Developmental Biology 343 (2010) 94–103

Contents lists available at ScienceDirect

Developmental Biology

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

prep1.2 and aldh1a2 participate to a positive loop required for branchial archesdevelopment in zebrafish

Enrico Vaccari, Gianluca Deflorian 1, Elisa Bernardi, Stefan Pauls 2, Natascia Tiso,Marino Bortolussi, Francesco Argenton ⁎Dipartimento di Biologia, Università degli Studi di Padova, Via Bassi 58B, Padova, Italy

⁎ Corresponding author.E-mail address: francesco.argenton@unipd.it (F. Arg

1 IFOM, FIRC Institute of Molecular Oncology, Via Ada2 School of Biological and Chemical Sciences. Queen

E14NS UK.

0012-1606/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.ydbio.2010.04.016

a b s t r a c t

a r t i c l e i n f o

Article history:Received for publication 13 March 2009Revised 15 April 2010Accepted 16 April 2010Available online 25 April 2010

Keywords:VertebrateZebrafishEndodermSegmentationPrepRetinoic acidPharyngeal arches

Segmentation is a key step in embryonic development. Acting in all germ layers, it is responsible for thegeneration of antero-posterior asymmetries. Hox genes, with their diverse expression in individual segments,are fundamental players in the determination of different segmental fates. In vertebrates, Hox gene productsgain specificity for DNA sequences by interacting with Pbx, Prep and Meis homeodomain transcriptionfactors. In this work we cloned and analysed prep1.2 in zebrafish. In-situ hybridization experiments showthat prep1.2 is maternally and ubiquitously expressed up to early somitogenesis when its expression patternbecomes more restricted to the head and trunk mesenchyme. Experiments of loss of function with prep1.2morpholinos change the shape of the hyoid and third pharyngeal cartilages while arches 4–7 and pectoralfins are absent, a phenotype strikingly similar to that caused by loss of retinoic acid (RA). In fact, we showthat prep1.2 is positively regulated by RA and required for the normal expression of aldh1a2 at later stages,particularly in tissues involved in the development of the branchial arches and pectoral fins. Thus, prep1.2and aldh1a2 are members of an indirect positive feedback loop required for pharyngeal endoderm andposterior branchial arches development. As the paralogue gene prep1.1 is more important in hindbrainpatterning and neural crest chondrogenesis, we provide evidence of a functional specialization of prep genesin zebrafish head segmentation and morphogenesis.

enton).mello 16, 20139 Milan, Italy.Mary University of London.

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

Introduction

The emergence of endodermal pouches and the integratedpharyngeal arches provided vertebrates with a rapidly evolvingstructure suited to enhance their adaptation. Indeed, pharyngealdevelopment and evolution involve derivatives from all three germlayers: neural crest, pharyngeal pouches, muscles and blood vessels.Craniofacial development starts with the migration of bilateralstreams of neural crest cells (NCC) in the prospective head region ofthe embryo. These cells migrate into the pharyngeal arches wherethey differentiate in skeletal structures including the jaws and theirsupport (Le Lievre and Le Douarin, 1975). Indeed, recent studies haveshown that pharyngeal endoderm playsmore roles in pharyngeal archdevelopment than previously acknowledged and is indeed responsi-ble for cranial NCC patterning (Barrallo-Gimeno et al., 2004;Begemann et al., 2001; Couly et al., 2002; Crump et al., 2004; Davidet al., 2002; Holzschuh et al., 2005; Kopinke et al., 2006; Piotrowski

and Nusslein-Volhard, 2000; Popperl et al., 2000; Ruhin et al., 2003;Trainor and Krumlauf, 2000). While studies of zebrafish mutants orknockout mice and molecular screenings in Xenopus have contributedto our understanding of endoderm formation (Alexander and Stainier,1999; Dickmeis et al., 2001; Gritsman et al., 1999; Yasuo and Lemaire,1999) the genetic andmolecular mechanisms controlling endodermalpatterning are less understood (reviewed in Grapin-Botton, 2005;Kiefer, 2003; Stainier, 2005). Interestingly, recent findings in zebrafishclearly demonstrate that retinoic acid (RA) is required for propermorphogenesis and segmentation of the endodermal pouches(Kopinke et al., 2006). RA modulates gene expression by ligand-activated transcription factors, the RA receptors (RARs) that can formheterodimers with retinoid X receptors (RXRs) (Chambon, 1996).RARs/RXRs bind as asymmetrically-oriented heterodimers to specificDNA target sequences that are named Retinoic Acid ResponsiveElements (RAREs). RARE sequences are composed of two directrepeats (DR) of a core hexameric motif, PuG(G/T)TCA (Leid et al.,1992; Mangelsdorf and Evans, 1995). A typical RARE sequence is a 5-bp-spaced direct repeat (DR5), however, RAR/RXR heterodimers alsobind to direct repeats separated by 1 bp (DR1) or 2 bp (DR2). It isnoteworthy that RA is known to regulate Hox genes by means ofRAREs and, by acting through Hox gene expression, it is also involvedin a/p patterning of the endoderm (Huang et al., 1998; Schubert et al.,2005; Yelon and Stainier, 2002).

95E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

Segmentation is an essential and evolutionarily conserved step inembryonic development. Acting in all three germ layers, it isresponsible for the generation and determination of different partsof the body. With their combined expression in different segments,Hox genes are key players in the determination of different segmentalfates (Krumlauf, 1994; Moens and Prince, 2002). The regionalizationof the vertebrate hindbrain, one of the best characterized models ofvertebrate segmentation, relies on the rhombomere-specific expres-sion of a subset of Hox genes. Hox gene products gain high specificityfor DNA target sequences by interacting with Pbc (Pbx in vertebrates)and other homeodomain proteins (Burglin, 1997; Kamps et al., 1990;Moens and Selleri, 2006; Nourse et al., 1990). Moreover, Prep (alsoknown as Pknox) and Meis homeodomain transcription factors of theMeinox family by interacting with Pbc are also involved in the Hoxregulation machinery (Berthelsen et al., 1998; Burglin, 1997). As Pbcmolecules use different domains to interact with Hox and Meinoxproteins, Pbc–Hox and Meinox––Pbc molecular interactions drive theformation of Meinox–Pbc–Hox trimers with high DNA bindingspecificity.

In zebrafish at least two partially redundant pbx genes (pbx2 andpbx4), three prep genes (prep1.1, prep1.2, and prep2) and five meisgenes, (meis1, meis2.1, meis2.2, meis3, meis4.1a and meis4.1b) areexpressed during early embryogenesis (Deflorian et al., 2004;Waskiewicz et al., 2001; Waskiewicz et al., 2002). The two pbxmembers cooperate with hox genes to drive the expression of earlyhindbrain patterning genes and, notably, the elimination of allhindbrain-expressed Pbx proteins (Pbx2 and Pbx4) uncovers ahindbrain ground state in which rhombomeres 2 to 6 (r2–r6) acquirean r1 identity (Waskiewicz et al., 2002). Dominant negative Meisderivatives and down-regulation of prep1.1 reduce the level of Pbxproteins (Choe et al., 2002; Deflorian et al., 2004; Mercader et al.,1999; Waskiewicz et al., 2001). On the other hand, meis1 gain-of-function increases the stability of Pbx proteins (Vlachakis et al., 2001;Waskiewicz et al., 2001).

Inactivation of pbx4/lazarus (lzr) in zebrafish leads to an embry-onic-lethal phenotype withmajor developmental defects in hindbrainsegmentation and cranial neural crest determination, the latter beingcaused by the lack of pharyngeal endoderm segmentation (Popperlet al., 2000). In a previous work, our group has demonstrated thatprep1.1 is necessary for hindbrain patterning and cranial neural crestchondrogenesis but dispensable for segmentation of the pharyngealendoderm (Deflorian et al., 2004). In this work we have found thatprep1.2 plays an essential role in pharyngeal endoderm segmentation,patterning of pharyngeal arches and pectoral fin bud development. Infact, prep1.2 loss of function phenotype is very similar to that obtainedby loss of RA (Grandel et al., 2002). Moreover, by manipulating theirlevels, we could demonstrate that RA, prep1.2 and aldh1a2 aremembers of an indirect positive feedback loop required for pharyngealendoderm and posterior branchial arches segmentation.

Material and methods

Fish maintenance, microinjection and treatments

Embryos were grown and staged at 28.5 °C according to standardprocedures (http://zfin.org). In the glucacon:GFP transgenic line Tg(gcga:GFP)ia1 (Zecchin et al., 2007) 2.7 kb of the 5′ region of theglucagon gene drives the expression of the GFP reporter. Thesequence of the zebrafish preproglucagon cDNA (Argenton et al.,1999; GenBank accession number: AJ133697) was used to design twoprimers, 5′-GTCAAACCCGGAGCCCTCG (GluB) and 5′-GGCAGTGG-GACGCAAATGC (GluC) to screen a zebrafish PAC library (DeutschesRessourcenzentrum fur Genomforschung RZPD, library number 706).This led to the isolation of the PAC clone BUSMP706F21177Q2 thatincluded the preproglucagon coding sequence. We sequenced the first2 kb upstream of the coding region by using the primers GluB, 5′-

TCATTTCATTGATGTAAATGACTACAG (Glu-rev1), 5′-ACATGTA-CAAGTTTGCTCACCTGG (Glu169Re), 5′-GGCTGTAGGCGATGGCTG(Glu176Re) and 5′-GCTCAACCTGCGATCTGTGTGG (Glu112Re). This2 kb region was amplified by 5′-CCACAAGTGTGCAGCATCCACC(Glu107Fw) and 5′-TTAACAGCCCAGTCTTCCAACACAC (Glu2146Re)and subcloned into the TOPO TA cloning vector (Invitrogen). By usingprimers 5′-GGGAGGAGGAATCGGATACAC (Glu114Re) and 5′-CGCAT-TATAGTGGCTATGCAGG (Glu160Re) we were able to sequenceanother 700 bp fragment upstream to the 2 kbp. The 700 bp wasthen amplified using the primer pair 5′-GGCATAGGCTAGTCAGCATG(Glu556Fw) and 5′-CTCACAGGCATCGGCGACTC (Glu1602Re). TheDNA resulting from amplification was cloned into the TOPO TAcloning vector (Invitrogen) and fused to the 2 kb fragment using anendogenous BglII-site. This procedure led to the isolation of 2.7 kb ofthe zebrafish preproglucagon promoter/enhancer that was cloned infront of a GFP coding sequence (pGI cloning vector, (Gilmour et al.,2002). The resulting construct was used to generate stable transgeniczebrafish lines. We were able to isolate two independent insertions ofthe transgene (glu10:GFP and glu36:GFP) displaying identical expres-sion patterns. Individuals used for this study belong to the glu36:GFPline, hereafter called glucagon:GFP, because GFP expression in this lineis stronger compared to the glu10:GFP line. For injection experiments,messengers of prep1.2 and prep1.1 were synthesized from linearizedvectors using SP6 RNA polymerase (mMessage mMachine in vitrotranscription kit, Ambion). Morpholino antisense oligos wereobtained from Gene Tools.

The morpholino (MO) oligos used in our experiments were:

MO1-prep1.2: 5′-GTCATCATAGTTACTGTTGCCGTGG-3′;MO1-prep1.2-5MIS: 5′-GTgATCtTAcTTAgTGTTGCgGTGG-3′;MO2-prep1.2: 5′-CTATAAAGCTGATCTTCAGCACCGC-3′MO2-prep1.2-6MIS: 5′-CTAaAAcCTcATtTTCAGgACgGC -3′

MO-p53: 5′ - GCGCCATTGCTTTGCAAGAATTG - 3′ (MO4-tp53 Robuet al., 2007); MO1-aldh1a2 (Grandel et al., 2002). The stock wasdiluted to working concentration of 1–4 mg/ml in Danieau solution,before the injection into the yolk of one-cell stage embryos. In allmicroinjection experiments phenol red or rodamine dextran wereused to check the efficiency of the procedure.

Quantification of transcriptional activities

To obtain the p50-RARE-Luc vector, the region containing the RAREwas cloned into HindIII and PstI sites of ptPRL-50-Luc (Argenton et al.,1996a) using the following primers:

5′-cccaagcttGCACAAAGTTTTGATTGACAGC-3′5′-aactgcagTTTCAATATTTGTCGGCTCATTT-3′

The plasmid obtained was sequenced to verify presence andaccuracy of prep1.2-RARE.

To obtain the negative control p50-RAREmut-Luc plasmid wemutagenized prep1.2-RARE sequence using “QuickChange™ XL Site-Directed Mutagenesis” (Stratagene). The primer used was:

5′-AATAGGAGTGTAAAAAATTGCAAAGATGATCTTCAAACGTTGC-3′.

Zebrafish embryos at 1-cell stage were injected with 30 pg of eachvector (p50-Luc, p50-RARE-Luc and p50-RAREmut-Luc). A batch ofinjected embryos was treated from 50% epiboly until early somito-genesis with 10−7 M RA. Embryos were rinsed at 24 hpf, divided ingroups of four and lysed in 0.02 ml of Report Lysis Buffer. Sampleswere frozen at −80 °C for 5 min and centrifuged at 12,000 g. Thesupernatant was assayed using the protocol described in Argentonet al. (1996a). For each injection and treatment we tested at least8 different lysates in duplicate.

96 E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

Real time quantitative PCR was done using 7500 Real Time PCRSystem (Applera). SYBR green was used to quantify cDNA. Primerswere:

prep1.2 forward: 5′-CAGTCAGGAGGACGGTTCGTCTA-3′;prep1.2 reverse: 5′-GGATAAGGGTGACCAATGTGCTG-3′;hoxb1b forward 5′-CCCCTAAAACAGTTAAAGTCGCCGA-3′;hoxb1b reverse: 5′-GTGAAACTCCTTCTCAAGTTCCGTG-3′;eF1 forward: 5′-GCGGTACTACTCTTCTTGATGCCC-3′;eF1 reverse: 5′-ACAGGTACAGTTCCAATACCTCCA-3′.

tbx1 and b-actin primers were the same as in Zhang et al. (2006).The relative expression of prep1.2, hoxb1b and tbx1 were computedwith respect to the amount of b-actin and eF1 in each sample.Standard amplification conditions were used.

For stage-specific RT-PCR (Fig. 1), prep1.2 forward, prep1.2 reverseand b-actin primers were used, omitting reverse transcriptase in thenegative control (24 hpf embryos).

Manipulation of RA synthesis and levels

Pharmacological inhibition of retinoic acid synthesis was per-formed by incubating live embryos in their chorions in embryomedium supplemented with DEAB (4-(diethylamino)benzaldehyde;Sigma-Aldrich; no. D86256) at final concentrations of 10−5 M understandard embryo culture conditions in the dark (Perz-Edwards et al.,2001). DEAB stock solution was diluted in DMSO (Fisher) and storedat a concentration of 10 mM at −20 °C. As control, other embryos

Fig. 1. Expression of prep1.2 during development. A, B: RT-PCR analysis showing amplificaexperiments with a prep1.2 antisense probe showing that prep1.2 is weakly and ubiquitouslyprep1.2 becomes more concentrated in cells localized on each side of the neural tube. H: At 2(forebrain andmidbrain), eyes and in cell clusters localized on each side of the NS, posteriorpectoral fin bud mesenchyme. J: From 48 hpf onwards prep1.2 expression decreases and is rvesicle; pba: prospective branchial arches; pfb: pectoral fin bud. D: Top viewwith ventral sidEmbryos are in lateral view.

were treated at the same time in 0.1% DMSO only. Exogenousapplication of RA was performed by incubating the embryos in asolution of RA 10−7 M in embryo medium containing 0.1% DMSO inthe dark (Zhang et al., 1996).

Characterization and radiation hybrid panel mapping of prep1.2

In the GenBank database for zebrafish expressed sequence tags wefound a 678 bp long EST AF38294 corresponding to a part of theprep1.2 cDNA. Using genomic sequence from clone BUSM1-69G10 wedesigned primers to amplify the complete coding sequence by twostep RT-PCR. The primers used to amplify prep1.2with or without theSTOP codon were:

oligo1: 5′-cgggatccGCCCCAGCGTTGTCCACGG-3′oligo2: 5′-cgggatccCAGCGAGTCGCTGGTCTCC-3′oligo3: 5′-cgggatccCTACAGCGAGTCGCTGGTC-3′.

The product was cloned in PCRII®-TOPO® Vector (Invitrogen) andsequenced. The full coding prep1.2was subcloned in the BamHI site ofpCS2+ (to obtain pCS-prep1.2) and in pCS2GFP (to obtain the pCS-prep1.2-GFP plasmid). The primers used to amplify the cDNA fragmentcontaining a prep1.2-specific riboprobe, used for in situ hybridization,were: 5 ′-CTATTGCAGGACAACAACTGG-3 ′ and 5 ′-CAGC-GAGTCGCTGGTCTCCAG-3′.

In order to clone and sequence the prep1.2 5′UTR region we used“5′/3′ RACE” kit (Roche®) and the following primers: 5′-CAGTAAC-TATGATGACATCCCC-3′; 5′-CATCAGATTGTCCAAATCCAGGG-3′; 5′-AGCAGCGCCAACAGGGGG-3′.

tion of prep1.2 from morula to 48 hpf stages. C–J: Whole-mount in situ hybridizationexpressed from 128 cells to tailbud stage (panels from C to F). G: during somitogenesis4 hpf prep1.2 transcripts remain weakly ubiquitous with stronger expression in the CNSto the otic vesicle. I: at 36 hpf prep1.2 is expressed in the head and in the pharyngeal andestricted to the head region. e: eye; h: hindbrain; m: midbrain; nt: neural tube; ov: otice to the left. E: Frontal view. F: lateral view. G and H: Embryos are in dorsal view. I and J:

97E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

To map the position of prep1.2 we used the Zebrafish/HamsterRadiation Hybrid Panel (Goodfellow T51 panel, Invitrogen). The panelwas screened by PCR using primers 5′-CTATTGCAGGACAACAACTGG-3′ and 5′-GAATTCGTAACACTTGGATGGCC-3′, and following themanufacturer's instructions. The retention profile of the PCR reactionwas analyzed by the RH Instant Mapper program (http://zfrhmaps.tch.harvard.edu/ZonRHmapper/Default.htm).

Whole-mount in situ hybridization and histochemistry

Embryos were staged according to Kimmel et al. (1995). Whole-mount in situ hybridizations were performed according to Thisse et al.(1993). Double staining whole-mount in situ hybridization wascarried out according to Hauptmann and Gerster (2000). Whole-mount immunohistochemistry was performed with anti-Zn5 anti-bodies essentially as described by Piotrowski and Nüsslein-Volhard(2000). The following digoxygenin- or fluorescein-labelled antisenseriboprobes were used: prep1.2, prep1.1 (Deflorian et al., 2004), dlx2a(Akimenko et al., 1994), her5 (Bally-Cuif et al., 2000), hoxb1b(Alexandre et al., 1996), foxD3 (Kelsh et al., 2000), dlx2a (Akimenkoet al., 1994), fgf3 (Kiefer et al., 1996), and aldh1a2 (Grandel et al.,2002). For cartilage staining, larvae were fixed overnight in 4% PBS-buffered p-formaldehyde, rinsed in distilled water and stainedovernight in a 0.1% Alcian blue solution. Larvae were then clearedby washing sequentially in 3% hydrogen peroxide and 70% glycerol,and whole mounted.

Results

Isolation, sequencing and mapping of prep1.2

The prep1.2 cDNA was isolated by RT-PCR from total RNA ofzebrafish embryos pooled at different developmental stages. Takingadvantage of the whole zebrafish genome assembly (www.ensebl.org/Danio_rerio), we found that the coding region of prep1.2 has 10exons, and an exon–intron organization identical to that of prep1.1.The prep1.2 cDNA encodes for a protein of 439 amino acids, with twoMeis homology regions (HR1–2) at the N-terminus and a conservedTALE homeodomain (HD) at the C-terminus (Supplemental Fig. 1).Sequence comparison at the amino acid level and phylogeneticanalysis of the Meis/Prep proteins family confirmed that zebrafishPrep1.2 is more similar to zebrafish Prep1.1 than zebrafish Prep2.Moreover, an extended glutamic acid stretch is present at the N-terminal region of Prep1.2, while a glutamic acid-rich region ispresent in the C-terminal region of both zebrafish and mammalianPrep1.1 proteins. This different feature might confer to Prep1.2specific functions. We mapped prep1.2 gene at 59 cRay on chromo-some 1, at 59 cRay from chromosome telomere, in a region syntenic tomouse chromosome 17 (Supplemental Fig. 1). The nearest markerwas ZC020N14.ya.

Expression of prep1.2 mRNA during embryogenesis

The expression profile of prep1.2 at different developmental stageswas analyzed by RT-PCR and in situ hybridisation. The transcripts ofprep1.2 could be amplified from total RNA from embryos at all stagesconsidered from cleavage to 48 hpf, indicating maternal and zygoticexpression of the gene (Fig. 1A). Using an antisense probe directedagainst prep1.2-specific sequences, spanning both the glutamic acidstretch and the HR1 domain, we found that prep1.2 mRNA isexpressed weakly and ubiquitously in 128 cell-, shield- and tailbud-stage embryos (Figs. 1C–F). At 18 hpf its expression is moreconcentrated in two stripes of mesenchyme flanking the neural tube(Fig. 1G). Then at 24 hpf prep1.2 transcripts were mainly found in theforebrain and midbrain, eyes and the prospective branchial region(Fig. 1H). From 36 hpf to 48 hpf prep1.2 expression decreased

markedly and was mainly restricted to the head and pectoral fins(Figs. 1I and J).

Morphological defects of prep1.2 knock-down embryos

In order to study the role of prep1.2 during zebrafish development,we designed an antisensemorpholino (MO1-prep1.2) complementaryto a 25-bp sequence in the 5′ region of prep1.2 mRNA spanning theATG. A second morpholino (MO5MIS-prep1.2) obtained by mutatingMO1-prep1.2 in 5 positionswas used as a negative control (Fig. 2A). Totest the morpholino specificity we injected one-cell embryos with20 pg of a prep1.2-GFP mRNA, encoding the Prep1.2-GFP fusionprotein, together with different concentrations of either MO1-prep1.2or MO5MIS-prep1.2. Embryos co-injected with 4 ng of MO1-prep1.2showed no GFP signal at 24 hpf, while those co-injected with MO5MIS-prep1.2 or receiving prep1.2-GFP mRNA alone expressed the reportergene in cell nuclei as expected for a transcription factor (Figs. 2B and C).In 5-day-old morphant larvae the branchial tissue was defective, theorganization of the jaw was impaired and the pectoral fins werestrongly reduced or completely absent (Figs. 2D and E, Table 1).Alcian blue staining of craniofacial cartilages of 5-day-old morphantlarvae confirmed and highlighted some aspects of prep1.2 morphantphenotype. Meckel's and hyoid cartilages were smaller andmisshaped, the third pharyngeal cartilage (p3) was strongly reducedand the more posterior ones (p4–7) were absent (Figs. 2F–I). Thephenotype was obtained with two independent morpholinostargeting different non-overlapping regions of prep1.2 mRNA, itwas highly reproducible, dose-dependent and never observed incontrol embryos injected with either MO1-prep1.2-5MIS or MO2-prep1.2-6MIS (Table1). Morpholino specificity has been also con-trolled with experiments in which the prep1.2 morphant phenotypehas been successfully rescued by injecting prep1.2 mRNA. Finally,control experiments with and without tp53 morpholino (Robu et al.,2007) showing that prep1.2 morphants have a small amount of celldeath that might partially contribute to the phenotype (seeSupplemental Fig. 2).

The induction and specification of neural crest cells in prep1.2morphants

In order to get insight into the role of prep1.2 in neural crest cell(NCC) development we analyzed the expression of NCC markers inprep1.2 morphants. As shown in Figs. 3A and B, the expression of thepre-migratory NCC marker foxD3was normal in prep1.2morphants at10 somites suggesting that prep1.2 knock-down does not affect theinduction of NCCs (see also Supplemental Fig. 3). Next, we analyzedthe expression of dlx2awhich is expressed in three distinct streams ofcranial NCCsmigrating into I) themandibular, II) the hyoid and III) theposterior pharyngeal arches (p3–7), respectively. In prep1.2 mor-phants a normal pattern of dlx2a expression was detected initially(Figs. 3C–F), but, at later stages, in the third stream we could detectonly the few anteriormost dlx2-expressing cells (Figs. 3G–J).

prep1.2 knock-down affects the pharyngeal endoderm segmentation

The endoderm is critically involved in patterning the pharyngealregion in zebrafish (Piotrowski and Nusslein-Volhard, 2000). In bothvgo (Piotrowski and Nusslein-Volhard, 2000) and lazarus (Popperlet al., 2000) mutants an impaired pharyngeal endoderm segmenta-tion is associated to a defective development of the pharyngealskeleton. To better dissect the role of prep1.2 in patterning thepharyngeal region we analyzed the development of pharyngealpouches in morphants obtained by injecting MO1-prep1.2 in gluca-gon:GFP transgenic zebrafish Tg(gcga:GFP)ia1 that express GFP in thepharyngeal endoderm (Zecchin et al., 2007). In such morphants,analyzed at 36 hpf, only the first and second pouches occur while the

Fig. 2. Morphological defects of prep1.2 morphant embryos. A: Schematic design ofantisense morpholinos and prep1.2-GFP construct. The morpholinos are complemen-tary to 25 bp in the 5′ region spanning the ATG (MO1-prep1.2) or to 25 bp in the 5′UTRregion (MO2-prep1.2); MO2-prep1.2 is non-overlapping with the 5′UTR of the prep1.2-GFP mRNA injected and can only block endogenous prep1.2 mRNA. Two controlmorpholino oligos were designed. They are mutated in 5 (MO1-prep1.2-5MIS) or 6(MO2-prep1.2-6MIS) positions. B, C: Each prep1.2 morpholino was co-injected withprep1.2-GFP mRNA to test its specificity; only MO1-prep1.2 blocks prep1.2-GFPtranslation in injected embryos (compare C with control morpholino in B) as revealedby the GFP signal in cell nuclei of the head region. D, E; day 5 larvae injected withMO1-prep1.2 display morphological defects: lack of branchial tissue, impaired jaworganization, absence of pectoral fins and pigmentation defects. F–I; Alcian bluestaining of 5-day-old larvae shows thatMekel's cartilage is normal while hyoid cartilagewas misshaped and reduced in size. Morphant embryos also display a reduction of thethird pharyngeal cartilage while the more posterior ones are completely absent. ba:branchial arches; pf: pectoral fin; cb: ceratobranchial; ch: ceratohyal; e: ethmoid plate;me: Meckel's cartilage; tr: trabeculae cranii. D and E: Embryos are in dorsal view. F andG: Embryos are in ventral view. H and I: embryos are in lateral view.

Table 1Morphological (A) and molecular marker (B) analysis of embryos injected withdifferent morpholinos and mRNAs, treated with DEAB or mutants of the nls strain.Numbers represent the amount of embryos analyzed in each batch. The abnormalphenotype is represented within the figures of this work. ba: branchial arches; pf:pectoral fin.

A

Dose of morpholino 2 ng 3 ng 4 ng

N. of embryos 87 76 229% with normal N. of ba 44.8 19.7 9.6% with 1st ba only 29.9 64.5 77.7% with reduced N. of ba 25.3 15.8 12.7% with normal pf 68.3 46.2 21% with no or reduced pf 31.7 53.8 79

B

Marker Allele/Morpholinoand/or mRNA

Treatment n Normalphenotype

Abnormalphenotype

fgf3 55 55 0DEAB 46 0 46 (100%)

MO1-prep1.1 16 16 0MO1-prep1.2 20 2 18 (90%)

dlx2a 38 38 0MO1-prep1.1 22 22 0MO1-prep1.2 38 5 33 (87%)MO2-prep1.2 108 21 87 (80%)MO2-prep1.26MIS 68 63 5 (7%)MO2-prep1.2+prep1.2-GFPmRNA

107 67 30 (28%)

Zn5 37 37 0DEAB 44 0 44 (100%)

MO1-prep1.2 30 3 27 (90%)prep1.2 50 50 0

RA 42 0 42 (100%)DEAB 40 1 39 (98%)

MO1-aldh1a2 24 3 21 (88%)nls 38 28 10 (26%)

prep1.1 22 22 0RA 20 20 0

MO1-aldh1a2 23 23 0AlcianBlue

MO2-prep1.2 104 25 79 (76%)MO2-prep1.26MIS 50 42 8 (16%)MO2-prep1.2+prep1.2-GFPmRNA

92 69 23 (25%)

98 E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

posterior ones (ep3–5) are lacking (Figs. 3K and L). The dependence ofposterior pharyngeal endoderm segmentation by prep1.2 was con-firmed using Zn5, an antibody specific for the pharyngeal endoderm(Figs. 3M and N) (Piotrowski andNusslein-Volhard, 2000).We furtherinvestigated the expression of Fgf3 which, together with Fgf8, is a keyregulator of cartilage development and is expressed in the pharyngeal

endoderm (Crump et al., 2004). It can be observed that in prep1.2morphants the impaired segmentation correlates with a loss of fgf3expression in the posterior pharyngeal region (Figs. 3O and P). Takentogether, these data suggest that Prep1.2 regulates the differentiationof posterior cranial NCCs through the control of posterior pharyngealendoderm segmentation.

Regulation of prep1.2 by retinoic acid

RA, which is involved in the control of hox genes expression, issynthesized from retinal by Retinaldehyde Dehydrogenase (Aldh). Inzebrafish aldh1a2 mutants (Begemann et al., 2001; Grandel et al.,2002), as in prep1.2morphants, the patterning of branchial arches andpectoral fins is impaired. We therefore investigated whether Prep1.2is involved in the RA pathway that controls the patterning of thesestructures in zebrafish.

As first step, we analysed the expression pattern of prep1.2 at 24hpf both in wild type embryos, incubated in RA from 50% epiboly toearly somitogenesis, and in embryos defective in the RA pathway. Itcan be observed (Figs. 4A, C and E), that prep1.2 transcription isanteriorized and increased in wild type RA-treated embryos, andselectively reduced in the prospective posterior pharyngeal regionof aldh1a2 morphants or mutants and DEAB-treated embryos

Fig. 3. Characterization of prep1.2 function. A–J: Expression pattern of pre-migratoryand migratory NCC markers. A, B: prep1.2 knock-down does not affect foxd3 expressionin 10 somite stage embryos (compare A and B); C–F: dlx2a expression in NCC isunaffected in prep1.2morphants analysed at 10 and 20 somite stages; G and H: at 24 hpfexpression of dlx2a is strongly downregulated in the third stream of cephalic NCC ofMO1-prep1.2 injected embryos; I and J: similar results are obtained in 48 hpf embryos:in prep1.2morphant embryos dlx2a expression disappears in posterior branchial arches(J). K–P: Failure of the posterior pharyngeal endoderm segmentation in prep1.2morphant embryos. K–N: Analysis of pharyngeal pouches development in the glu:GFPtransgenic line that express GFP ectopically in the pharyngeal endoderm and inembryos stained with Zn5 antibodies. L: in MO1-prep1.2 injected embryos analyzed at36 hpf, the more posterior pouches (ep3–ep5) do not develop (compare with control inK). Identical results were obtained when using Zn5 antibodies to stain for thepharyngeal endoderm (compare N with control in M, see also Supplemental Fig. 5). Oand P; lack of fgf3 expression in the pharyngeal endoderm of prep1.2morphants (F). A–F: dorsal view; G–P lateral view. b: branchial arches; ep: endodermal pouches; h: hyoidcartilage; m: Meckel's cartilage; ov: otic vesicle.

99E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

(Supplemental Fig. 4). Noteworthy, the expression of the prep1.1gene, the prep1.2 paralog, is unaffected in both RA-treated embryosand aldh1a2 morphants (Figs. 4B, D and F).

To substantiate the role of RA on prep1.2 expression, we assessedby qRT-PCR the transcription of prep1.2 in embryos treated witheither RA or the pan-Retinaldehyde Dehydrogenase inhibitor DEAB(Niederreither et al., 1999). We also tested the expressions of hoxb1b(Alexandre et al., 1996; Maves and Kimmel, 2005) and tbx1 (Zhanget al., 2006) as positive and negative controls, respectively. As shownin Fig. 4G, the expression of prep1.2 is more than doubled by RAtreatment, but unaffected by DEAB. This apparent paradox can beexplained considering that RA, added to the medium, is acting in thewhole embryo, including the anterior regions where prep1.2 expres-sion is normally RA-independent. Vice versa, RA depletion by DEABtreatment, would cause a decrease of prep1.2 levels only in the narrowregionwhere its expression is regulated by RA. It is to be observed thatthe expression of the positive control hoxb1b is markedly enhanced byRA and downregulated by DEAB as reported by Alexandre et al. (1996)and Maves et al., (2005), respectively, while the down-regulation oftbx1 by RA treatment, is in agreement with Zhang et al. (2006).

Identification and characterization of a 3′-RARE element in the firstintron of prep1.2

The upregulation of prep1.2mRNA by RA prompted us to verify thehypothesis of whether RA can directly regulate prep1.2. Since RAregulates its target genes by binding to Retinoic Acid ResponseElements (RAREs) located in their promoters or enhancers, wechecked whether such elements occurred in the 5′ region of theprep1.2 gene. Indeed, we found a putative DR1 RARE (AGTT-CAaAGTTCA) for RAR/RXR heterodimers located in the first introndownstream to the transcriptional start site and 1734 bp upstream tothe prep1.2 ATG.

To assess the activity of the prep1.2 RARE, a 500-bp sequence of theprep1.2 intron-1, encompassing the RARE, was cloned into a reporterplasmid (p50-Luc) bearing the luciferase gene driven by the basalpromoter (50 bp) of the trout prolactin gene (Argenton et al., 1996b).The resulting plasmid (p50-RARE-Luc) and a control plasmid in whichthe RARE was mutated in 4 positions (p50-RAREmut-Luc) wereinjected separately into one-cell stage embryos which were eithertreated with 10−7 RA from 50% epiboly to early somitogenesis or leftuntreated. The luciferase activity was then measured on pools ofinjected embryos homogenized at 24 hpf. As shown in Fig. 4H, theactivities of p50-Luc and p50-RAREmut-Luc in injected embryos weresimilar and unaffected by the RA treatment. Conversely, comparisonof the luciferase activities of p50-Luc and p50-RARE-Luc, indicates thatthe prep1.2 RARE enhances three times the basal activity of the troutprolactin promoter in RA-treated embryos, but represses the samepromoter in untreated embryos. The latter effect is consistent with thefact that RAR/RXR heterodimers act as silencers in the absence of RA(Horlein et al., 1995).

Common role of prep1.2 and RA in patterning the posterior pharyngealendoderm

As our experiments showed that the prep1.2 gene is positivelyregulated by RA, we reasoned that the pharyngeal endoderm defectsobserved in prep1.2 morphants could be related to impaired RAsignalling. We therefore compared pharyngeal endoderm develop-ment of embryos either treated with DEAB or injected with prep1.2morpholinos. As shown both by in situ hybridization, using a fgf3probe, and immuno-assay, using the Zn5 antibody, the resultingembryos displays an identical phenotype, failing to develop theposterior pharyngeal pouches (Fig. 5; compare Figs. 3M and N withSupplemental Fig. 5).

Fig. 4. prep1.2 expression is positively regulated by retinoic acid. A, C and E: In situ hybridization experiments showing that prep1.2mRNA in able to respond to RA treatment. In RA-treated embryos analyzed at 24 hpf the prep1.2 mRNA signal (blue, highlighted with a bar) is anteriorized and its intensity is significantly increased (C). In MO1-aldh1a2 injectedembryos, prep1.2 expression is reduced, especially in the prospective branchial region (bar) (E). prep1.1 expression does not respond to RA (D) and is unaffected in aldh1a2morphantembryos (F). G: qRT-PCR assay of prep1.2, hoxb1b and tbx1 expression in RA- and DEAB-treated embryos analyzed at 24 hpf. prep1.2 and hoxb1b are about 3-fold upregulated, whiletbx1 is downregulated, in embryos treated with RA. DEAB treatment does not affect the expression of prep1.2 and tbx1 while hoxb1b is downregulated. The expression levels arecompared with the housekeeping genes b-actin and eF1. H: Reporter analysis of the prep1.2 RARE. The basal promoter used in this study (p50-Luc) is unresponsive to RA. The 0.5 kbpintron-1 fragment bearing the putative prep1.2 RARE (p50-RARE-Luc) represses the basal promoter activity but confers a 7 fold RA-responsiveness. Mutagenesis of the RAREsequence of the 0.5 kbp intron-1 fragment restores the basal promoter activity.

100 E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

The endodermal pouch precursors are a cell population expressingher5 situated anterior to the midbrain–hindbrain boundary atcompletion of epiboly (Kopinke et al., 2006). As shown in Supple-

Fig. 5. Pharyngeal endoderm analysis of DEAB-treated embryos and prep1.2morphants. A–C:embryos (B) and MO1-prep1.2 morphants (C) lack the posterior endodermal pouches. ep:

mental Fig. 6, her5 expressionwas not altered in prep1-morphants andDEAB-treated embryos at 100% epiboly, excluding the involvement ofprep1.2 and RA in the specification of endodermal pouch precursors.

Zebrafish embryos are treated with DEAB from blastula stage up to 48 hpf. DEAB-treatedendodermal pouches.

Fig. 6. aldh1a2 and hoxb1b expression is regulated by prep1.2. A–H: Starting from 15somites to 24 hpf the expression of aldh1a2 in embryos injected with MO1-prep1.2 isreduced in branchial arches (black arrowhead), while aldh1a2 expression is unaffected inthe trunk. In situ hybridization of 24hpf embryos shows that aldh1a2expression is lost alsoin pectoral fin bud (*) of prep1.2 morphant embryos. I–L: At 24 hpf hoxb1b expression isdownregulated in branchial arches ofprep1.2morphant embryos.M,N:hoxb1b expressionisunaffectedat early developmental stages inMO1-prep1.2 injected embryos.Non-specifictoxic effects of morpholino were excluded by co-injecting MO-p53 (H and L). ov: oticvesicle; pba: prospective branchial arches; pfb: prospective fin bud.

101E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

Hence, we conclude that prep1.2 and RA are both involved in thesegmentation of the pharyngeal endoderm but not in its specification.

Prep1.2 regulates aldh1a2 in branchial arches and pectoral fin

The recent demonstration that Pbx and Meis proteins regulatealdh1a2 expression in zebrafish (French et al., 2007) prompted us toverify whether also prep1.2, a well known Hox and Pbx partner, isinvolved in the regulation of this gene (Berthelsen et al., 1998;Burglin, 1997). To this purpose we analysed aldh1a2 expression inprep1.2 morphants at several developmental stages. As shown inFig. 6, analysis between 15 somites to 24 hpf stages shows that theexpression of aldh1a2 in prep1.2 morphants is significantly reducedfrom the 20 somite stage in the prospective branchial region andpectoral fin buds, but unaffected in the head and trunk (Fig. 6A–H).These results demonstrate that prep1.2 is both regulated by RA andrequired for maintenance of aldh1a2 expression in tissues involved inthe development of the branchial arches and pectoral fins. Finally,assuming that the only function of prep1.2 in patterning of thebranchial skeleton is to control RA activity, we tested whether lowdoses of RA treatment could rescue the prep1.2 morphant phenotype.Results showed that RA is unable to rescue posterior endodermalpouches in the absence of Prep1.2 function (not shown) suggestingthat prep1.2 has more function in this region other than the control ofRA synthesis. Thus, prep1.2 and aldh1a2 participate in a compositepositive feedback loop required for pharyngeal endoderm patterning,as well as segmentation and development of posterior branchialarches.

In mouse, Hoxa1 and Hoxb1 are RA target genes which aredownregulated in the pharyngeal endoderm and mesoderm ofmutant aldh1a2 embryos (Niederreither et al., 2003). In order toconfirm that abnormal patterning of the posterior pharyngeal regionin prep1.2 morphants is due to the alteration of the RA-Hox pathway,we analysed the expression of the hoxb1b gene which is transcribed inthe prospective pharyngeal region from gastrulation to somitogenesis(Alexandre et al., 1996). As shown in Fig. 6, the expression of hoxb1bin prep1.2morphants, was unaffected at tailbud stage (Fig. 6 M and N)but specifically impaired in the prospective pharyngeal region at 24hpf (Fig. 6I–L).

Discussion

In the present work we examined the expression, function andregulation of prep1.2 in Danio rerio and compared it with its paralogueprep1.1. The analysis of two zebrafish prep1 genes gave us theopportunity to dissect some details on how duplicated genes havefacilitated the evolution of vertebrate-specific features, such aspharyngeal pouches in a model which is less affected by pleiotropyof phenotypes as the partial genetic redundancy of zebrafish genomehas led to subfunctionalization (Force et al., 1999).

Our in-silico analysis of the Prep1 sequences highlighted a relevantdifference between the N-terminus of the two Prep1 proteins(Supplemental Fig. 1), indicating that Prep1 proteins have acquiredspecific subfunctions during embryogenesis. Differences between theexpression patterns of the two prep1 genes were revealed by in situhybridization. While, prep1.1 mRNA is ubiquitously and uniformlyexpressed up to 48 hpf (Deflorian et al., 2004), prep1.2mRNA, startingfrom late somitogenesis, is mainly concentrated in specific tissues(Figs. 1G–I). Initially, the prep1.2morphants looked normal, but at 4–5dpf the morphant larvae displayed morphological defects such as lackof posterior branchial tissue, impaired organization of the jaw andcomplete absence of pectoral fins (Fig. 2 and Table 1). Alcian bluestaining of craniofacial cartilages of prep1.2 morphants showed thatMeckel's cartilage was quite normal, while the hyoid cartilage wasreduced in size and misshaped (Fig. 2F–I). Moreover, prep1.2morphants showed a strong reduction of the third pharyngeal

cartilage and the absence of the more posterior ones (p4–7)(Fig. 2G). Interestingly, a similar situation was observed in the pbx4mutant lazarus (lzr), in which the third and more posterior arch-

102 E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

derived cartilages are entirely absent and certain characteristics in themorphology of the second arch cartilages makes themmore similar tothose of the first arch (Popperl et al., 2000). Furthermore, pectoral finsfail to be specified in lzr as well. The defects observed in prep1.2morphants at the level of the branchial arches and fin bud weresurprisingly similar to those described in DEAB-treated embryos(Kopinke et al., 2006) and in the aldh1a2 mutant alleles no-fin (nof)and neckless (nls) (Grandel et al., 2002; Linville et al., 2004).

Recent studies demonstrated that loss of RA signalling atgastrulation causes a loss of the post-otic NCC due to an impairedspecification of posterior rhombomeres (Begemann et al., 2004;Costaridis et al., 1996; Gavalas and Krumlauf, 2000; Kopinke et al.,2006; Linville et al., 2004; Maves and Kimmel, 2005). However, whenRA synthesis is inhibited starting from 16 hpf, the hindbrain isproperly segmented, the post-otic NCC migrate normally in thepharyngeal region but the posterior pharyngeal pouches 3–6 fail todevelop (Kopinke et al., 2006). Comparing our results with theseobservations, we propose that Prep1.2, acting in a positive feedbackloopwith aldh1a2, is a key factor in the post-otic NCC differentiation, aprocess indeed depending on “late RA signalling”. Moreover,considering that a failure of pharyngeal endoderm development hasbeen also described in mouse and zebrafish mutants of pbx genes(Manley et al., 2004; Popperl et al., 2000) as well as in our prep1.2morphant embryos (Fig. 3), it is tempting to speculate that Prep1.2, inassociation with Pbx and Hoxb1b proteins, plays an essential role inthe Aldh1a2/RA-mediated morphogenesis of posterior pharyngealendoderm. Accordingly, an essential role of Pbx proteins has beenalready demonstrated in the maintenance of aldh1a2 expression inthe retina (French et al., 2007). Vice versa, in prep1.1 morphants thepharyngeal endoderm develops normally while hindbrain segmenta-tion is severely affected (Deflorian et al., 2004). Thus, our results onPrep1 functions indicate that prep1.2 knock-down prevents thechondrogenesis of the more posterior branchial arches by impairingthe formation of endodermal pouches, while the inactivation ofprep1.1 inhibits chondrogenesis in all pharyngeal arches, withoutperturbing the expression of the most cranial NCC specificationmarkers (Deflorian et al., 2004). On the other hand, while prep1.2 hasonly few essential function in the hindbrain (Supplemental Fig. 7),prep1.1 is necessary for hindbrain segmentation and patterning aswell as for neural crest chondrogenesis but is unresponsive to RA(Fig. 4) and has no role in pharyngeal endoderm segmentation(Deflorian et al., 2004).

Our data show that Prep1.2 is involved in several fundamentalaspects of zebrafish embryogenesis giving the first in vivo evidence ofan interaction between RA signalling and a member of the Meinoxfamily. prep1.2 is positively regulated by RA in vivo and a 3′-RAREelement situated in the first intron is, possibly, the mediator of suchpositive regulation. On the other hand, we show that prep1.2 isnecessary for the maintenance of aldh1a2 expression in thepharyngeal region. Thus, we have identified a regional feedbackloop with a core role in endoderm segmentation and subsequentspecification of cranial neural crest cells, which is based on RA andother known essential genetic factors acting in the pharyngeal region(namely Prep1.2, Pbx and Hoxb1b).

Finally, we would like to point out that the inability of RA to rescuePrep1.2 morphants is the evidence that Prep1.2, which is a possibledownstream target of RA, has several unique functions in branchialregion development and NCC patterning, other than control of RAsynthesis. On the other hand, the fact that prep1.2 mRNA injectionalone in unable to rescue no-fin/neckless (not shown) or raldhmorphants phenotype is also the proof that other factors activatedby RA (e.g. Hoxb1b) are essential for development of the branchialregion. In conclusion: one of the target of RA in the pharyngeal regionis prep1.2 and Prep1.2 has some essential functions in the pharyngealregion, different from Prep1.1, among which is Aldh1a2 expression inthe pharyngeal mesenchyme.

Acknowledgments

We are grateful to Francesco Blasi, Dirk Meyer and TatjanaPiotrowski, for critical reading of the manuscript and to the wholeZF community for its generous support with probes, morpholinos andsuggestions. We would like to thank Luigi Pivotti, Giovanna Pagninand Raffaella Picco for fish husbandry and laboratory management.This work is supported by EU LSH-CT-2003-503496 ZF MODELS andItalian MIUR Prin-2004055332.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.ydbio.2010.04.016.

References

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.

Alexander, J., Stainier, D.Y., 1999. A molecular pathway leading to endoderm formationin zebrafish. Curr. Biol. 9, 1147–1157.

Alexandre, D., Clarke, J.D., Oxtoby, E., Yan, Y.L., Jowett, T., Holder, N., 1996. Ectopicexpression of Hoxa-1 in the zebrafish alters the fate of the mandibular arch neuralcrest and phenocopies a retinoic acid-induced phenotype. Development 122,735–746.

Argenton, F., Arava, Y., Aronheim, A., Walker, M.D., 1996a. An activation domain of thehelix–loop–helix transcription factor E2A shows cell type preference in vivo inmicroinjected zebra fish embryos. Mol. Cell. Biol. 16, 1714–1721.

Argenton, F., Ramoz, N., Charlet, N., Bernardini, S., Colombo, L., Bortolussi, M., 1996b.Mechanisms of transcriptional activation of the promoter of the rainbow troutprolactin gene by GHF1/Pit1 and glucocorticoid. Biochem. Biophys. Res. Commun.224, 57–66.

Argenton, F., Zecchin, E., Bortolussi, M., 1999. Early appearance of pancreatic hormone-expressing cells in the zebrafish embryo. Mech Dev. 87 (1–2), 217–221.

Bally-Cuif, L., Goutel, C., Wassef, M., Wurst, W., Rosa, F., 2000. Coregulation of anteriorand posterior mesendodermal development by a hairy-related transcriptionalrepressor. Genes Dev. 14, 1664–1677.

Barrallo-Gimeno, A., Holzschuh, J., Driever, W., Knapik, E.W., 2004. Neural crest survivaland differentiation in zebrafish depends on mont blanc/tfap2a gene function.Development 131, 1463–1477.

Begemann, G., Marx, M., Mebus, K., Meyer, A., Bastmeyer, M., 2004. Beyond the necklessphenotype: influence of reduced retinoic acid signaling on motor neurondevelopment in the zebrafish hindbrain. Dev. Biol. 271, 119–129.

Begemann, G., Schilling, T.F., Rauch, G.J., Geisler, R., Ingham, P.W., 2001. The zebrafishneckless mutation reveals a requirement for raldh2 in mesodermal signals thatpattern the hindbrain. Development 128, 3081–3094.

Berthelsen, J., Zappavigna, V., Ferretti, E., Mavilio, F., Blasi, F., 1998. The novelhomeoprotein Prep1 modulates Pbx-Hox protein cooperativity. Embo J. 17,1434–1445.

Burglin, T.R., 1997. Analysis of TALE superclass homeobox genes (MEIS, PBC, KNOX,Iroquois, TGIF) reveals a novel domain conserved between plants and animals.Nucleic Acids Res. 25, 4173–4180.

Chambon, P., 1996. A decade of molecular biology of retinoic acid receptors. Faseb J. 10,940–954.

Choe, S.K., Vlachakis, N., Sagerstrom, C.G., 2002. Meis family proteins are required forhindbrain development in the zebrafish. Development 129, 585–595.

Costaridis, P., Horton, C., Zeitlinger, J., Holder, N., Maden, M., 1996. Endogenousretinoids in the zebrafish embryo and adult. Dev. Dyn. 205, 41–51.

Couly, G., Creuzet, S., Bennaceur, S., Vincent, C., Le Douarin, N.M., 2002. Interactionsbetween Hox-negative cephalic neural crest cells and the foregut endoderm inpatterning the facial skeleton in the vertebrate head. Development 129,1061–1073.

Crump, J.G., Maves, L., Lawson, N.D., Weinstein, B.M., Kimmel, C.B., 2004. An essentialrole for Fgfs in endodermal pouch formation influences later craniofacial skeletalpatterning. Development 131, 5703–5716.

David, N.B., Saint-Etienne, L., Tsang, M., Schilling, T.F., Rosa, F.M., 2002. Requirement forendoderm and FGF3 in ventral head skeleton formation. Development 129,4457–4468.

Deflorian, G., Tiso, N., Ferretti, E., Meyer, D., Blasi, F., Bortolussi, M., Argenton, F., 2004.Prep1.1 has essential genetic functions in hindbrain development and cranialneural crest cell differentiation. Development 131, 613–627.

Dickmeis, T., Mourrain, P., Saint-Etienne, L., Fischer, N., Aanstad, P., Clark, M., Strahle, U.,Rosa, F., 2001. A crucial component of the endoderm formation pathway,CASANOVA, is encoded by a novel sox-related gene. Genes Dev. 15, 1487–1492.

Force, A., Lynch, M., Pickett, F.B., Amores, A., Yan, Y.L., Postlethwait, J., 1999.Preservation of duplicate genes by complementary, degenerative mutations.Genetics 151 (4), 1531–1545.

French, C.R., Erickson, T., Callander, D., Berry, K.M., Koss, R., Hagey, D.W., Stout, J.,Wuennenberg-Stapleton, K., Ngai, J., Moens, C.B., Waskiewicz, A.J., 2007. Pbx

103E. Vaccari et al. / Developmental Biology 343 (2010) 94–103

homeodomain proteins pattern both the zebrafish retina and tectum. BMC Dev Biol.7, 85.

Gavalas, A., Krumlauf, R., 2000. Retinoid signalling and hindbrain patterning. Curr. Opin.Genet. Dev. 10, 380–386.

Gilmour, D.T., Maischein, H.M., Nusslein-Volhard, C., 2002. Migration and function of aglial subtype in the vertebrate peripheral nervous system. Neuron 34, 577–588.

Grandel, H., Lun, K., Rauch, G.J., Rhinn, M., Piotrowski, T., Houart, C., Sordino, P., Kuchler,A.M., Schulte-Merker, S., Geisler, R., Holder, N., Wilson, S.W., Brand, M., 2002.Retinoic acid signalling in the zebrafish embryo is necessary during pre-segmentation stages to pattern the anterior–posterior axis of the CNS and toinduce a pectoral fin bud. Development 129, 2851–2865.

Grapin-Botton, A., 2005. Antero-posterior patterning of the vertebrate digestive tract:40 years after Nicole Le Douarin's PhD thesis. Int J Dev Biol. 49, 335–47.

Gritsman, K., Zhang, J., Cheng, S., Heckscher, E., Talbot, W.S., Schier, A.F., 1999. The EGF-CFC protein one-eyed pinhead is essential for nodal signaling. Cell 97, 121–132.

Hauptmann, G., Gerster, T., 2000. Multicolor whole-mount in situ hybridization.Methods Mol. Biol. 137, 139–148.

Holzschuh, J., Wada, N., Wada, C., Schaffer, A., Javidan, Y., Tallafuss, A., Bally-Cuif, L.,Schilling, T.F., 2005. Requirements for endoderm and BMP signaling in sensoryneurogenesis in zebrafish. Development 132, 3731–3742.

Horlein, A.J., Naar, A.M., Heinzel, T., Torchia, J., Gloss, B., Kurokawa, R., Ryan, A., Kamei, Y.,Soderstrom,M., Glass, C.K., et al., 1995. Ligand-independent repression by the thyroidhormone receptormediated by a nuclear receptor co-repressor. Nature 377, 397–404.

Huang, D., Chen, S.W., Langston, A.W., Gudas, L.J., 1998. A conserved retinoic acidresponsive element in the murine Hoxb-1 gene is required for expression in thedeveloping gut. Development 125, 3235–3246.

Kamps, M.P., Murre, C., Sun, X.H., Baltimore, D., 1990. A new homeobox genecontributes the DNA binding domain of the t(1;19) translocation protein in pre-BALL. Cell 60, 547–555.

Kelsh, R.N., Schmid, B., Eisen, J.S., 2000. Genetic analysis of melanophore developmentin zebrafish embryos. Dev. Biol. 225, 277–293.

Kiefer, J.C., 2003. Molecular mechanisms of early gut organogenesis: a primer ondevelopment of the digestive tract. Dev. Dyn. 228, 287–291.

Kiefer, P., Strahle, U., Dickson, C., 1996. The zebrafish Fgf-3 gene: cDNA sequence,transcript structure and genomic organization. Gene 168, 211–215.

Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., Schilling, T.F., 1995. Stages ofembryonic development of the zebrafish. Dev. Dyn. 203, 253–310.

Kopinke, D., Sasine, J., Swift, J., Stephens, W.Z., Piotrowski, T., 2006. Retinoic acid isrequired for endodermal pouch morphogenesis and not for pharyngeal endodermspecification. Dev. Dyn. 235, 2695–2709.

Krumlauf, R., 1994. Hox genes in vertebrate development. Cell 78, 191–201.Le Lievre, C.S., Le Douarin, N.M., 1975. Mesenchymal derivatives of the neural crest:

analysis of chimaeric quail and chick embryos. J Embryol Exp Morphol. 34, 125–154.Leid, M., Kastner, P., Chambon, P., 1992. Multiplicity generates diversity in the retinoic

acid signalling pathways. Trends Biochem. Sci. 17, 427–433.Linville, A., Gumusaneli, E., Chandraratna, R.A., Schilling, T.F., 2004. Independent roles

for retinoic acid in segmentation and neuronal differentiation in the zebrafishhindbrain. Dev. Biol. 270, 186–199.

Mangelsdorf, D.J., Evans, R.M., 1995. The RXR heterodimers and orphan receptors. Cell83, 841–850.

Manley, N.R., Selleri, L., Brendolan, A., Gordon, J., Cleary, M.L., 2004. Abnormalities ofcaudal pharyngeal pouch development in Pbx1 knockout mice mimic loss of Hox3paralogs. Dev. Biol. 276, 301–312.

Maves, L., Kimmel, C.B., 2005. Dynamic and sequential patterning of the zebrafishposterior hindbrain by retinoic acid. Dev. Biol. 285, 593–605.

Mercader, N., Leonardo, E., Azpiazu, N., Serrano, A., Morata, G., Martinez, C., Torres, M.,1999. Conserved regulation of proximodistal limb axis development by Meis1/Hth.Nature 402, 425–429.

Moens, C.B., Prince, V.E., 2002. Constructing the hindbrain: insights from the zebrafish.Dev. Dyn. 224, 1–17.

Moens, C.B., Selleri, L., 2006. Hox cofactors in vertebrate development. Dev. Biol. 291,193–206.

Niederreither, K., Remboutsika, E., Gansmuller, A., Losson, R., Dolle, P., 1999. Expressionof the transcriptional intermediary factor TIF1alpha during mouse developmentand in the reproductive organs. Mech. Dev. 88, 111–117.

Niederreither, K., Vermot, J., Le Roux, I., Schuhbaur, B., Chambon, P., Dollé, P., 2003. Theregional pattern of retinoic acid synthesis by RALDH2 is essential for thedevelopment of posterior pharyngeal arches and the enteric nervous system.130 (11), 2525–2534.

Nourse, J., Mellentin, J.D., Galili, N., Wilkinson, J., Stanbridge, E., Smith, S.D., Cleary, M.L.,1990. Chromosomal translocation t(1;19) results in synthesis of a homeobox fusionmRNA that codes for a potential chimeric transcription factor. Cell 60, 535–545.

Perz-Edwards, A., Hardison, N.L., Linney, E., 2001. Retinoic acid-mediated geneexpression in transgenic reporter zebrafish. Dev. Biol. 229, 89–101.

Piotrowski, T., Nusslein-Volhard, C., 2000. The endoderm plays an important role inpatterning the segmented pharyngeal region in zebrafish (Danio rerio). Dev. Biol.225, 339–356.

Popperl, H., Rikhof, H., Chang, H., Haffter, P., Kimmel, C.B., Moens, C.B., 2000. lazarus is anovel pbx gene that globally mediates hox gene function in zebrafish. Mol Cell. 6,255–267.

Robu, M.E., Larson, J.D., Nasevicius, A., Beiraghi, S., Brenner, C., Farber, S.A., Ekker, S.C.,2007. p53 activation by knockdown technologies. PLoS Genet. 3, e78.

Ruhin, B., Creuzet, S., Vincent, C., Benouaiche, L., Le Douarin, N.M., Couly, G., 2003.Patterning of the hyoid cartilage depends upon signals arising from the ventralforegut endoderm. Dev. Dyn. 228, 239–246.

Schubert, M., Yu, J.K., Holland, N.D., Escriva, H., Laudet, V., Holland, L.Z., 2005. Retinoicacid signaling acts via Hox1 to establish the posterior limit of the pharynx in thechordate amphioxus. Development 132, 61–73.

Stainier, D.Y., 2005. No organ left behind: tales of gut development and evolution.Science 307, 1902–1904.

Thisse, C., Thisse, B., Schilling, T.F., Postlethwait, J.H., 1993. Structure of the zebrafishsnail1 gene and its expression in wild-type, spadetail and no tail mutant embryos.Development 119, 1203–1215.

Trainor, P.A., Krumlauf, R., 2000. Patterning the cranial neural crest: hindbrainsegmentation and Hox gene plasticity. Nat Rev Neurosci. 1, 116–124.

Vlachakis, N., Choe, S.K., Sagerstrom, C.G., 2001. Meis3 synergizes with Pbx4 andHoxb1b in promoting hindbrain fates in the zebrafish. Development 128,1299–1312.

Waskiewicz, A.J., Rikhof, H.A., Hernandez, R.E., Moens, C.B., 2001. Zebrafish Meisfunctions to stabilize Pbx proteins and regulate hindbrain patterning. Development128, 4139–4151.

Waskiewicz, A.J., Rikhof, H.A., Moens, C.B., 2002. Eliminating zebrafish pbx proteinsreveals a hindbrain ground state. Dev Cell. 3, 723–733.

Yasuo, H., Lemaire, P., 1999. A two-step model for the fate determination ofpresumptive endodermal blastomeres in Xenopus embryos. Curr. Biol. 9, 869–879.

Yelon, D., Stainier, D.Y., 2002. Pattern formation: swimming in retinoic acid. Curr. Biol.12, R707–R709.

Zecchin, E., Filippi, A., Biemar, F., Tiso, N., Pauls, S., Ellertsdottir, E., Gnugge, L., Bortolussi,M., Driever, W., Argenton, F., 2007. Distinct delta and jagged genes controlsequential segregation of pancreatic cell types from precursor pools in zebrafish.Dev. Biol. 301, 192–204.

Zhang, L., Zhong, T., Wang, Y., Jiang, Q., Song, H., Gui, Y., 2006. TBX1, a DiGeorgesyndrome candidate gene, is inhibited by retinoic acid. Int. J. Dev. Biol. 50, 55–61.

Zhang, Z., Balmer, J.E., Lovlie, A., Fromm, S.H., Blomhoff, R., 1996. Specific teratogeniceffects of different retinoic acid isomers and analogs in the developing anteriorcentral nervous system of zebrafish. Dev. Dyn. 206, 73–86.