 1–2 somite stage (TS12a). Merged confocal images. Six1 (green) is detected by anti-Six1 antibody in non-neural ectoderm (orange dotted line). (J) A schematicrepresentation of TS12a embryo showing the positions of the transverse sections (G–I) and a stripe of the Six1-positive ectoderm (orange dotted line). (K,L) Six1 distribution at5 somite stage (TS12b). Sagittal section (K) and high power view (L) showing Six1 in the ectoderm and the outer slope of the ANR (orange dotted line), endoderm, headmesoderm and somites. DAPI is used for nuclear staining (blue in G–I,K,L). In B,J–L, anterior is to the left. In E, anterior is to the right. ec: non-neural ectoderm positive for Six1mRNA/Six1 protein, en: endoderm, me: head mesoderm, np: neural plate, so: somites. Scale bars: 100 μm.

161S. Sato et al. / Developmental Biology 344 (2010) 158–171

ridge (ANR) (Cobos et al., 2001) was unclear. Simultaneousdetection of Six1 protein and Fgf8 mRNA, a specific marker of ANR(Shimamura and Rubenstein, 1997), confirmed overlapping of Six1protein and Fgf8 mRNA in the outer slope of the ANR (data notshown). Thus, Six1 expression was detected in the non-neuralectoderm surrounding the anterior neural plate, which reflects thePPR territory in chick (Litsiou et al., 2005), Xenopus (Schlosser andAhrens, 2004) and zebrafish (Bessarab et al., 2004). Furthermore,Six1 was also expressed in the underlying mesoderm and endodermin mouse as in chick embryos (Litsiou et al., 2005).

Identification of CNSs at Six1 locus

Our analysis revealed the expression of mouse Six1 in a horseshoe-shaped, non-neural ectoderm that resembles the PPR identified inother species. However, since no fate mapping or experimentalembryological studies have been carried out in mouse in relation tothe Six1-positive ectodermal domain at E8.0–E8.5, another piece ofevidence was needed to establish homology between the mouseectoderm and the previously described PPR of non-mammalianmodels. We reasoned that if we could identify a conserved enhancer

Fig. 2. Position of 16 evolutionarily conserved sequences around Six1 exons. (A,B) The VISTA plot of the 190 kb (A) or the central 26 kb (B) intervals containing mouse Six1. The plot shows conserved sequences between mouse and human(top), mouse and opossum (upper middle), mouse and chick (lower middle), and mouse and Xenopus (bottom). The abscissa represents the mouse sequence (base sequence) and the ordinate represents the percentage identity in a 100 bpwindow. Conserved regions above the level of 50%/100 bp are highlighted under the curve, with pink indicating a conserved non-coding sequence (CNS), blue, a conserved exon, and cyan, an untranslated region. Among such CNSs scatteredover the 150-kb region containing Six1 between the flanking Six6 and Six4, 16 are conserved in mammals (mouse, human and opossum) and either chick or Xenopus. They are termed Six1-8 to Six1-29 and the positions are indicated in Arabicnumerals on each plot. Six1-13 of Xenopus does not appear in this plot but its presence is confirmed by direct BLAST search. The genomic sequences were downloaded from Ensembl: human (NCBI 36 assembly, November 2005), mouse (NCBIm37 assembly, April 2007), opossum (MonDom5, Oct 2006), chick (release 2.1, May 2006) and Xenopus (assembly version 4.1, August 2005).

162S.Sato

etal./

Developm

entalBiology

344(2010)

158–171

163S. Sato et al. / Developmental Biology 344 (2010) 158–171

(s) at the mouse Six1 locus that directs gene expression in the Six1-positive ectoderm in mouse as well as PPR in chick and Xenopus,activity of such conserved enhancer in mouse would provide a basisfor the identity of the Six1-positive ectodermal region.

To identify such PPR-specific enhancer, we searched for CNSs atSix1 locus. We focused on CNSs located in the 150-kb region betweenthe neighboring Six6 and Six4 genes (Fig. 2A). Sequence blocks thatshowed higher than 50% identity over 100 bp in mammals and eitherchick or Xenopus were defined as CNSs. There were 16 such CNSs andthe majority clustered in the 20-kb region flanking Six1 exons(Fig. 2B).

Identification of a conserved enhancer that directs PPR-specific geneexpression in chick and Xenopus

To examine potential enhancer activity of 16 CNSs, thesequences from the C57BL/6J mouse genome were isolated andinserted into ptkEGFP (Uchikawa et al., 2003). EGFP reporterplasmids were introduced into the entire epiblast of chick embryos(Stage HH4-4+) by electroporation (Figs. 3A,B), cultured for 6–24 hand the enhancer activity was assessed as EGFP fluorescence.Furthermore, we confirmed the enhancer activity using multi-merized (2× and 4×) reporters.

Among the 16 CNSs, a single CNS (mSix1-14) activated EGFPexpression in the region surrounding the anterior neural plate at 6h.p.e. (Fig. 3B). Transverse sections showed that the expressiondomain was specific to the ectodermal layer (Figs. 3C,D) andimmunohistochemical detection of Sox2 revealed that EGFP signalwas mainly confined to the non-neural ectoderm flanking thethickened Sox2-positive neural plate (Fig. 3E). Interestingly, theexpression domain corresponded to the rostral part of the PPR whoseposterior limit lies beyond the level of Hensen's node (Streit, 2004).The homologous sequence isolated from chick (cSix1-14PCR) alsoactivated EGFP expression in a similar rostral domain in the non-neural ectoderm (Figs. 3J, S1). Among the remaining 15 CNSs, 7 CNSsshowed EGFP expression in specific subdomains within the Six1expression domains (as will be described elsewhere) while another 7CNSs did not show any EGFP expression and 1 CNS exhibited enhanceractivity seemingly-unrelated to Six1 (data not shown).

Then, we examined whether Six1-14 can activate gene expressionin the PPR in Xenopus (Figs. 3F–H). mSix1-14 was ligated intoISpBSIISK+betaGFP (Ogino et al., 2006), and transgenic embryoswere generated and EGFP expression was examined by whole mountin situ hybridization. As shown in Fig. 3G, mSix1-14 activated EGFPexpression in cells surrounding the rostral region of the open neuralplate in transgenic Xenopus embryos at Stage 15 (15 EGFP-positiveembryos/86 normally developed injected embryos). Sagittal sectionsshowed the hybridization signal in the rostral non-neural ectodermand weakly in the adjacent restricted rostral end of the neural plate(Fig. 3H), which corresponds to the PPR defined previously by theexpression of Six1 and Eya1 (Schlosser and Ahrens, 2004).

Finally, to provide further evidence that Six1-14 is a conservedenhancer that drives gene expression in the PPR, we examined thedistribution of cells labeled with Six1-14 in older chick embryos(Figs. 3I–P). To assess the position of Six1-14-labeled cells relativeto various placodes, antibodies against Pax6 (lens placode marker)and Pax3 (trigeminal placode marker) were used. Also, ptkEGFP-mSix1-21×2, which labels otic and epibranchial placodes (S. Sato,unpublished), was electroporated with ptkmRFP1-cSix1-14PCRx4.At 6 h.p.e., the mRFP1-positive cells were located in the rostral PPR(Fig. 3J). At 24 h.p.e., positive cells were located in the most rostralolfactory (Figs. 3K–M) and ventral adenohypophyseal (Fig. 3K)placode areas, ventral head ectoderm (Figs. 3K,M,N) and Pax6-labeled lens placode (Figs. 3K,L,N). However, positive cells wererarely observed in the more caudal Pax3-labeled trigeminalplacode (Figs. 3L,O) and otic/epibranchial placodes labeled by

EGFP driven by mSix1-21 (Figs. 3L,P). These findings stronglysuggest that Six1-14 is an enhancer specific to the rostral PPRconserved in tetrapods.

Enhancer activities of CNSs in transgenic mouse embryos

To examine whether Six1-14 is active and can account for theendogenous Six1 expression in mouse, we constructed transgeneswith lacZ reporter and generated transgenic mouse embryos (Fig. 4).At E8.0, the mSix1-14 fused to the mouse hsp68 promoter activatedlacZ expression in the rostral region of the non-neural ectoderm (9lacZ-positive embryos/20 transgenic embryos). More precisely, at thepresomite stage (TS11), lacZ expression was negative in the ectoderm(0/7, Fig. 4A). At the early somite stage (TS12a, 1–4 somites), lacZ wasexpressed in the rostral non-neural ectoderm, but the number ofembryos (4/8) and positive cells (about 5–50 cells/embryo) was verysmall (Figs. 4B,C). At the subsequent stage (TS12b, 5–7 somites), allembryos (5/5) were lacZ-positive and many lacZ-positive cells wereidentified also (Figs. 4D,E). The positive region was confined to be therostral part of the non-neural ectoderm underlying the anterior neuralplate (Fig. 4E) and was smaller than the endogenous Six1 mRNA/protein expression domain (Fig. 1). The lacZ-positive region in thenon-neural ectoderm was highly restricted even when a multi-merized mSix1-14 reporter was used (Fig. 5Ca). At E10.5, noreproducible lacZ expression was observed (0/11, Fig. 4F), suggestingthat the enhancer activity of mSix1-14 is transient and specific to therostral region of the non-neural ectoderm and does not stronglyactivate gene expression at later stages. The difference from theresults of short-term cell lineage trace in chick, in which mRFP1driven by cSix1-14PCR was detectable in placodes at 24 h.p.e.(Figs. 3I–P, S1) may reflect a difference in the activity of Six1-14between mouse and chick (Fig. S1), the presence of excess amount ofmRFP1 introduced by electroporation in chick and the markedincrease in the size of mouse embryos from E8.0 to E10.5. The aboveresults strongly suggest that the rostral PPR-specific enhancer (Six1-14) identified using chick and Xenopus accounts for part ofendogenous Six1 expression in the horseshoe-shaped domain inmouse. In addition, the results are consistent with the notion thatSix1-positive mouse ectoderm represents the conserved PPR thatexpresses an orthologous gene (Six1) driven by the same conservedenhancer (Six1-14).

Identification of essential cis-elements for mSix1-14

What are the regulatory mechanisms involved in the induction ofSix1 expression and the PPR itself? To provide an answer, wecharacterized the cis-elements and trans-acting factors required forthe activity of Six1-14.

As shown in Fig. 5A, alignment of Six1-14 from 4 tetrapod specieshighlighted conserved transcription factor binding sites: five Gata,three Sox and three homeodomain (HD) protein-binding sites. Toaddress the role of these sites, we introduced wild-type (EGFP) andmutated (mRFP1) reporters into chick embryos (Fig. 5Ba). Sox123mmutation (disrupted all Sox-binding sites) did not alter mRFP1expression driven bymSix1-14 in the PPR (data not shown). Similarly,mutations of a unique E-box and a Pitx-binding site had no discernibleeffect (data not shown). As shown in Fig. 5Bb, Gata1-5m mutation(disrupted all Gata-binding sites) reduced mRFP1 expression. How-ever, the most noticeable change was caused by HD123m mutation(disrupted all HD protein-binding sites): it completely abolishedmRFP1 expression (Fig. 5Bc). Single (HD1m, HD2m and HD3m) anddouble (HD12m) mutations were also analyzed, but they had only amarginal effect (data not shown). Then, to confirm the effect ofHD123mmutation on Six1-14 in mouse, we chose multimerized (4×)reporters. The wild-type Six1-14 frommouse and chick activated lacZexpression in the rostral PPR in all transgenic embryos at TS12

164 S. Sato et al. / Developmental Biology 344 (2010) 158–171

(mSix1-14: 5/5 transgenic embryos, cSix1-14: 3/3, Figs. 5Ca,Cb). LacZexpression was also detected in the anterior neural plate whenmultimerized enhancer was used. Given the strictly non-neuralnature of monomeric mSix1-14, we conceived this activity to beectopic although the exact reason is unclear (Fig. 4). As expected,HD123m mutation resulted in a complete loss of lacZ expression inthe rostral ectoderm (0/5, Fig. 5Cc). On the other hand, LacZexpression in the endoderm was not affected by HD123m mutation(Fig. 5Cc).

Identification of Dlx5 and Msx1 as important transcription factors

The above results indicated that the presence of at least one of thethree HD protein-binding sites was essential for the enhancer activityof Six1-14. Among the HD proteins implicated in the formation ofneural plate border or PPR, Dlx5 and Msx1 have been the best-characterized proteins (Esterberg and Fritz, 2009; McLarren et al.,2003; Phillips et al., 2006; Schlosser, 2008; Streit, 2007). Detailedanalyses in chick showed: 1) pronounced Dlx5 expression in the

Fig. 4. Six1-14 activated gene expression in the non-neural ectoderm inmouse embryos. (A–F) Enhancer activity of mSix1-14-hsp68-lacZpA inmouse.Whole mount embryo at TS11(A) showing lacZ expression in the endoderm and lack of such expression in the ectoderm.Whole mount embryo at TS12a (B) and sagittal section (C) showing weak lacZ expressionin a small number of cells in the rostral region of the non-neural (arrowheads) ectoderm. Whole mount embryo at TS12b (D) and a sagittal section (E) showing strong lacZexpression in the rostral end of the non-neural ectoderm (arrowheads). The approximate positions of sections (C,E) are indicated (red arrowheads in B,D). Endodermal expression oflacZ is detectable through TS11 (A) to TS12b (E). No specific lacZ expression is observed in transgenic embryo at E10.5 (F). A,B,D are frontal views. In C,E,F, anterior is to the left. en:endoderm, np: neural plate. Scale bars: 100 μm (C,E), 200 μm (A,B,D) and 1 mm (F).

165S. Sato et al. / Developmental Biology 344 (2010) 158–171

neural plate border at the anterior region while Msx1 expression wasin the posterior region, 2) extensive expression of Dlx5 in the anteriornon-neural ectoderm while Msx1 is expressed in the neural fold andsubsequently in the neural crest (McLarren et al., 2003; Pera et al.,1999; Streit, 2002; Streit and Stern, 1999). In addition, taking intoconsideration the effects of both proteins on the expression of PPRmarker genes (Six4 and Eya1) (Esterberg and Fritz, 2009; Kaji andArtinger, 2004; McLarren et al., 2003; Phillips et al., 2006; Schlosserand Ahrens, 2004; Solomon and Fritz, 2002), we focused on Dlx5 andMsx1 as potential candidate regulators of Six1-14.

As shown in Fig. 5D, both bacterially-expressed GST-Dlx5 andGST-Msx1 bound to three oligonucleotide probes (14-HD1 to HD3),each harboring one of the three HD protein-binding sites of Six1-14in vitro. The binding was specific since the addition of the wild-typecompetitor, but not the mutated competitors (14-HD1m to HD3m),inhibited complex formation. Then, are Dlx5 and Msx1 involved inthe regulation of Six1-14 in vivo? To investigate this, we first

Fig. 3. The conserved non-coding sequence Six1-14 activates gene expression in the PPR of chickby electroporation (A).Wholemount embryo at 6 h.p.e. with ptkEGFP-mSix1-14×4 showing EGview). Green channel (EGFP) is superimposed on bright-field image. Inset in B shows expressubiquitous expression of HcRed, in a broad area of the embryo. Transverse section showing un(arrowheads) in the non-neural ectoderm (Cb). Transverse section of another embryo at 6 h.p.(D). Green channel (EGFP) is superimposed onDIC image. Themajority of EGFP-positive cells ar(E). (F–H)Enhancer activity of Six1-14 inXenopus. Enhancer activitywas identifiedby transgeneEGFPmRNA distribution (arrowheads) is detected in the region surrounding the anterior neuraexpression in the anterior non-neural ectoderm or the PPR (arrowheads) (H). (I–P) DistributptkEGFP-mSix1-21×2 activate gene expression in the PPR and otic/epibranchial placodes, respeand 24 h.p.e. (K,L) in whole mount. Ventral (J,K) and dorsal (L) views. Merged images of cSix1superimposed on bright-field image. At 6 h.p.e., mRFP1-positive cells are found in the rostral Padenohypophyseal (K) placode areas and ventral head ectoderm (K) but are largely absent froembryo at 24 h.p.e. shown inK,L (the approximate positions of sections are indicated in L).Mergarea (M), lens placode stained by anti-Pax6 antibody (green, pseudocolor, N) and ventral hetrigeminal placode stained by anti-Pax3 antibody (green, pseudocolor, O) and otic/epibranchialPax6 antibody in the lens placode (white arrowheads in N). In contrast, no mRFP-positive cellpositive cell is labeled bymSix1-21 in the otic/epibranchial placodes (awhite arrowhead in P).magnification and the mRFP1 and DAPI channels were taken at the same exposure time. In B,Hplacode, ep: epibranchial placodes, hn: Hensen's node, le: lens placode, np: neural plate, ol: ol200 μm (H), 400 μm (B,G,J,K).

overexpressed Dlx5 in chick by electroporation (Figs. 6A–C,F). ThecDlx5 (Fig. 6B) and cDlx5HD fused to VP16 (Fig. 6C) augmented theco-electroporated reporter (mRFP1) expression under the control ofmSix1-14 in the PPR. Importantly, cDlx5 expanded mRFP1 expres-sion domain into the neural plate (Fig. 6B). The cDlx5HD fused toEngrailed inhibited reporter expression (data not shown). To obtainfurther evidence that Dlx5 activates mSix1-14, we chose aknockdown approach using the Stealth siRNA. Among three syn-thesized siRNAs against chick Dlx5 mRNA, the most effective siRNA(NM_204159_stealth_614) was selected for in vivo usage based onthe result in cultured cells (Fig. 7A). The Dlx5 siRNA electroporatedinto the anterior epiblast of chick embryo (as depicted in Fig. 6F)specifically reduced Dlx5 expression level at 6 h.p.e. (Figs. 7Ba, Bb)but did not alter the expression level of the closely related Dlx6(Figs. 7Bc, Bd). As shown in Fig. 7C, mRFP1 expression driven bymSix1-14×4 was clearly reduced by the Dlx5 siRNA (Fig. 7Cb)compared to the control (Fig. 7Ca). Quantitative analysis (Fig. 7D)

andXenopus. (A–E) Enhancer activity of Six1-14 in chick. Enhancer activitywas identifiedFP expression (arrowheads) in the region surrounding the anterior neural plate (B, dorsalion of HcRed (magenta, pseudocolor) from co-electroporated pCAG-HcRed, which drivesiform expression of HcRed (magenta) in the ectoderm (Ca) and specific EGFP expressione. with ptkEGFP-mSix1-14×2 showing EGFP expression (arrowheads) in a similar patterne located outside the neural plate labeledwith anti-Sox2 antibody (magenta, pseudocolor)sis (F). Frontal viewof a stage15embryo transgenic for ISpBSIISK+betaGFPmSix1-14 (G).l plate by in situ hybridization. Sagittal section (100-μm) of another embryo showing EGFPion of cells expressing mRFP1 driven by Six1-14 in chick. ptkmRFP1-cSix1-14PCRx4 andctively (I). Both reporterswere co-electroporated into the entire epiblast, examined at 6 (J)-14PCR-labeled cells (red, pseudocolor) and mSix1-21-labeled cells (green, pseudocolor)PR, while at 24 h.p.e, mRFP1-positive cells are present in the olfactory (K,L), lens (K,L) andm the otic/epibranchial placodes (L) labeled with EGFP. (M–P) Transverse sections of theed confocal images confirmed the presenceofmRFP1-positive cells in the olfactoryplacodead ectoderm (yellow arrows in N). In contrast, positive cells were mostly absent in theplacodes labeled bymSix1-21 (P). There aremultiplemRFP1-positive cells stainedby anti-is labeled with anti-Pax3 antibody in the trigeminal placode (O) and only a single mRFP1-DAPI is used for nuclear staining (blue inM–P). All confocal imageswere taken at the same,J–L, anterior is to the left. In C–E,G,H,M–P, dorsal side is to the top. ad: adenohypophysealfactory placode, ot: otic placode, ppr: pre-placodal region. Scale bars: 100 μm (C–E,M–P),

Fig. 5. Essential role of HD protein-binding sites on the activity of Six1 rostral PPR enhancer. (A) Alignment of the core regions of Six1-14. Binding sites for homeodomain (orangeshading), Sox (blue shading), Gata (yellow shading), Pitx (blue letters) and bHLH (E-box, green shading) are indicated. Numbers in brackets indicate the number of conserved sitesof each category. (B) Mutation analysis of mSix1-14 in chick. ptkEGFP-mSix1-14×4 (wild-type reporter, green) was co-electroporated with ptkmRFP1 constructs (magenta,pseudocolor): mSix1-14×4 (top, a), Gata1-5mx4 (middle, b) and HD123mx4 (bottom, c) were examined at 6 h.p.e. Wild-type reporters (EGFP in a–c, mRFP1 in a) mark the rostralPPR while Gata1-5m mutation reduces mRFP1 level (b). HD123m mutation abolishes mRFP1 expression (c). Dorsal views (a–c). Anterior is to the left. (C) Mutation analyses ofmSix1-14 in mouse. Wild-type [mSix1-14×4-tkintronlacZpA (a) and cSix1-14PCRx4-tkintronlacZpA (b)] and a mutated [mSix1-14HD123mx4-tkintronlacZpA (c)] transgenes wereused for transgenesis and lacZ expression was examined at E8.0. In sagittal sections, LacZ expression in the PPR at TS12 is observed in embryos injected with mouse (5 lacZ-positiveembryos/5 transgenic embryos, a) and chick (3/3, b) wild-type transgenes but absent in those carrying HD123m transgene (0/5, c). Asterisk indicates ectopic expression in theneural plate. lacZ expression in the endoderm is not affected (c). (D) Gel shift assay. The consensus binding sequences of Dlx (Feledy et al., 1999) andMsx1 (Catron et al., 1993), wild-type (14-HD1 to HD3) and mutated oligonucleotides (14-HD1m to HD3m) used as probes and competitors are shown (left). GST-cDlx5 (middle) and GST-cMsx1 (right) bind tothree probes containing HD protein-binding sites (HD1 to HD3) from mouse Six1-14. The unlabeled wild-type and mutated oligonucleotides (50-fold molar excess) were added tothe binding reaction in lanes 3, 4, 7, 8, 11, 12, 15, 16, 19, 20, 23, and 24. GST-cDlx5 (lanes 1–12) and GST-cMsx1 (lanes 13–24) show slightly different binding specificities. Thepositions of DNA/protein complexes and free probes are indicated. In B,C, anterior is to the left. Abbreviations as in Figs. 1, 3 and 4. Scale bars: 500 μm (B), 100 μm (C).

166 S. Sato et al. / Developmental Biology 344 (2010) 158–171

confirmed a significant reduction (40%, pb0.05, Fig. 7E). In contrast,overexpression of Msx1 inhibited mRFP1 expression driven bymSix1-14×4 (Fig. 6D). Interestingly, overexpressed Pax7, a criticalhomeodomain protein required for neural crest formation (Basch et

al., 2006), also inhibited reporter expression (Fig. 6E). Pax7 needsboth the homeodomain (TAATT) and paired domain (GTCAC)binding sequences for optimal binding, but the homeodomainbinding sequence alone is sufficient for binding and for transcription

Fig. 6. Regulation of the Six1 rostral PPR enhancer by overexpressed homeodomain proteins. (A–E) Overexpression of Dlx5, Msx1 and Pax7 in chick. The reporter ptkmRFP1-mSix1-14×4 was co-electroporated with pCAGIG constructs, and mRFP1 level (magenta, pseudocolor) relative to EGFP fluorescence (derived from an IRES-EGFP cassette of pCAGIG) wasexamined as depicted in F. EGFP fluorescence also indicates electroporated areas (F). Compared with the control pCAGIG (A), cDlx5 (B) and VP16-cDlx5HD (C) augment mRFP1expression in the PPR. Overexpressed cDlx5 expands mRFP1 expression domain into the neural plate (asterisk in B). In contrast, cMsx1 (D) and Pax7 (E) reduce the expression. (F)Experimental design used to assess the effect of co-electroporated pCAGIG constructs onmSix1-14. A–F are dorsal views, and anterior is to the top. Abbreviations as in Figs. 1, 3 and 4.Scale bars: 400 μm.

167S. Sato et al. / Developmental Biology 344 (2010) 158–171

activation (Kumar et al., 2009; White and Ziman, 2008). Thus, theresults indicated that Dlx5 activates Six1 expression by directlybinding to Six1-14 while Msx1 and Pax7 repress the enhancersuggesting their potential role as repressors.

Discussion

PPR is an ectodermal territory conserved in vertebrates

The PPR is considered a domain containing a pool of sensoryprecursor cells in non-mammalian vertebrates. However, the presenceof such domain in mammals or the existence of enhancers specific tothe PPR and the mechanisms that regulate their activity remainunknown. Here, we demonstrated the presence of a contiguoushorseshoe-shaped domain in the non-neural ectoderm expressingSix1 inmouse (Fig. 1). The Six1-positive region surrounded a large partof the neural plate and resembled the PPR described in chick, Xenopusand zebrafish. In addition, using in vivo enhancer assays, we identifieda conserved cis-element (Six1-14) at tetrapod Six1 loci that directsgene expression in the rostral PPR of chick and Xenopus (Fig. 3), and inthe rostral part of the horseshoe-shaped Six1-positive ectoderm in themouse (Fig. 4). We also found that transcription factors previouslyimplicated in PPR induction targeted Six1-14 and regulated Six1expression (Figs. 5–7, discussed below). Collectively, our findingsstrongly suggest that the PPR is an evolutionarily conservedsubdivision of early embryonic ectoderm together with the neuralplate, neural crest and epidermis in vertebrates.

Our study detected Six1 in the outer slope of the ANR as well as thecontiguous region of non-neural ectoderm (Fig. 1). This observation isconsistentwith the view that themost rostral region of the PPR, whichgives rise to adenohypophyseal placode medially and the olfactoryplacode laterally (Bhattacharyya et al., 2004; Cobos et al., 2001; Kouki

et al., 2001), abuts the rostral end of the neural plate excluding neuralcrest from the region (Baker and Bronner-Fraser, 2001; Couly and LeDouarin, 1985; Osumi-Yamashita et al., 1994; Schlosser and Ahrens,2004). Furthermore, the localization pattern of Six1 protein supportsprevious notions and strongly suggests that Six1 acts as a keytranscription factor in the PPR.

cis-Regulatory element controlling Six1 expression in the PPR

Six1-14 activates transcription in the PPR. However, the activity islimited to the rostral part of the PPR in chick, Xenopus and mouse. Theregion seems to correspond to a subregion of the PPR containingpresumptive placodal cells for anterior placodes (McCabe andBronner-Fraser, 2009), and cells labeled by cSix1-14 were indeedfound in those placodes (Figs. 3I–N). The results were unexpectedconsidering the fairly homogeneous expression of Six1 in theectoderm, and thus suggest that Six1-14 accounts for part of Six1expression in the PPR and that probably other cis-elements arerequired for Six1 expression in the caudal PPR and/or the entire PPR.There are at least two potential mechanisms for regulation of Six1 inthe posterior PPR. 1) Six1-14 is a conserved critical enhancer, andother CNS or coding sequences cooperate with Six1-14 to expand theenhancer activity towards the caudal PPR. 2) Six1-14 is a rostral PPR-specific enhancer, and parallel elements account for Six1 expression inthe caudal PPR. It is possible that different combinations orconcentrations of signaling molecules and transcription factors directSix1 expression separately in the rostral and the remaining parts of thePPR as described for Sox2 regulation in the central nervous system byfive separate enhancers (Uchikawa et al., 2003). Concerning othersequences that could act cooperatively or in parallel with Six1-14,there are three possibilities. a) ACNS located outside the 150-kb regionmay act as an enhancer as reported for several other genes (Kleinjan

Fig. 7. Knockdown of Dlx5 using Stealth siRNA in chick embryos. (A) Effect of Stealth RNAi on Dlx5 mRNA expression from the co-transfected pHM6-cDlx5 in HEK293 cells.Expression levels of chick Dlx5mRNA (at cycle 20) are normalized to human BETA-ACTINmRNA expression levels and shown relative to the value obtained from the negative controlStealth siRNA transfection (column 1). Data are mean values from two independent experiments. Column 1: Stealth RNAi Negative Control Medium GC Duplex (1.0), column 2:NM_204159_stealth_466 (0.29), column 3: NM_204159_stealth_614 (0.14), column 4: NM_204159_stealth_453 (0.28). Among three Stealth siRNAs against chick Dlx5,NM_204159_stealth_614 showed the strongest knockdown effect. (B) Specific knockdown of Dlx5 mRNA by Stealth siRNA in chick embryo. Embryos were electroporated with thenegative control siRNA (Stealth RNAi Negative Control Medium GC Duplex) (a,c) or the Dlx5 siRNA (NM_204159_stealth_614) (b,d). Expression pattern of Dlx5 (a,b) and Dlx6 (c,d)at 6 h.p.e examined by whole mount in situ hybridization. EGFP fluorescence from co-electroporated pCAGIG indicates approximate position of the electroporated areas (a′–d′). Dlx5expression level is specifically reduced in the electroporated area (asterisk, b). (C) Knockdown of Dlx5 by Stealth siRNA in chick. The siRNA against Dlx5 and the negative controlsiRNA were electroporated with the reporter ptkmRFP1-mSix1-14×4 and control pCAGIG. At 6 h.p.e., compared with negative control siRNA (Ca), the Dlx5 siRNA(NM_204159_stealth_614) (Cb) reduces the expression of reporter ptkmRFP1-mSix1-14×4 (magenta, pseudocolor) in the area marked with pCAGIG. (D,E) Quantification of theeffect of Stealth siRNA on mSix1-14 enhancer activity. The mean values of mRFP1 and EGFP channels of a fixed rectangular area (D) were measured. The mean mRFP1 levelsnormalized to the EGFP levels were shown relative to the value obtained from the negative control siRNA (E). The relative mRFP1 intensity of embryos that received cDlx5 siRNAiwas significantly lower (Student's t-test, pb0.05) than that of embryos that received negative control siRNA. All images are dorsal views, and anterior is to the top. Abbreviations as inFigs. 1, 3 and 4. Scale bars: 400 μm (Bd′,Cb,D), 800 μm (Bd).

168 S. Sato et al. / Developmental Biology 344 (2010) 158–171

and van Heyningen, 2005). b) A multipartite enhancer composed of acombination of conserved sequences may activate Six1 in the PPR. Theexistence of modular enhancers and contribution of coding sequencesto transcriptional regulation have been described (Lampe et al., 2008;Pappu et al., 2005; Tumpel et al., 2008). c) Functionally conserved(composed of similar sets of transcription factor binding sites) butdiverged (in terms of the exact numbers, order, orientation andspacing of those sites) sequencesmay also play a role in the conservedPPR-specific expression of Six1 (Meireles-Filho and Stark, 2009;Weirauch and Hughes, 2010).

Interplay of homeodomain proteins control Six1 expression in the PPR

What is the mechanism(s) that controls the enhancer activity ofSix1-14? Our results showed that Dlx5 activated Six1-14 while Msx1and Pax7 repressed it. Dlx proteins are implicated not only in thespecification of the neural plate border but also in the development/

induction of PPR and Six1 expression. First, Dlx genes are expressed inthe neural plate border and surrounding non-neural ectoderm andtheir expression continues in the PPR and the contiguous posteriornon-neural ectoderm, although which members of the family areexpressed differs among model vertebrates (Akimenko et al., 1994;McLarren et al., 2003; Pera et al., 1999; Quint et al., 2000; Solomon andFritz, 2002; Yang et al., 1998). In mice, Dlx5, Dlx6 and Dlx3 areexpressed in the border/PPR region (Quint et al., 2000; Yang et al.,1998). Second, in Xenopusmisexpression of a dominant negative Dlx3represses Six1 expression in the caudal PPR and an ectopic Six1induction in the ventral ectoderm by grafted neural plate requiresDlx3 in the ectoderm (Woda et al., 2003). Third, in chick embryos,Dlx5 misexpression in the non-neural ectoderm and neural tube leadsto upregulation of Six4 (McLarren et al., 2003). Six4 is the neighboringgene of Six1 (Fig. 2) and its expression pattern resembles that of Six1(Kawakami et al., 2000; Ozaki et al., 2001) possibly due to enhancersharing with Six1. In zebrafish, knockdown of dlx3b/4b abolishes

169S. Sato et al. / Developmental Biology 344 (2010) 158–171

six4.1 expression as well as eya1 in the PPR, and misexpression ofdlx3b induces their expression (Esterberg and Fritz, 2009). Thus,these previous studies are consistent with the model in which Dlxproteins bind to and activate the Six1 PPR enhancer. Our in vivoknockdown study identified Dlx5 as the major Dlx protein involved inthe regulation of Six1-14 in chick embryo.

Msx1 is also classified as a neural plate border specifier (Meule-mans and Bronner-Fraser, 2004; Streit, 2007). However, in contrast toDlx genes, its expression is pronounced in the posterior region of theborder and maintained in the neural fold, and it plays specific roles inneural crest formation instead of PPR (Monsoro-Burq et al., 2005;Streit, 2002; Streit, 2007; Streit and Stern, 1999). Moreover, amutually antagonistic relationship between Dlx and Msx proteinswas described (Givens et al., 2005; Zhang et al., 1997) that involvesheterodimerization and competitive occupancy of the same bindingsites. Importantly, in zebrafish, loss of function of dlx3b/4b leads tosevere reduction of olfactory, trigeminal and otic placodes (Kaji andArtinger, 2004; Solomon and Fritz, 2002) while simultaneous loss ofboth dlx3b/4b and msx genes results in the rescue of the placodalphenotype and six4.1 expression (Phillips et al., 2006). However, apartfrom the Bmp antagonist cv2 (Esterberg and Fritz, 2009), keydownstream target genes or enhancers controlled by Dlx and Msxremain unknown. Our results suggest that the Six1 rostral PPRenhancer is one of the direct targets of Dlx and Msx. Other as yetunidentified HD proteins could also contribute to the induction of Six1in the PPR. It is also possible that HD proteins required for neural crestformation contribute to the exclusion of Six1 expression from theneural crest territory. Indeed, overexpression of Pax7 repressedmSix1-14 enhancer most likely through Dlx/Msx binding sites(Fig. 6E). In addition, we noticed that the consensus binding sequence(CTAATTAG) of Gbx2, which has recently been shown to be involvedin the induction of neural crest (Berger et al., 2008; Li et al., 2009), issimilar to Dlx and Msx (Fig. 5D).

Six1 expression in the PPR is regulated positively by Fgf andnegatively by Bmp and Wnt (Ahrens and Schlosser, 2005; Brugmannet al., 2004; Glavic et al., 2004; Hong and Saint-Jeannet, 2007; Litsiouet al., 2005; Sjodal et al., 2007). However, there was no apparenttarget sequence responsive to these signals in the Six1 rostral PPRenhancer (Fig. 5). In contrast, the expression of Dlx and Msx genes isregulated by Bmp, Wnt and Fgf at the neural plate border (Litsiouet al., 2005; Monsoro-Burq et al., 2005; Sauka-Spengler and Bronner-Fraser, 2008; Streit, 2007; Taneyhill and Bronner-Fraser, 2005), andthe transactivation capacity of Dlx proteins can be modulated by Bmpthrough p38-mediated phosphorylation and increased interactionwith p300 (Ulsamer et al., 2008). Thus, it may be possible that thesesignaling inputs on Dlx and Msx genes and proteins contribute tocontrol Six1 expression in the PPR. Finally, we found that Gata-bindingsites in the Six1 PPR enhancer are also important for its activity(Figs. 5A,B). Gata2 and Gata3 are expressed in the early non-neuralectoderm (Manaia et al., 2000; Neave et al., 1995; Sheng and Stern,1999) and Gata protein can cooperate with HD proteins (Friedle andKnochel, 2002; Gordon et al., 1997; Schlosser, 2006). A recentidentification of the binding sequence for the HD of Six familyproteins, G(A/G)TATCA (Affolter et al., 2008; Noyes et al., 2008)containing a GATA motif, raises the interesting possibility that Six1and/or Six4 may also regulate the Six1 rostral PPR enhancer.

Conclusions

Wedemonstrated a conserved expression of Six1 in the horseshoe-shaped ectoderm of mouse and the presence of the conserved non-coding sequence that acts as an enhancer specific to the mouseectoderm and to the rostral PPR in chick and Xenopus. The resultssuggest that the Six1-positive mouse ectoderm is the PPR. Activity ofthe Six1 rostral PPR enhancer is positively and negatively regulated byhomeodomain proteins. These findings provide critical information

and tools to understand the molecular mechanism of early sensorydevelopment.

Acknowledgments

We thank Drs. Kazuki Nakao, Yasunori Hayashibara, and ShinichiAizawa for mouse transgenesis, Drs. Masanori Uchikawa and HisatoKondoh for plasmids, technical advice and discussions. We alsothank Drs. Robert Kosher, Andrew Groves, Atsushi Kuroiwa,Harukazu Nakamura, Hiroshi Sasaki, Connie Cepko and Gen Yamadafor materials, Rika Nakayama, Tomoe Bunno at CDB and all membersof the laboratory for the technical assistance. This work wassupported by grants from the MEXT of Japan (to K.K. and S.S.),Support Program for Scientific Research Platform in PrivateUniversities (SPSRP) (to K.K.) and the Naito Foundation (to S.S.).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.ydbio.2010.04.029.

References

Affolter, M., Slattery, M., Mann, R.S., 2008. A lexicon for homeodomain-DNArecognition. Cell 133, 1133–1135.

Ahrens, K., Schlosser, G., 2005. Tissues and signals involved in the induction of placodalSix1 expression in Xenopus laevis. Dev. Biol. 288, 40–59.

Akimenko, M.A., Ekker, M., Wegner, J., Lin, W., Westerfield, M., 1994. Combinatorialexpression of three zebrafish genes related to distal-less: part of a homeobox genecode for the head. J. Neurosci. 14, 3475–3486.

Bailey, A.P., Bhattacharyya, S., Bronner-Fraser, M., Streit, A., 2006. Lens specification isthe ground state of all sensory placodes, from which FGF promotes olfactoryidentity. Dev. Cell 11, 505–517.

Baker, C.V., Bronner-Fraser, M., 2001. Vertebrate cranial placodes I. Embryonicinduction. Dev. Biol. 232, 1–61.

Bard, J.L., Kaufman, M.H., Dubreuil, C., Brune, R.M., Burger, A., Baldock, R.A., Davidson, D.R., 1998. An internet-accessible database of mouse developmental anatomy basedon a systematic nomenclature. Mech. Dev. 74, 111–120.

Basch, M., Bronner-Fraser, M., García-Castro, M., 2006. Specification of the neural crestoccurs during gastrulation and requires Pax7. Nature 441, 218–222.

Berger, M., Badis, G., Gehrke, A., Talukder, S., Philippakis, A., Peña-Castillo, L., Alleyne, T.,Mnaimneh, S., Botvinnik, O., Chan, E., Khalid, F., Zhang,W., Newburger, D., Jaeger, S.,Morris, Q., Bulyk, M., Hughes, T., 2008. Variation in homeodomain DNA bindingrevealed by high-resolution analysis of sequence preferences. Cell 133, 1266–1276.

Bessarab, D.A., Chong, S.W., Korzh, V., 2004. Expression of zebrafish six1 during sensoryorgan development and myogenesis. Dev. Dyn. 230, 781–786.

Bhattacharyya, S., Bailey, A.P., Bronner-Fraser, M., Streit, A., 2004. Segregation of lensand olfactory precursors from a common territory: cell sorting and reciprocity ofDlx5 and Pax6 expression. Dev. Biol. 271, 403–414.

Bricaud, O., Collazo, A., 2006. The transcription factor six1 inhibits neuronal andpromotes hair cell fate in the developing zebrafish (Danio rerio) inner ear. J.Neurosci. 26, 10438–10451.

Brown, S.T., Wang, J., Groves, A.K., 2005. Dlx gene expression during chick inner eardevelopment. J. Comp. Neurol. 483, 48–65.

Brudno, M., Do, C.B., Cooper, G.M., Kim, M.F., Davydov, E., Green, E.D., Sidow, A.,Batzoglou, S., 2003. LAGAN and multi-LAGAN: efficient tools for large-scalemultiple alignment of genomic DNA. Genome Res. 13, 721–731.

Brugmann, S.A., Moody, S.A., 2005. Induction and specification of the vertebrateectodermal placodes: precursors of the cranial sensory organs. Biol. Cell 97,303–319.

Brugmann, S.A., Pandur, P.D., Kenyon, K.L., Pignoni, F., Moody, S.A., 2004. Six1 promotesa placodal fate within the lateral neurogenic ectoderm by functioning as both atranscriptional activator and repressor. Development 131, 5871–5881.

Catron, K.M., Iler, N., Abate, C., 1993. Nucleotides flanking a conserved TAAT core dictatethe DNA binding specificity of three murine homeodomain proteins. Mol. Cell Biol.13, 2354–2365.

Chen, B., Kim, E.H., Xu, P.X., 2009. Initiation of olfactory placode development andneurogenesis is blocked in mice lacking both Six1 and Six4. Dev. Biol. 326, 75–85.

Christophorou, N.A., Bailey, A.P., Hanson, S., Streit, A., 2009. Activation of Six1 targetgenes is required for sensory placode formation. Dev. Biol. 336, 327–336.

Cobos, I., Shimamura, K., Rubenstein, J.L., Martinez, S., Puelles, L., 2001. Fate map of theavian anterior forebrain at the four-somite stage, based on the analysis of quail-chick chimeras. Dev. Biol. 239, 46–67.

Couly, G.F., Le Douarin, N.M., 1985. Mapping of the early neural primordium in quail-chick chimeras. I. Developmental relationships between placodes, facial ectoderm,and prosencephalon. Dev. Biol. 110, 422–439.

Esterberg, R., Fritz, A., 2009. dlx3b/4b are required for the formation of the preplacodalregion and otic placode through local modulation of BMP activity. Dev. Biol. 325,189–199.

170 S. Sato et al. / Developmental Biology 344 (2010) 158–171

Esteve, P., Bovolenta, P., 1999. cSix4, a member of the six gene family of transcriptionfactors, is expressed during placode and somite development. Mech. Dev. 85,161–165.

Feledy, J.A., Morasso, M.I., Jang, S.I., Sargent, T.D., 1999. Transcriptional activation by thehomeodomain protein distal-less 3. Nucleic Acids Res. 27, 764–770.

Ferrari, D., Sumoy, L., Gannon, J., Sun, H., Brown, A.M., Upholt, W.B., Kosher, R.A., 1995.The expression pattern of the Distal-less homeobox-containing gene Dlx-5 in thedeveloping chick limb bud suggests its involvement in apical ectodermal ridgeactivity, pattern formation, and cartilage differentiation. Mech. Dev. 52, 257–264.

Frazer, K.A., Pachter, L., Poliakov, A., Rubin, E.M., Dubchak, I., 2004. VISTA: computationaltools for comparative genomics. Nucleic Acids Res. 32, W273–W279.

Friedle, H., Knochel, W., 2002. Cooperative interaction of Xvent-2 and GATA-2 in theactivation of the ventral homeobox gene Xvent-1B. J. Biol. Chem. 277,23872–23881.

Givens, M.L., Rave-Harel, N., Goonewardena, V.D., Kurotani, R., Berdy, S.E., Swan, C.H.,Rubenstein, J.L., Robert, B., Mellon, P.L., 2005. Developmental regulation ofgonadotropin-releasing hormone gene expression by the MSX and DLX home-odomain protein families. J. Biol. Chem. 280, 19156–19165.

Glavic, A., Maris Honore, S., Gloria Feijoo, C., Bastidas, F., Allende, M.L., Mayor, R., 2004.Role of BMP signaling and the homeoprotein Iroquois in the specification of thecranial placodal field. Dev. Biol. 272, 89–103.

Gordon, D.F., Lewis, S.R., Haugen, B.R., James, R.A., McDermott, M.T., Wood, W.M.,Ridgway, E.C., 1997. Pit-1 and GATA-2 interact and functionally cooperate toactivate the thyrotropin beta-subunit promoter. J. Biol. Chem. 272,24339–24347.

Graham, A., Blentic, A., Duque, S., Begbie, J., 2007. Delamination of cells from neurogenicplacodes does not involve an epithelial-to-mesenchymal transition. Development134, 4141–4145.

Gu, G., Wells, J.M., Dombkowski, D., Preffer, F., Aronow, B., Melton, D.A., 2004. Globalexpression analysis of gene regulatory pathways during endocrine pancreaticdevelopment. Development 131, 165–179.

Hamburger, V., Hamilton, H.L., 1951. A series of normal stages in the development of thechick embryo. J. Morphol. 88, 49–92.

Hong, C.S., Saint-Jeannet, J.P., 2007. The activity of Pax3 and Zic1 regulates three distinctcell fates at the neural plate border. Mol. Biol. Cell 18, 2192–2202.

Ikeda, K., Ookawara, S., Sato, S., Ando, Z., Kageyama, R., Kawakami, K., 2007. Six1 isessential for early neurogenesis in the development of olfactory epithelium. Dev.Biol. 311, 53–68.

Inoue, M., Kamachi, Y., Matsunami, H., Imada, K., Uchikawa, M., Kondoh, H., 2007. PAX6and SOX2-dependent regulation of the Sox2 enhancer N-3 involved in embryonicvisual system development. Genes Cells 12, 1049–1061.

Ishihara, T., Ikeda, K., Sato, S., Yajima, H., Kawakami, K., 2008a. Differential expression ofEya1 and Eya2 during chick early embryonic development. Gene Expr. Patterns 8,357–367.

Ishihara, T., Sato, S., Ikeda, K., Yajima, H., Kawakami, K., 2008b. Multiple evolutionarilyconserved enhancers control expression of Eya1. Dev. Dyn. 237, 3142–3156.

Jacobson, A.G., 1963. The determination and positioning of the nose, lens and ear. III.Effects of reversing the antero-posterior axis of epidermis, neural plate and neuralfold. J. Exp. Zool. 154, 293–303.

Kaji, T., Artinger, K.B., 2004. dlx3b and dlx4b function in the development of Rohon-Beard sensory neurons and trigeminal placode in the zebrafish neurula. Dev. Biol.276, 523–540.

Kawakami, K., Sato, S., Ozaki, H., Ikeda, K., 2000. Six family genes—structure andfunction as transcription factors and their roles in development. Bioessays 22,616–626.

Kimura, C., Takeda, N., Suzuki, M., Oshimura, M., Aizawa, S., Matsuo, I., 1997. Cis-actingelements conserved between mouse and pufferfish Otx2 genes govern theexpression in mesencephalic neural crest cells. Development 124, 3929–3941.

Kleinjan, D.A., van Heyningen, V., 2005. Long-range control of gene expression:emerging mechanisms and disruption in disease. Am. J. Hum. Genet. 76, 8–32.

Konishi, Y., Ikeda, K., Iwakura, Y., Kawakami, K., 2006. Six1 and Six4 promote survival ofsensory neurons during early trigeminal gangliogenesis. Brain Res. 1116, 93–102.

Kouki, T., Imai, H., Aoto, K., Eto, K., Shioda, S., Kawamura, K., Kikuyama, S., 2001.Developmental origin of the rat adenohypophysis prior to the formation of Rathke'spouch. Development 128, 959–963.

Kozlowski, D.J., Murakami, T., Ho, R.K., Weinberg, E.S., 1997. Regional cell movementand tissue patterning in the zebrafish embryo revealed by fate mapping with cagedfluorescein. Biochem. Cell Biol. 75, 551–562.

Kozlowski, D.J., Whitfield, T.T., Hukriede, N.A., Lam, W.K., Weinberg, E.S., 2005. Thezebrafish dog-eared mutation disrupts eya1, a gene required for cell survival anddifferentiation in the inner ear and lateral line. Dev. Biol. 277, 27–41.

Kumar, D., Shadrach, J., Wagers, A., Lassar, A., 2009. Id3 is a direct transcriptional targetof Pax7 in quiescent satellite cells. Mol. Biol. Cell 20, 3170–3177.

Laclef, C., Souil, E., Demignon, J., Maire, P., 2003. Thymus, kidney and craniofacialabnormalities in Six 1 deficient mice. Mech. Dev. 120, 669–679.

Lampe, X., Samad, O.A., Guiguen, A., Matis, C., Remacle, S., Picard, J.J., Rijli, F.M.,Rezsohazy, R., 2008. An ultraconserved Hox-Pbx responsive element resides in thecoding sequence of Hoxa2 and is active in rhombomere 4. Nucleic Acids Res. 36,3214–3225.

Li, B., Kuriyama, S., Moreno, M., Mayor, R., 2009. The posteriorizing gene Gbx2 is a directtarget of Wnt signalling and the earliest factor in neural crest induction.Development 136, 3267–3278.

Li, X., Oghi, K.A., Zhang, J., Krones, A., Bush, K.T., Glass, C.K., Nigam, S.K., Aggarwal, A.K.,Maas, R., Rose, D.W., Rosenfeld, M.G., 2003. Eya protein phosphatase activityregulates Six1–Dach–Eya transcriptional effects in mammalian organogenesis.Nature 426, 247–254.

Litsiou, A., Hanson, S., Streit, A., 2005. A balance of FGF, BMP and WNT signallingpositions the future placode territory in the head. Development 132, 4051–4062.

Loots, G.G., Ovcharenko, I., 2004. rVISTA 2.0: evolutionary analysis of transcriptionfactor binding sites. Nucleic Acids Res. 32, W217–W221.

Manaia, A., Lemarchandel, V., Klaine, M., Max-Audit, I., Romeo, P., Dieterlen-Lievre, F.,Godin, I., 2000. Lmo2 and GATA-3 associated expression in intraembryonichemogenic sites. Development 127, 643–653.

Martin, K., Groves, A.K., 2006. Competence of cranial ectoderm to respond to Fgfsignaling suggests a two-step model of otic placode induction. Development 133,877–887.

Matsuda, T., Cepko, C.L., 2004. Electroporation and RNA interference in the rodentretina in vivo and in vitro. Proc. Natl. Acad. Sci. U. S. A. 101, 16–22.

Matsunaga, E., Araki, I., Nakamura, H., 2001. Role of Pax3/7 in the tectumregionalization. Development 128, 4069–4077.

McCabe, K.L., Bronner-Fraser, M., 2009. Molecular and tissue interactions governinginduction of cranial ectodermal placodes. Dev. Biol. 332, 189–195.

McLarren, K.W., Litsiou, A., Streit, A., 2003. DLX5 positions the neural crest andpreplacode region at the border of the neural plate. Dev. Biol. 259, 34–47.

Meireles-Filho, A., Stark, A., 2009. Comparative genomics of gene regulation-conservation and divergence of cis-regulatory information. Curr. Opin. Genet.Dev. 19, 565–570.

Meulemans, D., Bronner-Fraser, M., 2004. Gene-regulatory interactions in neural crestevolution and development. Dev. Cell 7, 291–299.

Monsoro-Burq, A.H., Wang, E., Harland, R., 2005. Msx1 and Pax3 cooperate to mediateFGF8 andWNT signals during Xenopus neural crest induction. Dev. Cell 8, 167–178.

Nagy, A., Gertsentein, M., Vintersten, K., Behringer, R., 2003. Manipulating the MouseEmbryo: a Laboratory ManualThird Edition. Cold Spring Harbor Laboratory Press.

Neave, B., Rodaway, A., Wilson, S.W., Patient, R., Holder, N., 1995. Expression ofzebrafish GATA 3 (gta3) during gastrulation and neurulation suggests a role in thespecification of cell fate. Mech. Dev. 51, 169–182.

Nica, G., Herzog, W., Sonntag, C., Nowak, M., Schwarz, H., Zapata, A.G., Hammerschmidt,M., 2006. Eya1 is required for lineage-specific differentiation, but not for cellsurvival in the zebrafish adenohypophysis. Dev. Biol. 292, 189–204.

Noyes, M., Christensen, R., Wakabayashi, A., Stormo, G., Brodsky, M., Wolfe, S., 2008.Analysis of homeodomain specificities allows the family-wide prediction ofpreferred recognition sites. Cell 133, 1277–1289.

Ogino, H., Fisher, M., Grainger, R.M., 2008. Convergence of a head-field selector Otx2and Notch signaling: a mechanism for lens specification. Development 135,249–258.

Ogino, H., McConnell, W.B., Grainger, R.M., 2006. High-throughput transgenesis inXenopus using I-SceI meganuclease. Nat. Protoc. 1, 1703–1710.

Ohyama, T., Groves, A.K., Martin, K., 2007. The first steps towards hearing: mechanismsof otic placode induction. Int. J. Dev. Biol. 51, 463–472.

Oliver, G., Wehr, R., Jenkins, N.A., Copeland, N.G., Cheyette, B.N., Hartenstein, V.,Zipursky, S.L., Gruss, P., 1995. Homeobox genes and connective tissue patterning.Development 121, 693–705.

Osumi-Yamashita, N., Ninomiya, Y., Doi, H., Eto, K., 1994. The contribution of bothforebrain and midbrain crest cells to the mesenchyme in the frontonasal mass ofmouse embryos. Dev. Biol. 164, 409–419.

Ovcharenko, I., Loots, G.G., Giardine, B.M., Hou, M., Ma, J., Hardison, R.C., Stubbs, L.,Miller, W., 2005. Mulan: multiple-sequence local alignment and visualization forstudying function and evolution. Genome Res. 15, 184–194.

Ozaki, H., Nakamura, K., Funahashi, J., Ikeda, K., Yamada, G., Tokano, H., Okamura, H.O.,Kitamura, K., Muto, S., Kotaki, H., Sudo, K., Horai, R., Iwakura, Y., Kawakami, K.,2004. Six1 controls patterning of the mouse otic vesicle. Development 131,551–562.

Ozaki, H., Watanabe, Y., Takahashi, K., Kitamura, K., Tanaka, A., Urase, K., Momoi, T.,Sudo, K., Sakagami, J., Asano, M., Iwakura, Y., Kawakami, K., 2001. Six4, a putativemyogenin gene regulator, is not essential for mouse embryonal development. Mol.Cell. Biol. 21, 3343–3350.

Papanayotou, C., Mey, A., Birot, A.M., Saka, Y., Boast, S., Smith, J.C., Samarut, J., Stern, C.D.,2008. A mechanism regulating the onset of Sox2 expression in the embryonicneural plate. PLoS Biol. 6, e2.

Pappu, K.S., Ostrin, E.J., Middlebrooks, B.W., Sili, B.T., Chen, R., Atkins, M.R., Gibbs, R.,Mardon, G., 2005. Dual regulation and redundant function of two eye-specificenhancers of the Drosophila retinal determination gene dachshund. Development132, 2895–2905.

Pera, E., Stein, S., Kessel, M., 1999. Ectodermal patterning in the avian embryo:epidermis versus neural plate. Development 126, 63–73.

Phillips, B.T., Kwon, H.J., Melton, C., Houghtaling, P., Fritz, A., Riley, B.B., 2006. ZebrafishmsxB, msxC andmsxE function together to refine the neural–nonneural border andregulate cranial placodes and neural crest development. Dev. Biol. 294, 376–390.

Quint, E., Zerucha, T., Ekker, M., 2000. Differential expression of orthologous Dlx genesin zebrafish and mice: implications for the evolution of the Dlx homeobox genefamily. J. Exp. Zool. 288, 235–241.

Ruf, R.G., Xu, P.X., Silvius, D., Otto, E.A., Beekmann, F., Muerb, U.T., Kumar, S., Neuhaus, T.J., Kemper, M.J., Raymond Jr., R.M., Brophy, P.D., Berkman, J., Gattas, M., Hyland, V.,Ruf, E.M., Schwartz, C., Chang, E.H., Smith, R.J., Stratakis, C.A., Weil, D., Petit, C.,Hildebrandt, F., 2004. SIX1 mutations cause branchio-oto-renal syndrome bydisruption of EYA1–SIX1–DNA complexes. Proc. Natl. Acad. Sci. U. S. A. 101,8090–8095.

Sahly, I., Andermann, P., Petit, C., 1999. The zebrafish eya1 gene and its expressionpattern during embryogenesis. Dev. Genes Evol. 209, 399–410.

Sasaki, H., Hogan, B.L., 1996. Enhancer analysis of the mouse HNF-3 beta gene:regulatory elements for node/notochord and floor plate are independent andconsist of multiple sub-elements. Genes Cells 1, 59–72.

171S. Sato et al. / Developmental Biology 344 (2010) 158–171

Sato, S., Nakamura, M., Cho, D.H., Tapscott, S.J., Ozaki, H., Kawakami, K., 2002.Identification of transcriptional targets for Six5: implication for the pathogenesis ofmyotonic dystrophy type 1. Hum. Mol. Genet. 11, 1045–1058.

Sauka-Spengler, T., Bronner-Fraser, M., 2008. A gene regulatory network orchestratesneural crest formation. Nat. Rev. Mol. Cell Biol. 9, 557–568.

Schlosser, G., 2006. Induction and specification of cranial placodes. Dev. Biol. 294,303–351.

Schlosser, G., 2008. Do vertebrate neural crest and cranial placodes have a commonevolutionary origin? Bioessays 30, 659–672.

Schlosser, G., Ahrens, K., 2004. Molecular anatomy of placode development in Xenopuslaevis. Dev. Biol. 271, 439–466.

Schlosser, G., Awtry, T., Brugmann, S.A., Jensen, E.D., Neilson, K., Ruan, G., Stammler, A.,Voelker, D., Yan, B., Zhang, C., Klymkowsky, M.W., Moody, S.A., 2008. Eya1 and Six1promote neurogenesis in the cranial placodes in a SoxB1-dependent fashion. Dev.Biol. 320, 199–214.

Schug, J., 2008. Using TESS to predict transcription factor binding sites in DNA sequence.Curr. Protoc. Bioinform. Chapter 2, Unit 2 6.

Sheng, G., Stern, C.D., 1999. Gata2 and Gata3: novel markers for early embryonicpolarity and for non-neural ectoderm in the chick embryo. Mech. Dev. 87, 213–216.

Shimamura, K., Rubenstein, J.L., 1997. Inductive interactions direct early regionalizationof the mouse forebrain. Development 124, 2709–2718.

Sjodal, M., Edlund, T., Gunhaga, L., 2007. Time of exposure to BMP signals plays a key rolein the specification of the olfactory and lens placodes ex vivo. Dev. Cell 13, 141–149.

Solomon, K.S., Fritz, A., 2002. Concerted action of two dlx paralogs in sensory placodeformation. Development 129, 3127–3136.

Streit, A., 2002. Extensive cell movements accompany formation of the otic placode.Dev. Biol. 249, 237–254.

Streit, A., 2004. Early development of the cranial sensory nervous system: from acommon field to individual placodes. Dev. Biol. 276, 1–15.

Streit, A., 2007. The preplacodal region: an ectodermal domain with multipotentialprogenitors that contribute to sense organs and cranial sensory ganglia. Int. J. Dev.Biol. 51, 447–461.

Streit, A., Stern, C.D., 1999. Establishment and maintenance of the border of the neuralplate in the chick: involvement of FGF and BMP activity. Mech. Dev. 82, 51–66.

Takemoto, T., Uchikawa, M., Kamachi, Y., Kondoh, H., 2006. Convergence of Wnt andFGF signals in the genesis of posterior neural plate through activation of the Sox2enhancer N-1. Development 133, 297–306.

Taneyhill, L.A., Bronner-Fraser, M., 2005. Dynamic alterations in gene expression afterWnt-mediated induction of avian neural crest. Mol. Biol. Cell 16, 5283–5293.

Tumpel, S., Cambronero, F., Sims, C., Krumlauf, R., Wiedemann, L.M., 2008. A regulatorymodule embedded in the coding region of Hoxa2 controls expression inrhombomere 2. Proc. Natl. Acad. Sci. U. S. A. 105, 20077–20082.

Uchikawa, M., Ishida, Y., Takemoto, T., Kamachi, Y., Kondoh, H., 2003. Functionalanalysis of chicken Sox2 enhancers highlights an array of diverse regulatoryelements that are conserved in mammals. Dev. Cell 4, 509–519.

Ulsamer, A., Ortuno, M.J., Ruiz, S., Susperregui, A.R., Osses, N., Rosa, J.L., Ventura, F., 2008.BMP-2 induces Osterix expression through up-regulation of Dlx5 and itsphosphorylation by p38. J. Biol. Chem. 283, 3816–3826.

Weirauch, M., Hughes, T., 2010. Conserved expression without conserved regulatorysequence: the more things change, the more they stay the same. Trends Genet. 26,66–74.

White, R., Ziman, M., 2008. Genome-wide discovery of Pax7 target genes duringdevelopment. Physiol. Genomics 33, 41–49.

Woda, J.M., Pastagia, J., Mercola, M., Artinger, K.B., 2003. Dlx proteins position theneural plate border and determine adjacent cell fates. Development 130, 331–342.

Xu, P.X., Adams, J., Peters, H., Brown, M.C., Heaney, S., Maas, R., 1999. Eya1-deficientmice lack ears and kidneys and show abnormal apoptosis of organ primordia. Nat.Genet. 23, 113–117.

Xu, H., Dude, C.M., Baker, C.V., 2008. Fine-grained fate maps for the ophthalmic andmaxillomandibular trigeminal placodes in the chick embryo. Dev. Biol. 317,174–186.

Yang, L., Zhang, H., Hu, G., Wang, H., Abate-Shen, C., Shen, M.M., 1998. An early phase ofembryonic Dlx5 expression defines the rostral boundary of the neural plate. J.Neurosci. 18, 8322–8330.

Yokouchi, Y., Ohsugi, K., Sasaki, H., Kuroiwa, A., 1991. Chicken homeobox gene Msx-1:structure, expression in limbbudsandeffectof retinoic acid.Development113,431–444.

Zhang, H., Hu, G., Wang, H., Sciavolino, P., Iler, N., Shen, M.M., Abate-Shen, C., 1997.Heterodimerization of Msx and Dlx homeoproteins results in functional antago-nism. Mol. Cell. Biol. 17, 2920–2932.

Zheng, W., Huang, L., Wei, Z.B., Silvius, D., Tang, B., Xu, P.X., 2003. The role of Six1 inmammalian auditory system development. Development 130, 3989–4000.

Zou, D., Silvius, D., Fritzsch, B., Xu, P.X., 2004. Eya1 and Six1 are essential for early stepsof sensory neurogenesis in mammalian cranial placodes. Development 131,5561–5572.