Zebrafish admp Is Required to Restrict the Size of the Organizer and to Promote Posterior and Ventral Development

Transcription

1 DEVELOPMENTAL DYNAMICS 222: (2001) BRIEF COMMUNICATIONS Zebrafish admp Is Required to Restrict the Size of the Organizer and to Promote Posterior and Ventral Development ZSOLT LELE, MATTHIAS NOWAK, AND MATTHIAS HAMMERSCHMIDT* Max-Planck-Institut fü rimmunbiologie, Stü beweg 51, Freiburg, Germany ABSTRACT Bone morphogenetic proteins (Bmps) and their roles during early dorsoventral patterning of the vertebrate embryo are well understood.theroleandregulationofamoredistant member of this family, the anti-dorsalizing morphogenetic protein (Admp), however, are less clear. Here, we report the isolation and characterization of zebrafish admp. Unlike other bmps, admp is exclusively expressed on the dorsal side. Expressionstartsatblastulastagesintheregionof the organizer, giving rise to anterior neuroectoderm and axial mesoderm. During the course of gastrulation, both the neuroectodermal and the mesodermal admp transcripts vanish in an anterior-posterior wave. The maintenance of admp expression is positively influenced by Nodal signaling and by Bozozok (Boz), an organizer-promoting homeodomain protein acting as a repressor of earlybmp2bexpression.despitethepositiveeffect of boz on admp expression, Boz and Admp have rather opposite effects on zebrafish patterning, as revealedingain-andloss-of-functionexperiments. Upon overexpression, admp has Bmp-like activitiescausingasmallerorganizerandenhancedventral specification, very similar to the phenotype caused by the loss of boz function in mutant embryos. Antisense-based admp knockdown, on the other side, leads to an enlarged organizer and impaired ventral and posterior development, as observed in embryos after boz overexpression. This finding indicates that admp is required for the development of embryonic structures normally suppressed by organizer activities. The seeming discrepancybetweentheregulativeandfunctional relationship of boz and admp is discussed, and models are proposed according to which Admp might be part of a negative feedback loop to pattern and confine the organizer region Wiley-Liss, Inc. Key words: Danio rerio; Admp; Bmp; Bozozok; Nodal; organizer INTRODUCTION Earlydorsoventral(DV)patterningisoneofthemost thoroughly studied areas of developmental biology. There are many unanswered questions, however, regardingtheintegrationofdifferentsignalingpathways affecting this process. Bmp signaling is one of the key factors establishing dorsoventral (DV) polarity during late blastula and early gastrula stages. Studies in Xenopus and zebrafish have demonstrated that Bmps act as morphogens inducing ventral cell fates in adosedependent manner (Harland and Gerhart, 1997; Dosch et al., 1997). On the dorsal side, bmp expression is repressed by the homeodomain protein Bozozok (Boz, also called Dharma or Nieuwkoid; Koos and Ho, 1999; Fekany-Lee et al., 2000). In addition, organizer cells secrete Bmp inhibitors such as Chordin, Noggin, and Follistatin (Mullins, 1999). The anti-dorsalizing morphogenetic protein (Admp), first isolated from Xenopus laevis (Moos et al., 1995), has been one of the most controversial members of the Bmp family. Unlike the rest of the family members, it is exclusively expressed on the future dorsal side of the embryo, initially in all three germ layers, then restricted to the axial mesoderm. Moreover, Admp is not inhibited by adominant-negative type IBMP receptor, Noggin or Chordin, but by Follistatin, suggesting that it does not act by means of the classic Bmp signaling pathway (Dosch and Niehrs, 2000). Results obtained with dominant negative versions indicate that Admp promotes formation of trunk and tail by inhibiting head-organizingactivities(doschandniehrs,2000).in chick embryos, Admp was shown to act as an inhibitor of anode-inducing center, thereby preventing the formation of supernumerary organizers (Joubin and Stern, 1999). However, the regulation of admp expression and its interconnection with other regulators of early patterning have remained unclear. Here, we report the isolation and characterization of azebrafish admp homologue. We show that its expression is posi- Grant sponsor: Max-Planck Society; Grant sponsor: Human Frontier Science Program; Grant number: 0354/1999-M. Dr. Lele s present address is Department of Anatomy and Developmental Biology, University College London, Gower Street, London, WC1E 6BT UK. *Correspondence to: Matthias Hammerschmidt, Max-Planck-Institut fü rimmunbiologie, Stü beweg 51, D Freiburg, Germany. hammerschmid@immunbio.mpg.de Received 10 August 2001; Accepted 12 September 2001 Published online 2November WILEY-LISS, INC. DOI /dvdy.1222

2 682 LELE ET AL. Fig. 1. Expression of admp in wild-type zebrafish embryos analyzed by whole-mount in situ hybridization: (A) 40% epiboly, animal pole view; (B) shield stage, lateral view; (C) shield stage, dorsal view; (D) 70% epiboly, lateral view; (E) 70% epiboly, dorsal view; (F) tailbud stage, dorsal view (the embryo is slightly tilted upward to expose the tailbud region); (G) 8-somite stage, lateral view; (H) 8-somite stage, dorsal view of the tailbud region. Arrows in C,E,F point to expression in the mesoderm; arrowheads in C E point to expression in the neuroectoderm. tively regulated by the Bmp repressor Boz (Koos and Ho, 1999; Ryu et al., 2001), but that Admp itself has effects opposite to those of Boz. Thus, the phenotype of boz mutants can be phenocopied by overexpression of admp, whereas boz overexpression phenocopies traits of admp morphants. We present models according to which Admp acts like (and maybe even in concert with) Bmps on the dorsal side of the zebrafish embryo to attenuate organizer activities. RESULTS AND DISCUSSION By using sequence information from zebrafish ESTs, we carried out RACE-PCR to clone a full-length 1174 bp long admp cdna from gastrula stage RNA (see Experimental Procedures section; gene bank accession no. AF ). The predicted amino acid sequence is 67% identical to Xenopus Admp and 61% identical to chick Admp, but only 31% identical to zebrafish Bmp7, the closest related zebrafish protein, suggesting that it is indeed a zebrafish Admp homologue. BLAST and SMART searches identified an N-terminal signal sequence, a Tgf propeptide (amino acids [aa] ) with a consensus cleavage site, and a Tgf domain (aa ). Expression of zebrafish admp starts at 30% epiboly in the organizer region on the dorsal side of the embryo (Fig. 1A). At early and midgastrula stages, admp mrna is present in both the anterior neural ectoderm and the axial mesoderm (Fig. 1B E). During the further course of development, admp transcripts progressively vanish in an anteroposterior (AP) wave, starting in the neuroectoderm and followed by the mesoderm. Thus, at the tail bud stage, admp expression has become confined to the posterior half of the notochord anlage, whereas the neuroectoderm at the same AP levels is devoid of admp transcripts (Fig. 1F). Finally, at the 8-somite stage, admp expression in both the mesoderm and the neuroectoderm has become restricted to a few posterior axial cells (Fig. 1G,H), whereas by the 15-somite stage, admp transcripts have disappeared completely. This spatial and temporal expression pattern of zebrafish admp is very similar to that of admp during Xenopus embryogenesis (Moos et al., 1995). To address the function of admp during zebrafish development, we carried out gain- and loss-of-function experiments, injecting mrna or morpholino antisense oligonucleotides into one- to four-cell stage embryos. Upon injection of admp mrna, we observed two distinct phenotypes, depending on the injection site. First, we obtained moderately ventralized embryos with an increased number of blood cells and a split ventral tail fin (obtained in 37% of the injected embryos; 88 of 238; Fig. 2B). In addition, we obtained embryos characterized by a lack of head and anterior notochord, but no signs of ventralization (Fig. 2C; 31% of the injected embryos; 74 of 238). Finally, approximately 5% of the injected embryos (12 of 238) showed a combination of both phenotypes (Fig. 2D). The two distinct phenotypes appear to result from an unequal distribution of the injected mrna (as confirmed by co-injections of lineage tracers in a small number of embryos; not shown). When ectopically expressed on the ventral side, Admp

3 Fig. 2. admp overexpression leads to a reduction of the organizer and a ventralization of the embryo; admp knock-down leads to an enlarged organizer and a dorsalization of the embryo. A G: Live embryos at 60 hr postfertilization (hpf). H S: Whole-mount in situ hybridizations at indicated stages and with indicated probes. (A,H K) Un-injected controls; (B D,L O) embryos after injection of admp mrna ( pg/embryo); (E G,P S) embryos after injection of admp antisense morpholino oligonucleotide ( pmol/embryo). Arrows in B,D point to enlarged blood islands, a sign of ventralization; arrow in C to U-shaped somites, resulting from the absence of the notochord; arrowheads in C,D to the lost head; arrow in E to the broadened notochord; and arrowheads in F,G to the reduced tail. The shortened axis in admp mrna-injected embryos (B D) is due to the partial loss of the dorsal axis, resulting in impaired extension movements, similarly to the situation in dino mutants. H,L,P: gsc, 60% epiboly, animal view, dorsal right. Length of the gsc expression domain along the dorsoventral axis is indicated by bars. I,M,Q: flh, 70% epiboly, vegetal view, dorsal right. J,N,R: gsc (indicated by arrows) and eve1, 70% epiboly, vegetal view, dorsal right. Note that in comparison to P, the gsc expression domain in R has become narrower and thicker, most likely due to dorsal convergence movements of gastrulation. K,O,S: cdx1, 90% epiboly, lateral view, dorsal right. Length of the cdx1 along the animalvegetal axis on the ventral side of the embryos is indicated by bars.

4 684 LELE ET AL. appears to act as an enhancer of Bmp signaling, generating ventralized phenotypes as earlier described for dino (din) mutants, which lack the secreted Bmp inhibitor Chordin (Hammerschmidt et al., 1996; Schulte- Merker et al., 1997). However, when overexpressed dorsally, the site of normal transient admp expression, it generates a phenotype very similar to that of bozozok (boz) mutants (Fekany et al., 1999). During normal development, the boz gene is activated in dorsal cells by the canonical Wnt signaling pathway (Ryu et al., 2001). It encodes a homeodomain protein repressing bmp2b expression in dorsal cells of blastula stage embryos (Koos and Ho, 1999). Compared with din mutants, boz mutant embryos display a more severe head truncation, suggesting that boz, in addition to its bmp-repressing function, has a head-inducing activity, probably by inhibiting Wnt signals (Fekany-Lee et al., 2000). During gastrula stages, admp mrna-injected embryos have a smaller organizer, as revealed by the narrower expression domains of goosecoid (gsc), a marker for organizer cells giving rise to the anterior-most dorsal mesoderm (Stachel et al., 1993), and floating head (flh, Talbot et al., 1995), an organizer marker maintained in more posterior axial mesoderm, the notochord (Fig. 2H J,L N). A similar reduction in the gsc and flh expression domains has also been described for boz mutants (Fekany et al., 1999; Koos and Ho, 1999). In addition, admp mrna-injected embryos display a dorsal expansion of the expression domains of ventral marker genes, such as wnt8 (Kelly et al., 1995; data not shown) and eve1 (Joly et al., 1993; Fig. 2J,N), similar to the situation observed both in dino and boz mutants (Hammerschmidt et al., 1996; Koos and Ho, 1999). Thus, it seems that Admp primarily acts as an antagonist of organizer formation, leading to the later reduction of organizer-derived structures, such as head and notochord. The expansion of ventral fates might occur secondarily, due to the reduced generation of dorsalizing signals from the organizer, such as the Bmp inhibitors Chordin and Noggin, or the Wnt inhibitor Dickkopf (Hashimoto et al., 2000; Shinya et al., 2000). Alternatively, or in addition, it might result from a direct ventralizing effect of ectopically expressed admp, which would suggest that Admp can act in a Bmp-like manner. For admp loss-of-function analysis, we injected zebrafish embryos with antisense morpholino oligonucleotides (MOs) against the 5 -region of zebrafish admp. MOs have been proven to efficiently knock down specific gene products during zebrafish development by blocking translation (Nasevicius and Ekker, 2000). Increasing amounts of injected MOs produced a gradual loss of tail and trunk (Fig. 2E G), in parallel to a broadening of the notochord (Fig. 2E). A loss of posterior structures also occurs in dorsalized embryos, as caused by the loss of Bmp signaling (Mullins, 1999) or by overexpression of boz (Yamanaka et al., 1998; Koos and Ho, 1998). Like in embryos injected with boz mrna, but unlike in bmp mutants, admp morphant embryos display a significant enlargement of the organizer (Fig. 2P R), further supporting the notion that Admp acts as a negative regulator of organizer formation. This enlarged organizer in admp morphants should lead to an overproduction of dorsalizing signals, causing impaired ventral specification revealed by the reduced expression of eve1 (Fig. 2R). As in the various bmp mutants, this reduced ventral specification is accompanied by impaired posterior development, as anticipated by reduced expression of the posterior marker cdx1 (Joly et al., 1992) on the ventral side of admp morphant embryos (Fig. 2S). Altogether, the admp gain- and loss of function studies suggest that Admp, despite its dorsal expression, displays a dorsal-antagonizing activity opposite to that of Boz and other organizer factors. In light of this seeming contradiction between expression and activity, we were wondering how admp expression is regulated. Thus far, the regulation of admp expression had not been addressed in other species (Moos et al., 1995; Joubin and Stern, 1999). Taking advantage of zebrafish mutants, we studied the dependence of admp expression on three different factors required for organizer formation or function: Boz itself; Nodal signals, which act as mesoderm inducers (Feldman et al., 1998) and which have been reported to have a negative effect on head formation in Xenopus (Piccolo et al., 1999); and the Bmp inhibitor Chordin, which is generated by organizer cells and which is required for dorsal specification in dorsolateral cells adjacent to the organizer (Piccolo et al., 1996; Schulte-Merker et al., 1997; Miller-Bertoglio et al., 1997). To examine the effect of Nodal signaling on admp expression, we used the zebrafish one-eyed-pinhead (oep) mutant (Zhang et al., 1998; Gritsman et al., 1999). oep encodes a Cripto-like membrane-associated protein essential for Tgf /Nodal signaling. Embryos lacking both maternal and zygotic oep (mzoep) lack the mesendoderm of head and trunk, very similar to the zebrafish nodal-related1;nodal-related2 (squint;cyclops) double mutant (Feldman et al., 1998). As shown in Figure 3D, admp expression appears relatively normal in mzoep mutants of the 45% epiboly stage, indicating that Nodal signaling is not required for the initial induction of admp expression. However, at midgastrula stages (Fig. 3E,F), no admp transcripts can be detected in mzoep mutants, although wild-type embryos of the same stage show expression in both the mesoderm and the neuroectoderm. This finding indicates that maintenance of admp expression in the neural ectoderm requires Nodal signaling and/or the presence of functional mesendoderm. This positive effect of Nodals on admp expression is consistent with the finding in Xenopus that Nodal signaling like Admp antagonizes head formation (Piccolo et al., 1999). In contrast to Nodals and Admp, Boz promotes and is required for head formation (Yamanaka et al., 1998; Koos and Ho, 1998; Fekany et al., 1999; Koos and Ho, 1999). Nevertheless, boz and mzoep mutants display a very similar admp expression pattern. As in mzoep

5 Admp DURING ZEBRAFISH ORGANIZER FORMATION 685 Fig. 3. admp expression in positively regulated by Nodal signaling, Bozozok, and Chordin: (A C,J) wild-type, (D F) mzoep; (G I) boz m168 ; (K,L) din tt250. (A,G) 40% epiboly, animal view; (D) 45% epiboly, animal view; (B,E,H) 70% epiboly, animal view; (C,F,I,L) 70% epiboly, lateral view; (J,K) 70% epiboly, dorsal view. Note that in the din mutant shown in L, admp expression is restricted to the mesoderm in the inner cell layers, whereas the neuroectoderm in the outer cell layers lacks admp transcripts (in contrast to the wild-type [WT] sibling shown in C). mutants, early activation of admp expression is not affected in boz mutant embryos, whereas by 70% epiboly admp mrna has disappeared, indicating that maintenance of admp expression depends on boz function. We cannot rule out an even earlier requirement of boz for the initiation of admp expression, as the observed early expression of admp in zygotic boz mutants could be driven by maternally supplied boz gene products, similarly to how it has been shown for chordin (Fekany-Lee et al., 2000).

6 686 LELE ET AL. As pattern and regulation of chordin expression (Miller-Bertoglio et al., 1997) show significant similarities to those of admp, we wondered whether admp is regulated by boz by means of chordin. Therefore, we examined the expression of admp in din mutants. As in boz mutant embryos, early expression of admp is not affected by the loss of Chordin function (not shown). At 70% epiboly, however, admp mrna does not disappear completely in din mutants. Rather, only the neuroectodermal expression domain is lost while expression in the axial mesoderm remains unaffected, suggesting that at least the notochord-specific expression of admp is independent of chordin. Altogether, our data suggest that Oep, Boz, and Chordin are not required for the initiation of admp expression. Rather it seems likely that admp expression is initiated by -catenin mediated Wnt signaling, in parallel rather than downstream of boz (Ryu et al., 2001). However, it is noteworthy that the maintenance of admp expression is positively influenced by Boz, which promotes head and notochord formation and represses bmp expression, whereas Admp itself blocks head and notochord formation and can act like Bmps. Thus, it seems that boz, by enhancing admp expression, establishes a negative feedback loop to antagonize its own activity during organizer formation and function. Such a negative feedback might be necessary to restrict organizer activity, consistent with conclusions about the role of Admp drawn from results obtained in chick and Xenopus. In chick, it has been proposed that Admp present in Henson s node, the chick organizer, inhibits node-inducing signals from the primitive streak, thereby confining the organizer region and preventing the formation of additional or larger organizers (Joubin and Stern, 1999), whereas in Xenopus, Admp has been proposed to govern patterning within the organizer by inhibiting head-organizing and promoting trunk-organizing activities (Dosch and Niehrs, 2000). But how can such a negative feedback work, given that Admp and organizer activities are, at least temporarily, present in the same dorsal cells? With Admp being the only antagonist, a fine-tuning in the temporal and spatial regulation of Admp and organizer activities would be necessary to allow proper organizer function. This mechanism is very well possible, because gsc and other organizer-specific genes start to be expressed earlier than admp. Detailed comparative analyses of the expression patterns of admp and organizer genes have to be carried out to test this notion. Alternatively, it is possible that Admp fulfills its organizer-restricting action in concert with other signals. We propose that these partners could be Bmps. Comparing the expression patterns of bmps and admp at late blastula stages (Fig. 1A and Dick et al., 2000), it appears that two zones are established within the early organizer; a central zone expressing admp, but not bmps (due to the bmp-repressing action of boz), and a lateral zone coexpressing bmps and admp. If Admp can indeed act in concert with Bmps (as suggested by the ventralizing effect of admp; see above), this differential expression pattern should affect organizer acitivity, with Admp and Bmps from the lateral zone antagonizing organizer factors, thereby leading to a spatial restriction of organizer activity. In boz mutants, due to the lack of bmp repression, bmps and admp are also coexpressed in central regions of the future organizer. According to our model, this should cause the complete loss of organizer identity, as indeed observed in severe cases of boz mutant phenotypes (Fekany et al., 1999; Koos and Ho, 1999). Consistent with this notion, a loss of the organizer, followed by a loss of head and notochord, can also be obtained upon ectopic early expression of bmps on the dorsal side of wild-type zebrafish embryos (Nikaido et al., 1997; Hammerschmidt, M., unpublished data). A similar interaction of Admp and Bmps could also account for a spatial and temporal restriction of organizer activities during later stages of development. Starting shortly after the onset of gastrulation, bmps are also expressed dorsally, so that at midgastrula stages, bmp4 and bmp7 are coexpressed with admp in anterior regions of the dorsal axis, whereas bmp2b and admp are coexpressed at its posterior end (Dick et al., 2000). In addition, some of the putative Bmp receptors (Hammerschmidt, M., unpublished data) and the putative downstream transcription factor Smad1 (Dick et al., 1999) are present in the dorsal midline of gastrulating embryos. In the future, several experiments have to be carried out to test this model. First, it has to be studied whether Admp and Bmps do indeed act in a synergistic manner. Furthermore, a detailed comparative analysis of the expression patterns of admp, the various bmps, and active organizer factors has to be performed, followed by specific ectopic expression studies of bmps in different zones of the organizer. Finally, it will be interesting to investigate how the initiation of admp expression and its progressive loss in anterior regions of the axis is regulated. These studies should help to resolve the seeming paradox why ventralizing agents are present and active within the dorsal axis and the organizer. EXPERIMENTAL PROCEDURES Cloning of Zebrafish admp A BLAST search of the TIGR zebrafish gene index revealed an EST (ID number fe05b04.y1; GeneBank accession no. AW174751) with high similarity to the 5 region of Xenopus and chick admp. By using this sequence information, we designed a nondegenerated sense primer covering the presumptive admp start codon (GCAATGTTCTTTGCAATGTTGTCCAC), which was used to amplify the entire admp coding region by means of 3 -RACE PCR with gastrula-stage RNA and a 3 -RACE kit according to manufacturer s (Clontech) protocols. The amplified fragment was cloned into pc- RII vector (Invitrogen) and sequenced, by using the Big Dye cycle sequencing protocol and ABI Prism 310 automated sequencer.

7 Admp DURING ZEBRAFISH ORGANIZER FORMATION 687 In Situ Hybridization Probe synthesis and in situ hybridizations were carried out as previously described (Hammerschmidt et al., 1996). For admp probe synthesis, plasmid pcriiadmp was linearized with NotI and transcribed with SP6 RNA polymerase. Sense RNA and Morpholino Injections The full-length admp cdna was cloned from pcrii into pcs2 by means of high-fidelity PCR introducing EcoRI - XbaI restriction sites. For sense RNA synthesis, pcs2-admp plasmid was linearized with NotI and transcribed by using the SP6 Message Machine Kit (Ambion) according to manufacturer s protocols. The following morpholino oligonucleotides were purchased from Gene-Tools and injected into one- to fourcell stage embryos at a concentration of 66 M or133 M in 1 Danieau buffer (Nasevicius and Ekker, 2000). (1) admp MO: 5 -TGGACAACATTGtAAAGAA- CATTGC-3 (complementary to 3 to 22 of cdna; contains one mismatch at nucleotide indicated by small letter to avoid self-hybridization); (2) admp 4-mismatch (4 mm) control MO: 5 -TGGAgAACtTTGtAAAGtACAaTGC-3 ; (3) admp sense control MO: 5 - GCAATGTTCTTTGCAATGTTGTC-3. Increasing the concentration or volume of injected admp MO led to a corresponding increase in the strength of the observed phenotype, whereas the same amounts of control morpholino oligonucleotides had no effects. ACKNOWLEDGMENTS We thank W. Driever for boz m168 ; S. Wilson and M. Concha for mzoep embryos; and J.-S. Joly (eve1, cdx1), R. Moon (wnt8), and W. Talbot (flh) for in situ probes. Z.L. was supported by an EMBO Long-Term Postdoctoral Fellowship. REFERENCES Dick A, Meier A, Hammerschmidt M Smad1 and Smad5 have distinct roles during dorsoventral patterning of the zebrafish embryo. Dev Dyn 216: Dick A, Hild M, Bauer H, Imai Y, Maifeld H, Schier A, Talbot W, Bowemeester T, Hammerschmidt M Essential role of Bmp7 (snailhouse) and its prodomain in dorsoventral patterning of the zebrafish embryo. Development 127: Dosch R, Niehrs C Requirement for Anti-dorsalizing morphogenetic protein in organizer patterning. Mech. Dev. 90: Dosch R, Gawantka V, Delius H, Blumenstock C, Niehrs C Bmp4 acts as a morphogen in dorsoventral patterning in Xenopus. Development 124: Fekany K, Yamanaka Y, Leung T, Sirotkin HI, Topczewski J, Gates MA, Hibi M, Renucci A, Stemple D, Radbill A, et al The zebrafish bozozok locus encodes Dharma, a homeodomain protein essential for induction of gastrula organizer and dorsoanterior embryonic structures. Development 126: Fekany-Lee K, Gonzalez E, Miller-Bertoglio V, Solnica-Krezel L The homeobox gene bozozok promotes anterior neuroectoderm formation in zebrafish through negative regulation of BMP2/4 and Wnt pathways. Development 127: Feldman B, Gates MA, Egan ES, Dougan ST, Rennebeck G, Sirotkin HI, Schier AF, Talbot WS Zebrafish organizer development and germ-layer formation require nodal-related signals. Nature 395: Gritsman K, Zhang J, Gheng S, Heckscher E, Talbot WS, Schier AF The EGF-CFC protein one-eyed pinhead is essential for nodal signaling. Cell 97: Hammerschmidt M, Pelegri F, Mullins MC, Kane DA, van Eeden FJM, Granato M, Brand M, Furutani-Seiki M, Haffter P, Heisenberg C-P, et al dino and mercedes, two genes regulating dorsal development in the zebrafish embryo. Development 123: Harland R, Gerhart J Formation and function of Spemann s organizer. Annu Rev Cell Dev Biol 13: Hashimoto H, Itoh M, Yamanaka Y, Yamashita S, Shimizu T, Solnicakrezel L, Hibi M, Hirano T Zebrafish Dkk1 functions in forebrain specification and axial mesendoderm formation. Dev Biol 217: Joly J-S, Maury M, Joly C, Duprey P, Boulkebache H, Condamine H Expression of a zebrafish caudal homeobox gene correlates with the establishment of posterior lineages at gastrulation. Differentiation 50: Joly J-S, Joly C, Schulte-Merker S, Boulkebache H., Condamine H The ventral and posterior expression of the homeobox gene eve1 is perturbed in dorsalized and mutant embryos. Development 119: Joubin K, Stern CD Molecular interactions continuously define the organizer during the cell movements of gastrulation. Cell 98: Kelly GM, Greenstein P, Erezyilmaz DF, Moon RT Zebrafish wnt8 and wnt8b share a common activity but are involved in distinct developmental pathways. Development 121: Koos DS, Ho RK The nieuwkoid gene characterizes and mediates a Nieuwkoop center-like activity in the zebrafish. Curr Biol 8: Koos DS, Ho RK The nieuwkoid/dharma homeobox gene is essential for bmp2b repression in the zebrafish pregastrula. Dev Biol 215: Miller-Bertoglio VE, Fisher S, Sánchez A, Mullins MC, Halpern ME Differential regulation of chordin expression domains in mutant zebrafish. Dev Biol 192: Moos M Jr, Wang S, Krinks M Anit-dorsalizing protein is a novel Tgf- homologue expressed in the Spemann organizer. Development 121: Mullins MC Embryonic axis formation in the zebrafish. Methods Cell Biol 59: Nasevicius A, Ekker SC Effective targeted gene knockdown in zebrafish. Nat Genet 26: Nikaido M, Tada M, Saji T, Ueno N Conservation of BMP signaling in zebrafish mesoderm patterning. Mech Dev 61: Piccolo SY, Sasai Y, Lu B, De Robertis EM A possible molecular mechanism for Spemann organizer function: inhibition of ventral signals by direct binding of Chordin to BMP-4. Cell 85: Piccolo S, Agius E, Leyns L, Bhattacharyya S, Grunz H, Bouwmeester T, DeRobertis EM The head inducer Cerberus is a multifunctional antagonist of Nodal, BMP, and Wnt signals. Nature 397: Ryu S-L, Fujii R, Yamanaka Y, Shimizu T, Yabe T, Hirata T, Hibi M, Hirano T Regulation of dharma/bozozok by the Wnt pathway. Dev Biol 231: Schulte-Merker S, Lee LJ, McMahon AP, Hammerschmidt M The zebrafish organizer requires chordino. Nature 387: Shinya M, Eschbach C, Clark M, Lehrach H, Furutani-Seiki M Zabrafish Dkk1, induced by the pre-mbt wnt signaling, is secreted from the prechordal plate and patterns the anterior neural plate. Mech Dev 98:3 17. Stachel SE, Grunwald DJ, Myers PZ Lithium perturbation and goosecoid expression identify a dorsal specification pathway in pregastrula zebrafish. Development 117: Talbot WS, Trevarrow B, Halpern ME, Melby AE, Farr G, Postlethwait JH, Jowett T, Kimmel CB, Kimelman D A homeobox gene essential for zebrafish notochord development. Nature 378: Yamanaka Y, Mizuno T, Sasai Y, Kishi M, Takeda H, Kim C-H, Hibi M, Hirano T A novel homeobox gene, dharma, can induce the organizer in a non-cell-autonomous manner. Genes Dev 12: Zhang J, Talbot WS, Schier AF Positional cloning identifies zebrafish one-eyed-pinhead as a permissive EGF-related ligand required during gastrulation. Cell 92: