We report an analysis of forebrain determination and patterning in the zebrafish Danio rerio. In order to study these events, we isolated zebrafish homologs of two neural markers, odd-paired-like (opl), which encodes a zinc finger protein, and fkh5, which encodes a forkhead domain protein. At mid-gastrula, expression of these genes defines a very early pattern in the presumptive neurectoderm, with opl later expressed in the telencephalon, and fkh5 in the diencephalon and more posterior neurectoderm. Using in vitro explant assays, we show that forebrain induction has occurred even earlier, by the onset of gastrulation (shield stage). Signaling from the early gastrula shield, previously shown to be an organizing center, is sufficient for activation of opl expression in vitro. In order to determine whether the organizer is required for opl regulation, we removed from late blastula stage embryos either the presumptive prechordal plate, marked by goosecoid (gsc) expression, or the entire organizer, marked by chordin (chd) expression. opl was correctly expressed after removal of the presumptive prechordal plate and consistently, opl was correctly expressed in one-eyed pinhead (oep) mutant embryos, where the prechordal plate fails to form. However, after removal of the entire organizer, no opl expression was observed, indicating that this region is crucial for forebrain induction. We further show that continued organizer function is required for forebrain induction, since beads of BMP4, which promotes ventral fates, also prevented opl expression when implanted during gastrulation. Our data show that forebrain specification begins early during gastrulation, and that a wide area of dorsal mesendoderm is required for its patterning.

Explant and transplant studies in amphibian, mouse and chick embryos indicate that neural determination has commenced by mid-gastrula stages (Ang et al., 1994; Holtfreter and Hamburger, 1955; Jones and Woodland, 1989; Roberts et al., 1991; Sive et al., 1989). Analysis using molecular markers is making it clear that, even before mid-gastrula, the presumptive neurectoderm has begun to be set aside (for example, see Kuo et al., 1998; Nakata et al., 1997). It is not clear, however, when the neurectoderm becomes patterned. During gastrulation, tissue initially specified as anterior neural (forebrain) includes at least some of the prospective posterior neurectoderm, suggesting that commitment to posterior neural fates may occur by converting anterior to more posterior tissue (Eyal-Giladi, 1954; Sive et al., 1989; reviewed in Kolm and Sive, 1997).

Despite the prevailing idea that forebrain is determined prior to more posterior neurectoderm, little information exists regarding the earliest patterning decisions within the presumptive forebrain. During neural plate stages, patterned gene expression divides the forebrain primordium into subregions (Bally-Cuif and Boncinelli, 1997; Simeone et al., 1993), while after neural tube closure, constrictions along the neural tube morphologically subdivide the future forebrain into prosomeres. Prosomeres can be grouped into two subregions: the diencephalon posteriorly, and more anteriorly, the secondary prosencephalon including the telencephalon (Puelles and Rubenstein, 1993; Rubenstein et al., 1994). However, it is not known when forebrain patterning begins, what genes divide the forebrain field initially, prior to neural plate formation, or what sources of patterning signals are responsible for this division.

Forebrain tissue can be induced by dorsal mesendoderm, also called the ‘organizer’, that has been found in all vertebrates. Organizing centers include the amphibian organizer (Spemann and Mangold, 1964), chick Hensen’s node (Storey et al., 1992; Waddington, 1932), the mouse node (Beddington, 1994) and the teleost shield (Oppenheimer, 1936; Sagerström et al., 1996; Shih and Fraser, 1996). The organizer can be separated into distinct domains which induce primarily anterior or primarily posterior neural tissue in Xenopus (Gerhart, 1996; Mangold and Spemann, 1927; Zoltewicz and Gerhart, 1997) and in teleosts other then zebrafish (Eakin, 1939; Oppenheimer, 1959). In Xenopus, anterior endoderm is a source of anterior ectodermal inducing signals (Bouwmeester et al., 1996; Bradley et al., 1996) and may be equivalent to the extraembryonic visceral endoderm in mouse that is required for forebrain formation (Thomas and Beddington, 1996; Varlet et al., 1997). In zebrafish, a recent report indicates that the anterior-most row of cells within the forming neural plate contains forebrain patterning signals (Houart et al., 1998).

Several molecules that are able to induce anterior neural tissue in Xenopus have been described (Kolm and Sive, 1997; Sasai and De Robertis, 1997). These include chordin, noggin and follistatin, which act by antagonizing the bone morphogenetic protein (BMP) signaling pathway through direct binding of BMPs (Piccolo et al., 1996; Zimmerman et al., 1996). Other inducers of anterior neural tissue are cerberus (Bouwmeester et al., 1996; Glinka et al., 1997), and dickkopf-1 (Glinka et al., 1998). Both are thought to antagonize the wnt signaling pathway, which is known to posteriorize neural tissue (McGrew et al., 1997). Corroborative data for the role of BMP-antiBMP signaling pathways have been obtained in zebrafish (Fisher et al., 1997; Hammerschmidt et al., 1996; Kishimoto et al., 1997; Schulte-Merker et al., 1997).

We have been studying neural commitment in the zebrafish, an excellent vertebrate paradigm (Driever et al., 1994; Kimmel, 1989; Rossant and Hopkins, 1992), where brain morphogenesis and patterning have been well described (Kimmel, 1993; Macdonald et al., 1994). Using explant assays we showed that zebrafish neurectoderm was not specified by late blastula; however, by early gastrula, the embryonic shield, which is a mixture of ectodermal and mesendodermal cell types, is committed to express both anterior and posterior neural markers (Sagerström et al., 1996). Since the shield is also a source of inducing signals (Sagerström et al., 1996; Shih and Fraser, 1996), it was not clear whether this reflected commitment of the early gastrula neurectoderm or induction of ectoderm during the culture period. Elegant transplant studies have shown that presumptive forebrain is not yet irreversibly committed by early gastrula (Woo and Fraser, 1997), but have not addressed whether initial forebrain induction has taken place at this stage.

In order to study determination and patterning of the zebrafish forebrain, we isolated zebrafish homologs of opl and fkh5, two genes defined as early neural-specific markers by subtractive cloning from Xenopus (Patel and Sive, 1996). We show here that zebrafish opl is expressed in the anterior neural plate which includes the presumptive telencephalon, and fkh5 in the presumptive diencephalon. Using in vitro explant assays (Grinblat et al., 1998; Sagerström et al., 1996) we analysed the timing of forebrain determination and the source of signals that pattern this region. We found that forebrain patterning begins early during gastrulation, and that the entire organizer, but not the prechordal plate, is necessary for forebrain determination in zebrafish.

Isolation and characterization of opl and fkh5 cDNA clones

A cDNA library, prepared from early neurula (9-16 hpf) embryos by R. Riggleman and D. Grunwald, was screened at low stringency (50°C; Church and Gilbert, 1984) with partial probes from the Xenopus opl and fkh5 cDNA. Complete nucleotide sequence was obtained for 3 zebrafish opl cDNAs (insert sizes 2.33, 2.33 and 1.86 kb) and 2 zebrafish fkh5 cDNAs (insert sizes 2.13 and 2.0 kb). Both contained the open reading frame predicted from other species, and were flanked by untranslated regions (for opl: 241 nt upstream and 693 nt downstream; for fkh5: 235 nt upstream and 1,010 nt downstream). Accession numbers are AF052435 for opl and AF052651 for fkh5.

Microdissections and explant culture

Explants were microdissected from embryos produced by natural crosses (Westerfield, 1995) and cultured at 29°C in 1× MBS supplemented with 50 μg/ml gentomycin and chemically defined lipids (0.01×, Gibco-BRL). Details of the dissection were as previously described (Grinblat et al., 1998; Sagerström et al., 1996). Late blastula (sphere stage) animal cap explants were cultured in groups of 10, early gastrula (shield stage) epiblast explants were cultured in groups of 5, and shield explants were cultured singly. Shield explants for conjugates were dissected from embryos lineage labeled with lysine fixable fluorescein dextran (FLDX, 10,000 Mr; Molecular Probes). Each FLDX-labeled shield was conjugated to a group of 5 unlabeled animal caps within 30 minutes after dissection.

Microdissections to remove dorsal margin

The dorsal margin was identified by rotating embryos at 30-35% epiboly on their sides in the dissection dish until the blastoderm margin appeared clearly thinner on one side than on the opposite side (Schmitz and Campos-Ortega, 1994). With the embryo held in this orientation, a cut was made to mark the location. The embryo was then rotated and additional cuts were made to remove the regions shown in Fig. 7A. One of these cuts punctured the yolk cell and released some of the dorsally located yolk. Mock dissected control embryos were generated by making incisions on three sides, including the incision which punctured the yolk. Equivalent amounts of dorsal yolk were released in mock-dissected embryos. Operated embryos were cultured until unoperated controls reached 90% epiboly stage.

Bead implantation

Coated Sepharose beads (Biorad Affi-gel blue, 75-150 μm diameter) were prepared as follows: after rinsing in PBS, 1 μl of the beads were incubated with 5 μl of a solution of BMP4 (5 mg/ml in 0.1% trifluoroacetic acid), or BSA (10 mg/ml in water) for 1 hour at room temperature. Just prior to implantation, the beads were transferred to a dissection plate filled with culture medium. An incision was made in the dorsal epiblast of a shield stage embryo, leaving the yolk cell intact, and a bead was pushed through the incision toward the animal pole, between the epiblast and the yolk cell. Incisions healed well and epiboly was not significantly affected.

Relative quantitative RT-PCR analysis

RNA was isolated and reverse transcribed as previously described by Sagerström et al. (1996). PCR amplification and analysis were carried out as described by Sagerström et al. (1996) for the number of cycles empirically defined as being within the linear range of amplification for individual primer pairs. After electrophoresis, a PhosphorImager gel reader and software (Fuji) was used to quantitate PCR products. PCR primers used to amplify gene-specific products are listed in Sagerström et al. (1996) and Table 1.

Table 1.

PCR primers used to amplify gene-specific products

PCR primers used to amplify gene-specific products
PCR primers used to amplify gene-specific products

Whole-mount in situ hybridization

In situ hybridization of whole embryos was carried out according to previously published protocol (Sagerström et al., 1996). The orange precipitate was formed when BCIP (Sigma) and 4-iodonitrotetrazolium violet (INT) (Molecular Probes) were added together at the developing step. The purple precipitate resulted from a combination of NBT (Research Organics) and BCIP.

Zebrafish opl is an early anterior neural marker

In a recent subtractive cloning screen for early neural markers in Xenopus (Kuo et al., 1998; Patel and Sive, 1996), one of the earliest neural-specific genes identified was a gene encoding a zinc finger protein, opl, so called because it resembles Drosophila odd-paired (odd-paired-like). Since few ectoderm-restricted markers of early neural pattern were available in zebrafish, we isolated the zebrafish homolog of opl. Zebrafish opl encodes a 441 amino acid protein with a zinc finger domain which is highly similar (77%) to that of the Drosophila gene odd paired (opa) (Benedyk, 1994) (Fig. 1). Within vertebrates opl is highly conserved along the entire length of the protein. Zebrafish opl protein is 89% identical both to Xenopus opl (Kuo et al., 1998) and to its probable mouse homolog, Zic1 (Aruga et al., 1994).

Fig. 1.

Zebrafish opl is highly conserved. Alignment of the deduced amino acid sequences of the zebrafish opl protein and those of its putative homologs: Xenopus opl (Kuo et al., 1998), mouse Zic1 (Aruga et al., 1994) and Drosophila opa (Benedyk, 1994). Positions of identity are shaded. Amino acid positions are designated on the left hand side of the sequences. The five zinc finger domains found within the proteins are indicated with a line above.

Fig. 1.

Zebrafish opl is highly conserved. Alignment of the deduced amino acid sequences of the zebrafish opl protein and those of its putative homologs: Xenopus opl (Kuo et al., 1998), mouse Zic1 (Aruga et al., 1994) and Drosophila opa (Benedyk, 1994). Positions of identity are shaded. Amino acid positions are designated on the left hand side of the sequences. The five zinc finger domains found within the proteins are indicated with a line above.

Expression of zebrafish opl began at mid-gastrula, as seen both by northern analysis (data not shown) and by whole-mount in situ hybridization (Fig. 2A). Strong expression was restricted to an arc at the anterior of the future neural plate. Much weaker expression outlining the lateral edges of the future neural plate was detectable more posteriorly upon overstaining (not shown). Expression persisted in this domain until late gastrula (Fig. 2B), when the opl expression domain directly abutted but did not overlap ventral ectoderm, as seen in embryos double stained for opl (orange) and gta3 (purple), a marker of ventral ectoderm (Neave et al., 1995) (Fig. 2C). After completion of gastrulation at tailbud stage, strong opl expression persisted in the presumptive forebrain (Fig. 2D). As somitogenesis progressed, opl expression persisted in this domain, with expression in the more posterior neural keel upregulated (arrowhead in Fig. 2E). By the end of somitogenesis, in prim-5 stage embryos, opl was strongly expressed throughout the brain and more weakly in the spinal cord (Fig. 2F). Sections of late gastrula (90% epiboly) stage embryos show that opl expression was restricted to the epiblast (ectoderm) (Fig. 2G). Sections of prim-5 stage embryos, which have formed the neural tube, show that opl expression was restricted to the dorsal neural tube in both diencephalon (Fig. 2H) and rhombencephalon (Fig. 2I).

Fig. 2.

Zebrafish opl is expressed in the anterior neural plate and dorsal neural tube. Staged embryos were stained for opl RNA (purple, except in C, where it is orange) and gta3 RNA (purple in C) using whole-mount in situ hybridization. (A) Mid-gastrula (75% epiboly) embryo, dorsal view, dotted line indicates blastoderm margin, dashed line indicates anterior neural boundary. (B) Late gastrula (90% epiboly), dorsal view; dotted line indicates blastoderm margin, arrow indicates plane of section shown in G. (C) Late gastrula (90% epiboly), anterior view. (D) Early neurula (tailbud) embryo, dorsal view, dashed line outlines the neural plate. (E) Early somitogenesis (5 somites), side view, arrowhead indicates a weak posterior domain of staining. (F) Late somitogenesis (prim-5), side view, arrows indicate planes of section shown in H and I. (G) Transverse section through a late gastrula (90% epiboly) stage embryo as shown in D, showing that opl expression is confined to the epiblast (e) and is excluded from the hypoblast (h) (mesendoderm) layers (bracketed). Arrowheads mark dorsal and ventral limits of the blastoderm. (H,I) Transverse section through (H) the diencephalon, (I) the rhombencephalon; note restriction of expression to dorsal neural tube, arrowheads mark dorsal and ventral limits of the blastoderm. a, anterior; d, dorsal; v, ventral; p, posterior; t, telencephalon; d, diencephalon; m, mesencephalon; r, rhombencephalon; s, spinal cord.

Fig. 2.

Zebrafish opl is expressed in the anterior neural plate and dorsal neural tube. Staged embryos were stained for opl RNA (purple, except in C, where it is orange) and gta3 RNA (purple in C) using whole-mount in situ hybridization. (A) Mid-gastrula (75% epiboly) embryo, dorsal view, dotted line indicates blastoderm margin, dashed line indicates anterior neural boundary. (B) Late gastrula (90% epiboly), dorsal view; dotted line indicates blastoderm margin, arrow indicates plane of section shown in G. (C) Late gastrula (90% epiboly), anterior view. (D) Early neurula (tailbud) embryo, dorsal view, dashed line outlines the neural plate. (E) Early somitogenesis (5 somites), side view, arrowhead indicates a weak posterior domain of staining. (F) Late somitogenesis (prim-5), side view, arrows indicate planes of section shown in H and I. (G) Transverse section through a late gastrula (90% epiboly) stage embryo as shown in D, showing that opl expression is confined to the epiblast (e) and is excluded from the hypoblast (h) (mesendoderm) layers (bracketed). Arrowheads mark dorsal and ventral limits of the blastoderm. (H,I) Transverse section through (H) the diencephalon, (I) the rhombencephalon; note restriction of expression to dorsal neural tube, arrowheads mark dorsal and ventral limits of the blastoderm. a, anterior; d, dorsal; v, ventral; p, posterior; t, telencephalon; d, diencephalon; m, mesencephalon; r, rhombencephalon; s, spinal cord.

These data identify opl as a very early neural marker, initially specific to the anterior neurectoderm, and later restricted to the dorsal region of the entire brain.

opl and fkh5 divide the future forebrain into anterior and posterior domains

We isolated the zebrafish homolog of another gene defined in the Xenopus subtraction study, the forkhead domain gene fkh5 (Gamse and Sive, unpublished data), which has also been isolated in mouse (Kaestner et al., 1996; Labosky et al., 1997; Wehr et al., 1997). We used in situ hybridization to compare the expression pattern of fkh5 RNA with that of opl. Zebrafish fkh5 (purple stain) was first expressed in mid-gastrula embryos in a wide stripe in the posterior dorsal ectoderm (Fig. 3A). At this stage, there was a slight gap between the posterior border of the opl (orange stain) expression domain and the anterior border of the fkh5 expression. By late gastrula (Fig. 3B) fkh5 was strongly expressed in an anterior domain which lay immediately posterior to, but did not overlap with, the opl-expressing territory, and was more weakly expressed in a separate posterior domain. Posteriorly, the strong anterior domain of fkh5 expression overlapped with the expression domain of pax2, a mesencephalic marker (Krauss et al., 1991) (data not shown). Both the anterior and posterior expression domains of fkh5 were maintained through neural keel formation which begins at tailbud stage (Fig. 3C). The anterior edge of the fkh5 domain remained directly juxtaposed with the posterior edge of the opl domain. A transverse section through the anterior expression domain at this stage (Fig. 3D) showed that fkh5 expression was restricted to the epiblast (ectoderm).

Fig. 3.

opl and fkh5 are expressed in adjacent domains that mark presumptive telencephalon and diencephalon at gastrula. Staged embryos were stained for opl RNA (orange) and fkh5 RNA (purple) using whole-mount in situ hybridization. (A) Mid-gastrula (75% epiboly) embryo, dorsal view, dotted line marks blastoderm margin; dashed line outlines anterior neural plate. (B) Late gastrula (90% epiboly) embryo, dorsal view; ad, anterior domain of fkh5 expression; pd, posterior domain of fkh5 expression. (C) Early neurula (tailbud) embryo, dorsal view, arrow shows plane of section shown in D. (D) Transverse section through a tailbud stage embryo. Note that staining is strong in the epiblast (e) and excluded from hypoblast (h). (E) Early somitogenesis (5 somite) embryo, side view.(F) Late somitogenesis (prim-5) embryo, side view, asterisks indicate weak fkh5 staining. a, anterior; t, telencephalon; d, diencephalon; m, mesencephalon; p, posterior; r, rhombencephalon; s, spinal cord.

Fig. 3.

opl and fkh5 are expressed in adjacent domains that mark presumptive telencephalon and diencephalon at gastrula. Staged embryos were stained for opl RNA (orange) and fkh5 RNA (purple) using whole-mount in situ hybridization. (A) Mid-gastrula (75% epiboly) embryo, dorsal view, dotted line marks blastoderm margin; dashed line outlines anterior neural plate. (B) Late gastrula (90% epiboly) embryo, dorsal view; ad, anterior domain of fkh5 expression; pd, posterior domain of fkh5 expression. (C) Early neurula (tailbud) embryo, dorsal view, arrow shows plane of section shown in D. (D) Transverse section through a tailbud stage embryo. Note that staining is strong in the epiblast (e) and excluded from hypoblast (h). (E) Early somitogenesis (5 somite) embryo, side view.(F) Late somitogenesis (prim-5) embryo, side view, asterisks indicate weak fkh5 staining. a, anterior; t, telencephalon; d, diencephalon; m, mesencephalon; p, posterior; r, rhombencephalon; s, spinal cord.

The domains of opl and anterior fkh5 expression in the late gastrula embryo were remarkably similar in shape and position to the presumptive telencephalon and diencephalon, respectively, as defined by fate mapping (Woo and Fraser, 1995). Interestingly, a recent fine scale fate mapping of the late gastrula neural plate indicates that opl is expressed primarily in the future retina, which is initially adjacent to the presumptive telencephalon (Varga and Westerfield, personal communication). Later, opl is not expressed in the developing eye but is expressed in the telencephalon. The close proximity between the future retinal and telencephalic fields makes these domains indistinguishable in our assays, and we will therefore refer to opl as a telencephalic marker.

Strong fkh5 expression persisted through early somitogenesis, when it was restricted to the newly formed ventral diencephalon, mesencephalon, and spinal cord (Fig. 3E). By the end of somitogenesis, anterior fkh5 expression comprised two weak vertical stripes in the brain, one in the diencephalon and one in the posterior mesencephalon (Fig. 3F). The posterior domain was also downregulated, except for the posterior spinal cord and tailbud.

In summary, these data showed that opl and fkh5 expression divides the presumptive neurectoderm of the gastrula embryo into two distinct domains: the anterior domain, marked by opl, which includes the prospective telencephalon, and the posterior domain, marked by fkh5, which includes the prospective diencephalon, mesencephalon, and posterior hindbrain/spinal cord.

Neural specification takes place by early gastrula (shield) stage

Using explant assays and neural markers, which are expressed starting at late gastrula, we previously established that neural specification has not yet occurred at late blastula (sphere stage, 4 hpf) (Sagerström et al., 1996). We wished to repeat this analysis using the earlier, mid-gastrula onset, neural markers opl and fkh5. We also wanted to ask whether early gastrula (shield stage, 6 hpf) caps were specified as neural tissue. Explants of animal caps from sphere stage embryos or presumptive anterior neurectoderm from shield stage embryos were cultured in saline, in groups of 5 to 10, until control embryos reached mid-somitogenesis (10 somite stage, 14 hpf). Explants were then analyzed by a reverse transcriptase-PCR (RT-PCR) assay for expression of a panel of ectodermal markers (see Materials and Methods).

The results shown in Fig. 4B are representative of 2-5 experiments for each marker. Late blastula animal caps (lane 1) were strongly specified to express the ventral ectodermal marker gta3 (Neave et al., 1995), and weakly specified to express early markers of anterior dorsal ectoderm opl, fkh5 and otx2 (Li et al., 1994). These caps were not specified to express the neural marker pax6 (Krauss et al., 1991) that is activated later during development, at late gastrula, in accord with our previous results (Sagerström et al., 1996). By early gastrula (shield stage) (lane 2), animal caps were much more weakly specified to express gta3 (by 3 to 35 fold, 5 fold on average, relative to sphere stage caps). In contrast, caps were much more strongly specified to express opl and otx2 at this stage than at earlier stages, with levels of opl upregulated by 3 to 30 fold (an average of 6 fold), and levels of otx2 by 5 to 8 fold (an average of 6 fold) relative to sphere stage caps. This strong specification occurred 2 hours prior to the normal onset of expression of these genes at mid-gastrula. Specification of the ability to express fkh5 did not increase significantly in shield stage caps relative to sphere stage caps. Specification of shield caps for expression of pax6, a neural marker with a later onset of expression than opl or fkh5, was observed for the first time at this stage.

Fig. 4.

Explant assays show that neural specification occurs by early gastrula. (A) Schematic outline of the explant assay. Animal caps were explanted from sphere stage, 30% epiboly stage, or shield stage embryos. Caps were combined in groups of 5-10 and cultured until control embryos reached mid-somitogenesis, when they were assayed by a reverse transcriptase-PCR (RT-PCR) assay for expression of marker genes, as described in Materials and Methods. (B) RT-PCR analysis of gene expression in animal cap explants. The following markers were assayed: gta3 (Neave et al., 1995), a marker of ventral ectoderm; opl, a marker of prospective telencephalon; fkh5, a marker of prospective diencephalon, mesencephalon and spinal cord; otx2 (Li et al., 1994), a marker of prospective telencephalon, diencephalon, and mesencephalon; pax6 (Krauss et al., 1991), a marker of prospective diencephalon, hindbrain and spinal cord, and α-tubulin (G. Conway, personal communication), a loading control. Each lane represents a pool of five explants. The data shown here is representative of 2 experiments each for gta3, otx2 and pax6, 3 experiments for fkh5 and 5 experiments for opl. Lane 1: explants from sphere stage embryos; lane 2: explants from shield stage embryos; lane 3: whole embryo controls. (C) RT-PCR analysis of opl expression in animal cap explants from a later blastula stage. Dissections were done as in A using embryos at 30-35% epiboly. Lane 1: explants from sphere stage embryos; lane 2: explants from 30-35% epiboly stage embryos; lane 3: explants from shield stage embryos.

Fig. 4.

Explant assays show that neural specification occurs by early gastrula. (A) Schematic outline of the explant assay. Animal caps were explanted from sphere stage, 30% epiboly stage, or shield stage embryos. Caps were combined in groups of 5-10 and cultured until control embryos reached mid-somitogenesis, when they were assayed by a reverse transcriptase-PCR (RT-PCR) assay for expression of marker genes, as described in Materials and Methods. (B) RT-PCR analysis of gene expression in animal cap explants. The following markers were assayed: gta3 (Neave et al., 1995), a marker of ventral ectoderm; opl, a marker of prospective telencephalon; fkh5, a marker of prospective diencephalon, mesencephalon and spinal cord; otx2 (Li et al., 1994), a marker of prospective telencephalon, diencephalon, and mesencephalon; pax6 (Krauss et al., 1991), a marker of prospective diencephalon, hindbrain and spinal cord, and α-tubulin (G. Conway, personal communication), a loading control. Each lane represents a pool of five explants. The data shown here is representative of 2 experiments each for gta3, otx2 and pax6, 3 experiments for fkh5 and 5 experiments for opl. Lane 1: explants from sphere stage embryos; lane 2: explants from shield stage embryos; lane 3: whole embryo controls. (C) RT-PCR analysis of opl expression in animal cap explants from a later blastula stage. Dissections were done as in A using embryos at 30-35% epiboly. Lane 1: explants from sphere stage embryos; lane 2: explants from 30-35% epiboly stage embryos; lane 3: explants from shield stage embryos.

In order to narrow the window of time during which neural specification occurred even further, we compared specification for opl expression in animal caps removed from sphere stage (4 hpf), from 30-35% epiboly (very late blastula, 4 hours 45 minutes postfertilization) and from shield stage (6 hpf) embryos (Fig. 4A). As shown in Fig. 4C, opl expression was specified at equivalently low levels in caps taken from sphere stage (lane 1) and from 30-35% epiboly (lane 2) embryos, and increased to high levels in shield stage embryos (lane 3). In summary, we conclude that neural specification takes place during a span of about 1 hour 15 minutes between the 30-35% epiboly and shield stages. This includes approximately 55 minutes before and 20 minutes after the onset of gastrulation.

Dorsal ectoderm is specified as anterior neural and patterned by early gastrula

Having established that the presumptive neurectoderm of the shield stage embryo is specified as neural, we next asked whether any anteroposterior (A/P) pattern had been specified in this tissue. As diagrammed in Fig. 5A, four distinct regions of shield stage (6 hpf) ectoderm were explanted separately: ventral (V), anterior dorsal (A), posterior dorsal (P), and lateral (L), where V will give rise primarily to epidermis, A to telencephalon and retina with some diencephalon, P to diencephalon, retina and mesencephalon, and L to rhombencephalon. These assignments were made based on a fate map of Woo and Fraser (Woo and Fraser, 1995), constructed by dye-labeling cells at early gastrula (shield stage). The shield region (S), which contains a mixture of ectoderm, mesoderm and endoderm cells (Shih and Fraser, 1995), was also explanted. Explants were either harvested immediately, or cultured in groups of 5 until control embryos reached tailbud (10 hpf) or 10 somite (14 hpf) stage, when they were analysed for expression of marker genes using an RT-PCR assay (see Materials and Methods).

Fig. 5.

Forebrain pattern is specified at early gastrula. (A) Schematic outline of the explant assay. Four segments of early gastrula (shield stage) epiblast were isolated: ventral epiblast (V), anterior dorsal epiblast (A), posterior dorsal epiblast (P) and lateral epiblast (L). Explants of embryonic shield (S) contained a mixture of presumptive ectoderm, mesoderm and endoderm. All explants were combined in groups of 5 and harvested immediately (B) or cultured until control embryos reached tailbud or 10 somite (C) stage before analysis of marker gene expression by RT-PCR (see Methods). (B) RT-PCR analysis of explants harvested immediately after dissection and assayed for expression of dorsal mesendodermal markers gsc (Stachel et al., 1993; Thisse et al., 1994) and lim1 (Toyama et al., 1995), both markers of the axial hypoblast of the shield (prospective prechordal plate), ntl (Schulte-Merker et al., 1994), a marker of the entire mesendoderm, and chd (Miller-Bertoglio et al., 1997) a marker of the dorsal mesendodorm. α-tubulin (G. Conway, personal communication) was used as loading control. The data shown here is representative of 2 experiments. Lane 1: ventral explant; lane 2: anterior dorsal explant; lane 3: posterior dorsal explant; lane 4: shield explant. (C) RT-PCR analysis of cultured explants assayed for expression of ectodermal markers gta3 (Neave et al., 1995), a marker of ventral ectoderm, opl, a marker of prospective telencephalon, fkh5, a marker of prospective diencephalon, mesencephalon and spinal cord, krx20 (Oxtoby and Jowett, 1993), a marker of prospective rhombencephalon, and α-tubulin (G. Conway, personal comunication), a loading control. Each lane represents a pool of five explants. The data shown is representative of 3-4 independent experiments. Lane 1: ventral explants; lane 2: anterior dorsal explants; lane 3: posterior dorsal explants; lane 4: shield explants; lane 5: lateral ectoderm explants; lane 6: whole embryo control. NT: not tested.

Fig. 5.

Forebrain pattern is specified at early gastrula. (A) Schematic outline of the explant assay. Four segments of early gastrula (shield stage) epiblast were isolated: ventral epiblast (V), anterior dorsal epiblast (A), posterior dorsal epiblast (P) and lateral epiblast (L). Explants of embryonic shield (S) contained a mixture of presumptive ectoderm, mesoderm and endoderm. All explants were combined in groups of 5 and harvested immediately (B) or cultured until control embryos reached tailbud or 10 somite (C) stage before analysis of marker gene expression by RT-PCR (see Methods). (B) RT-PCR analysis of explants harvested immediately after dissection and assayed for expression of dorsal mesendodermal markers gsc (Stachel et al., 1993; Thisse et al., 1994) and lim1 (Toyama et al., 1995), both markers of the axial hypoblast of the shield (prospective prechordal plate), ntl (Schulte-Merker et al., 1994), a marker of the entire mesendoderm, and chd (Miller-Bertoglio et al., 1997) a marker of the dorsal mesendodorm. α-tubulin (G. Conway, personal communication) was used as loading control. The data shown here is representative of 2 experiments. Lane 1: ventral explant; lane 2: anterior dorsal explant; lane 3: posterior dorsal explant; lane 4: shield explant. (C) RT-PCR analysis of cultured explants assayed for expression of ectodermal markers gta3 (Neave et al., 1995), a marker of ventral ectoderm, opl, a marker of prospective telencephalon, fkh5, a marker of prospective diencephalon, mesencephalon and spinal cord, krx20 (Oxtoby and Jowett, 1993), a marker of prospective rhombencephalon, and α-tubulin (G. Conway, personal comunication), a loading control. Each lane represents a pool of five explants. The data shown is representative of 3-4 independent experiments. Lane 1: ventral explants; lane 2: anterior dorsal explants; lane 3: posterior dorsal explants; lane 4: shield explants; lane 5: lateral ectoderm explants; lane 6: whole embryo control. NT: not tested.

Since we wished to determine the specification state of pure ectoderm, we first asked whether explants were free of mesodermal contamination by harvesting explants immediately after dissection and analysing them for expression of mesodermal markers. A representative experiment (one of two with very similar results) is shown in Fig. 5B. The V explant (lane 1) was not specified to express the dorsal mesodermal markers gsc (Stachel et al., 1993; Thisse et al., 1994), lim1 (Toyama et al., 1995), or chd (Miller-Bertoglio et al., 1997), and was only weakly specified to express the pan mesodermal marker ntl (Schulte-Merker et al., 1994). The A explant did not express any of the 4 markers (lane 2), while very low (less then 1% of S expression) levels of gsc and chd were seen in the P explant (lane 3). Since both gsc and chd are expressed in the ectoderm soon after this stage, starting at mid to late gastrula (Miller-Bertoglio et al., 1997; Thisse et al., 1994), and since lim1 and ntl RNA was not detected in the P explant, the presence of small quantities of gsc and chd RNA in the P explant at early gastrula is likely to be due to low level ectodermal expression. In contrast, all four mesodermal markers were highly expressed in the S explant (lane 4), consistent with the fate map and our previous specification analysis (Sagerström et al., 1996).

In order to determine whether A/P polarity already exists in the dorsal ectoderm at shield stage, prior to the onset of opl or fkh5 expression, we cultured explants until tailbud or 10 somite stage and analyzed them for marker gene expression (Fig. 5A). Results of a representative experiment (of 4 which gave similar results) are shown in Fig. 5C. V explants (lane 1) were strongly specified to express gta3, but only weakly specified to express the dorsal markers opl and fkh5. In contrast, the A explant (lane 2) was weakly specified to express gta3 (with levels in the V explant 10 fold higher than those of the A explant) and strongly specified to express opl. The P explant (lane 3), like the A explant, was weakly specified to express gta3 and strongly specified to express opl, but in contrast to the A explant, it was also strongly specified for fkh5 expression (with levels of fkh5 15 fold higher in the cultured P than in the A explant). In addition, the P explant was less strongly specified to express opl than was the A explant (with levels approximately 1.5 fold lower in the P than in the A explant). The shield explant S (lane 4) was not specified to express opl or high levels of gta3, but was strongly specified to express fkh5. We also examined the specification state of the lateral ectoderm, L (lane 5) that is fated to form rhombencephalon. In order to assay expression of the rhombencephalic marker krx20 (Oxtoby and Jowett, 1993), whose expression is activated at early neurula and reaches high levels by mid-somitogenesis, explants were cultured until control embryos reached the 10 somite stage. L explants were specified to express intermediate levels of gta3 expression and strongly specified to express opl and fkh5, suggesting that this explant straddles the boundary between dorsal and ventral ectoderm. In contrast, krx20 expression was not detected in this explant or in the V, A, and P explants. We have previously shown that isolated shields (S explant) go on to activate high levels of krx20 (Sagerström et al., 1996).

In summary, this analysis showed that early during gastrulation the presumptive neurectoderm is not only specified, but contains a latent anteroposterior pattern. Specifically, telencephalic (marked by opl) and more posterior neural (marked by fkh5) regions are determined in discrete parts of the neurectoderm. Since a rhombencephalic marker (krx20) was not specified to be expressed at this stage, we assign the fkh5 expression to mark the presumptive diencephalon and mesencephalon.

The early gastrula shield can induce and pattern anterior neural tissue

Having shown that the neurectoderm near the animal pole is strongly specified as telencephalon by early gastrula, we asked what source of inducing signals might be responsible for this specification. By early gastrula the presumptive telencephalon is not yet underlain by involuting mesendoderm (Kimmel et al., 1990), suggesting either that signals arise from a source at a distance, particularly the dorsal mesendoderm (organizer), or that a proximal tissue other than the organizer is the source of inducing signals. We first asked whether the early gastrula shield, which contains organizer activity (Sagerström et al., 1996; Shih and Fraser, 1996), can induce opl during zebrafish embryogenesis. Animal caps were isolated from late blastula embryos and either cultured alone in groups of 10 or as conjugates with shields isolated from early gastrula (shield stage, 6 hpf) embryos (Fig. 6A). Shield explants cultured alone served as controls. Explants were cultured until control embryos from which the shields were taken reached the 5 somite stage, and assayed for opl expression by in situ hybridization. In order to distinguish inducing and responding tissues after culture, embryos from which the shield was removed for conjugation were lineage labeled with FITC-conjugated dextran (FLDX) that was later visualized by immunohistochemistry (see Materials and Methods).

Fig. 6.

Early gastrula organizer (shield) induces opl expression in animal caps. (A) Schematic outline of the conjugation experiment. Animal cap explants from late blastula (sphere stage) embryos were cultured either in isolation, in groups of 10, or as conjugates with shields explanted from early gastrula (shield stage) embryos. Each conjugate was made with one shield and 5 animal caps. As a control, shields were cultured alone. For conjugates, embryos from which shields were isolated were lineage labeled with FLDX (pale blue; see Methods) in order to distinguish inducing from responding tissues. Explants were cultured until shield-donor embryos reached the 5 somite stage, at which time the explants were harvested and stained for opl RNA (purple) by in situ hybridization. (B) Cultured explants, stained for the presence of opl RNA (purple) and lineage label (blue).(a) Two groups of 10 animal caps cultured in isolation. (b) Three conjugates representative of the typical outcomes of the experiment. Dots mark the anterior and posterior edges of the animal cap-derived tissues. (c) Shield explants cultured in isolation. A total of 38 conjugates were made in 7 independent experiments. Induction of opl was observed in 58% of them, and was always restricted to the animal cap-derived portion. No detectable opl expression has been observed in more than 40 shield explants generated in at least 5 experiments. Abbreviations: a, anterior; p, posterior.

Fig. 6.

Early gastrula organizer (shield) induces opl expression in animal caps. (A) Schematic outline of the conjugation experiment. Animal cap explants from late blastula (sphere stage) embryos were cultured either in isolation, in groups of 10, or as conjugates with shields explanted from early gastrula (shield stage) embryos. Each conjugate was made with one shield and 5 animal caps. As a control, shields were cultured alone. For conjugates, embryos from which shields were isolated were lineage labeled with FLDX (pale blue; see Methods) in order to distinguish inducing from responding tissues. Explants were cultured until shield-donor embryos reached the 5 somite stage, at which time the explants were harvested and stained for opl RNA (purple) by in situ hybridization. (B) Cultured explants, stained for the presence of opl RNA (purple) and lineage label (blue).(a) Two groups of 10 animal caps cultured in isolation. (b) Three conjugates representative of the typical outcomes of the experiment. Dots mark the anterior and posterior edges of the animal cap-derived tissues. (c) Shield explants cultured in isolation. A total of 38 conjugates were made in 7 independent experiments. Induction of opl was observed in 58% of them, and was always restricted to the animal cap-derived portion. No detectable opl expression has been observed in more than 40 shield explants generated in at least 5 experiments. Abbreviations: a, anterior; p, posterior.

Fig. 7.

Dorsal mesendoderm is required for the gastrula expression of opl. (A) Schematic outline of the extirpation experiment. The dorsal margin of a late blastula embryo, morphologically distinct from the ventral margin at 30% epiboly, was removed by microdissection. Three distinct regions were removed: the stippled region, corresponding approximately to the chd-expressing region, the hatched region, corresponding approximately to the gsc-expressing region, and the solid gray region, corresonding to the animal portion of the presumptive organizer region. Resulting embryos were cultured until control embryos reached 90% epiboly when they were stained by in situ hybridization (B,C) for genes listed below (see also Results and Materials and Methods). (B) (a,b) Control wild-type embryos; (c,e) control microsurgery; (d,f) removal of the presumptive shield/organizer region. (a) Dorsal view of a wild-type embryo at early gastrula (shield stage) stained for gsc (purple) and chd RNA (orange); (b) dorsal view of a wild-type embryo at late gastrula (90% epiboly) stained for chd (purple) and opl RNA (orange). (c) Embryo generated by removing the animal portion of the presumptive organizer region (solid gray in Fig. 7A), cultured until 90% epiboly stage equivalent and stained for opl (orange) and chd (purple). 5 out of 5 embryos obtained in 1 experiment expressed opl and chd. (d) Embryo generated by complete removal of the presumptive organizer region (stippled in A), cultured until 90% epiboly stage equivalent, and stained for opl (orange) and chd (purple). 8 out of 23 embryos generated in 4 experiments failed to express chd, and were also devoid of opl RNA. The remaining 15 out of 23 embryos expressed chd and opl, probably due to incomplete organizer removal. (e) Mock dissected embryos at 90% epiboly stage equivalent, stained for opl (orange) and gsc (purple). 11 out of 11 embryos generated in 4 experiments expressed both opl and gsc.(f) Embryo generated by removing part of the presumptive organizer region (hatched in A), cultured to 90% epiboly equivalent, and stained for opl (orange) and gsc (purple). 10 out of 17 embryos generated in 4 experiments failed to express gsc, but were able to express and pattern opl. In addition, 1 embryo was obtained which was free of both gsc and opl staining. (C) Analysis of one eyed pinhead mutant embryos. Embryos resulting from a cross between oepte57 heterozygotes (39 total) were analysed at tailbud stage by in situ hybridization for expression of opl (orange) and gsc (purple). (a) Wild-type embryo, representative of 30 out of 39 embryos; (b) sibling embryo homozygous for the oep mutant allele, identified by failure to express gsc (9 out of 39 embryos). opl was patterned but expressed at lower levels.

Fig. 7.

Dorsal mesendoderm is required for the gastrula expression of opl. (A) Schematic outline of the extirpation experiment. The dorsal margin of a late blastula embryo, morphologically distinct from the ventral margin at 30% epiboly, was removed by microdissection. Three distinct regions were removed: the stippled region, corresponding approximately to the chd-expressing region, the hatched region, corresponding approximately to the gsc-expressing region, and the solid gray region, corresonding to the animal portion of the presumptive organizer region. Resulting embryos were cultured until control embryos reached 90% epiboly when they were stained by in situ hybridization (B,C) for genes listed below (see also Results and Materials and Methods). (B) (a,b) Control wild-type embryos; (c,e) control microsurgery; (d,f) removal of the presumptive shield/organizer region. (a) Dorsal view of a wild-type embryo at early gastrula (shield stage) stained for gsc (purple) and chd RNA (orange); (b) dorsal view of a wild-type embryo at late gastrula (90% epiboly) stained for chd (purple) and opl RNA (orange). (c) Embryo generated by removing the animal portion of the presumptive organizer region (solid gray in Fig. 7A), cultured until 90% epiboly stage equivalent and stained for opl (orange) and chd (purple). 5 out of 5 embryos obtained in 1 experiment expressed opl and chd. (d) Embryo generated by complete removal of the presumptive organizer region (stippled in A), cultured until 90% epiboly stage equivalent, and stained for opl (orange) and chd (purple). 8 out of 23 embryos generated in 4 experiments failed to express chd, and were also devoid of opl RNA. The remaining 15 out of 23 embryos expressed chd and opl, probably due to incomplete organizer removal. (e) Mock dissected embryos at 90% epiboly stage equivalent, stained for opl (orange) and gsc (purple). 11 out of 11 embryos generated in 4 experiments expressed both opl and gsc.(f) Embryo generated by removing part of the presumptive organizer region (hatched in A), cultured to 90% epiboly equivalent, and stained for opl (orange) and gsc (purple). 10 out of 17 embryos generated in 4 experiments failed to express gsc, but were able to express and pattern opl. In addition, 1 embryo was obtained which was free of both gsc and opl staining. (C) Analysis of one eyed pinhead mutant embryos. Embryos resulting from a cross between oepte57 heterozygotes (39 total) were analysed at tailbud stage by in situ hybridization for expression of opl (orange) and gsc (purple). (a) Wild-type embryo, representative of 30 out of 39 embryos; (b) sibling embryo homozygous for the oep mutant allele, identified by failure to express gsc (9 out of 39 embryos). opl was patterned but expressed at lower levels.

A representative set of explants is shown in Fig. 6B. Cultured late blastula animal caps expressed low levels of opl, distributed randomly (Fig. 6Ba), consistent with RT-PCR data shown in Fig 4. When animal caps were conjugated with a shield, two outcomes were observed (Fig. 6Bb). In 22 out of 38 conjugates (58%), generated in 7 experiments, opl was strongly induced in the portion of the animal cap juxtaposed to the most anterior region of the shield (arrowhead). In the remaining 16 conjugates (32%) no opl expression was observed. In both cases, the randomly distributed patches of weak opl expression observed in isolated caps were suppressed. We previously showed that shield could induce expression of the anterior neural markers otx2 and pax6 in blastula caps with a frequency similar to the frequency of opl induction observed here (Sagerström et al., 1996). However, while otx2 and pax6 were expressed by isolated shield in culture (Sagerström et al., 1996), opl was not (Fig. 6Bc) (0 out of 40 explants analyzed). Since otx2 and pax6 are also expressed in more posterior domains than opl, this suggests that the shield is not capable of autonomously activating the most anterior neural fates.

In summary, these data demonstrated that although the early gastrula shield cannot autonomously activate expression of opl, it can efficiently induce opl expression in blastula stage ectoderm.

The entire late blastula organizer, but not the future prechordal plate, is required for activation of opl

Having demonstrated that the early gastrula organizer (shield) can induce anterior neural tissue, we asked whether it is also required for induction of this tissue. Since we had shown that anterior neural tissue was specified by early gastrula stages, but had not yet been specified at very late blastula (30-35% epiboly, Fig. 4C), we asked whether removal of the presumptive shield from younger embryos prevented opl expression. The presumptive shield region of the late blastula organizer was identified morphologically and the entire region, or portions thereof, was removed. Operated embryos were cultured until the end of gastrulation (90% epiboly equivalent, 9 hpf) and assayed for opl expression by in situ hybridization (Fig. 7A).

The late blastula organizer was identified morphologically at the asymmetric onset of epiboly, which results in a thinner margin on the dorsal side then on the ventral side (Schmitz and Campos-Ortega, 1994). This difference can be most easily observed at late blastula, the 30-35% epiboly stage. In a pilot experiment using eight embryos, we confirmed the correlation by marking the thinner side with FLDX injected into the adjacent yolk. When observed at shield stage, all embryos had formed shields near the position marked by the injection (data not shown). Additionally, in order to determine how much tissue to remove, we mapped the late blastula organizer using in situ hybridization of two markers, gsc (purple) which marks the presumptive prechordal plate (Stachel et al., 1993) and chd (orange) that marks the entire dorsal mesendoderm (organizer region) (Miller-Bertoglio et al., 1997). At 30% epiboly, chd is expresed in a domain which extends from the blastoderm margin to about 30% of the distance between the margin and the animal pole, while gsc expression is restricted to the vegetal half of the chd domain (Fig. 7Ba).

Using the early asymmetry in epiboly as a landmark, we first removed an area which included the entire chd expression domain at 30% epiboly (entire stippled area in Fig. 7A, see Materials and Methods). This surgery did not remove cells at the animal pole, which contain forebrain precursors (Kimmel and Warga, 1987). In order to control for the effects of trauma due to surgery, embryos were generated in which a significant portion of the organizer domain (>80%, solid gray area in Fig. 7A) was removed, while the vegetal-most portion of it was left intact. Cultured embryos were processed by in situ hybridization to detect simultaneously expression of opl (orange) and chd (purple). Although epiboly was abnormal in control embryos, all of the control embryos expressed both chd and opl (5 out of 5 embryos generated in 1 experiment) with opl expressed anterior to the chd domain, as in normal embryos (Fig. 7Bc). In contrast, all of the embryos from which the late blastula organizer had been successfully removed (8 out of 23 operated embryos generated in 4 experiments) failed to express opl (Fig. 7Bd), although they appeared healthy, with no indication of excessive cell death as visualized by acridine orange staining at mid-gastrula (data not shown). The remaining 15 out of 23 operated embryos expressed both chd and opl, suggesting that the complete organizer had not been removed.

In order to test the requirement for the presumptive prechordal plate in opl activation, we removed a smaller portion of the late blastula organizer (hatched area in Fig. 7A). Control embryos were generated by mock dissection (see Materials and Methods). Embryos were cultured until late gastrula stage equivalent and stained for opl (orange) and gsc (purple). Both in controls (11 embryos) (Fig. 7Be) and in most embryos from which gsc-expressing cells had been successfully removed (10 out of 17 embryos generated in 4 experiments) (Fig. 7Bf) opl was expressed and patterned correctly. One of 17 embryos was free of both opl and gsc expression. Both opl and gsc were expressed in the remaing 6 of 17 operated embryos, probably indicating that the mesendodermal gsc domain had not been completely removed. In order to confirm independently that the prechordal plate (gsc-expressing region) is dispensable for opl expression and patterning, we assayed opl expression in embryos mutant at the one eyed pinhead (oep) locus (Hammerschmidt et al., 1996a; Schier et al., 1997). oepte57 embryos activate gsc normally, but from 40% epiboly onward fail to maintain gsc expression, and do not form prechordal plates. oep mutant embryos do contain some organizer tissue, since they form shields and notochords. Sibling embryos derived from a cross between oep heterozygous parents were stained for opl (orange) and gsc (purple) (Fig. 7C). Wild-type embryos, which expressed gsc, (30 out of 39 tested) showed normally patterned opl expression (Fig. 7Ca). oep embryos that did not express gsc (9 out of 39 tested) also expressed opl in a correct pattern, albeit at lower levels than the wild-type embryos (Fig. 7Cb).

Two interpretations of these results are possible. The first is that the entire organizer, but not the prechordal plate, is required for opl expression starting at 30-35% epiboly. An alternative interpretation is that, in the process of removing the organizer, all precursors of the opl-expressing population were ablated, either directly or through cell death due to surgical trauma. Several lines of evidence support the first interpretation. First, fate mapping data strongly suggest that some (animal pole) cells fated as forebrain were not removed with the organizer. Kimmel and Warga (1987) showed that animal pole cells of a mid-blastula stage embryo remain at the animal pole until early gastrula and later contribute to forebrain structures. These cells can be induced to express opl in explants (Fig. 6), and therefore could presumably have been induced in the embryo, had appropriate inducing signals been present after surgery. Furthermore, when at least 80% of the mesendodermal region expressing chd was removed, leaving only a very thin strip of cells at the blastoderm margin as well as the animal pole cells, opl was efficiently induced. This indicates that failure to express opl after complete organizer removal was not a consequence of surgical trauma. Finally, using acridine orange staining (Furutani-Seiki et al., 1996), we have seen no increase in cell death between mock-dissected and organizer-ablated embryos.

In summary, these data demonstrated that the dorsal mesendodermal region marked by chd expression, the putative organizer, is required for normal opl activation. However, the presumptive prechordal plate is not required for correctly patterned opl expression, although it is required for high level expression of this gene.

When applied during gastrulation, BMP4 can ablate opl expression

Having shown that the late blastula organizer was required for opl activation, we next wanted to know whether organizer activity was also required during gastrulation for opl patterning and expression. This could not be done by removal of the early gastrula shield since this structure regenerates after removal at this stage (Y. G. and H. S., unpublished results). Bone Morphogenetic Proteins (BMPs), namely, BMP2, BMP4 and BMP7, promote ventral determination and must normally be antagonized by organizer-derived signaling to activate dorsal fates (Hawley et al., 1995; Sasai et al., 1995; Schmidt et al., 1995; Suzuki et al., 1997; Hammerschmidt et al., 1996; Kishimoto et al., 1997). We made use of BMP4 to ask whether signaling by the organizer during gastrulation is required for opl expression. Sepharose beads were coated with BMP4 protein (see Materials and Methods) and implanted into the dorsal side of early gastrula (shield, 6 hpf) stage embryos (Fig. 8A). Control beads were coated with bovine serum albumin (BSA). Embryos were cultured until late gastrula (90% epiboly, 9 hpf) and assayed for opl expression by in situ hybridization.

Fig. 8.

BMP4 inhibits opl expression in whole embryos. (A) Diagram of the experiment. Sepharose beads, coated with BSA or BMP4 (see Methods), were inserted into the dorsal side of shield stage embryos, between the yolk cell and the epiblast. Embryos were harvested at 90% epiboly stage equivalent and assayed for expression of opl (orange) by in situ hybridization. (B) Dorsoanterior views of representative embryos in which control BSA-coated (a) or BMP4-coated (b) Sepharose beads had been implanted. (A) A control embryo with implanted BSA-coated bead (arrowhead). Seven such embryos were obtained in 3 experiments. (b) Embryos obtained after implantation of BMP4 coated beads. 11 embryos generated in 3 experiments had BMP4 beads located dorsally (arrowheads); 3 of these did not express opl (like the embryo on the right), the other 8 showed partial ablation of the opl domain proximal to the bead (like the embryo on the left). a, anterior.

Fig. 8.

BMP4 inhibits opl expression in whole embryos. (A) Diagram of the experiment. Sepharose beads, coated with BSA or BMP4 (see Methods), were inserted into the dorsal side of shield stage embryos, between the yolk cell and the epiblast. Embryos were harvested at 90% epiboly stage equivalent and assayed for expression of opl (orange) by in situ hybridization. (B) Dorsoanterior views of representative embryos in which control BSA-coated (a) or BMP4-coated (b) Sepharose beads had been implanted. (A) A control embryo with implanted BSA-coated bead (arrowhead). Seven such embryos were obtained in 3 experiments. (b) Embryos obtained after implantation of BMP4 coated beads. 11 embryos generated in 3 experiments had BMP4 beads located dorsally (arrowheads); 3 of these did not express opl (like the embryo on the right), the other 8 showed partial ablation of the opl domain proximal to the bead (like the embryo on the left). a, anterior.

After embryo culture, control (BSA-coated) beads and BMP4-coated beads were found at similar positions with respect to the blastoderm margin. Control beads had no effect on opl expression (7 out of 7 embryos) (arrowhead in Fig. 8Ba). In contrast, opl expression was partially or completely ablated in the vicinity of a BMP4-coated bead in all cases (11 out of 11 embryos) (arrowheads in Fig. 8Bb), and the opl expression domain was located more caudally than in controls, that is, toward the blastoderm margin. In 8 of the 11 embryos, opl expression was only partially ablated by the BMP4-coated bead, while in 3 out of 11 embryos it was ablated completely. opl-expressing cells were never found adjacent to a BMP4-coated bead, however, we were not able to determine how close the bead needed to be to the opl domain in order to inhibit opl expression, except that beads needed to be located dorsally.

In summary, these experiments indicate that BMP4 antagonizes opl expression, consistent with a continued role for organizer-derived signaling in activating opl during gastrula stages.

Our examination of zebrafish forebrain determination makes three points. First, using two early neural markers, opl and fkh5, we have shown a very early molecular division of the forebrain field into an anterior region, the future telencephalon, and a more posterior region that includes the future diencephalon. Second, we have found that induction of these regions takes place by early gastrula, earlier than known previously. Third, using explant and extirpation assays, we have shown that telencephalic induction can be effected by the embryonic shield, and that a broad region of the zebrafish organizer is required for activation of forebrain determination. These findings have implications for the timing of and tissue interactions required for commitment to forebrain fates in other vertebrates.

Territories of the presumptive forebrain appear during gastrulation

During gastrulation, the homeobox-containing gene otx2 demarcates the entire presumptive forebrain and midbrain in zebrafish (Li et al., 1994), Xenopus (Blitz and Cho, 1995; Pannese et al., 1995) and mouse (Simeone et al., 1993). After neural tube closure, multiple genes demarcate different transverse and longitudinal domains within the forebrain (Rubenstein et al., 1994). However, the issue of when the forebrain is first divided into territories has not been clarified. At late gastrula, the opl expression domain appears to correspond to the presumptive telencephalon while the anterior portion of the fkh5 expression domain appears to correspond to the presumptive diencephalon as defined by fate mapping (Woo and Fraser, 1995). Together, the expression patterns of opl and fkh5 define a very early molecular subdivision of the presumptive forebrain which precedes any morphological subdivision.

Starting at mid-gastrula, zebrafish opl and fkh5 are expressed in patterns which are very similar, both spatially and temporally, to those of their Xenopus homologs (Gamse and Sive, unpublished data). However, in Xenopus, opl and fkh5 are also strongly expressed by the onset of gastrulation, at an earlier stage than they are in zebrafish, suggesting that some levels of gene control are different between these organisms (Gamse and Sive, unpublished data). Mouse homologs of opl (Zic1) and fkh5 are expressed in analogous domains during early development: Zic1 in the neurectoderm and then in the dorsal neural tube (Nagai et al., 1997) and fkh5 in the diencephalon, mesencephalon, rhombencephalon and spinal cord (Labosky et al., 1997; Wehr et al., 1997). These similarities may suggest that both the signaling pathways, which lead to the patterned expression of genes in the early neurectoderm, and the functions of these genes are conserved among vertebrates.

Timing of forebrain patterning

We showed that neurectoderm is specified by early gastrula (shield) stage, and that some neural pattern has also been specified by this stage. Three distinct domains of specification are present in the early gastrula ectoderm, shown as a ‘specification map’ in Fig. 9A. This map was drawn by summarizing the results of specification assays using explants made at different positions of the early gastrula ectoderm (Fig. 5). The overlap between ‘specification domains’ in this map reflects that none of the markers used was specified exclusively in one explant type. Instead, markers were specified primarily in one explant type, but also weakly in neighboring explants. This may be due to lack of sharp boundaries between specification domains or to our inability to explant precisely the same group of cells from different embryos.

Fig. 9.

A model of early neural induction and patterning in zebrafish. (A) Comparison of the early gastrula specification map obtained in this study with a fate map constructed by labeling cells at early gastrula (shield) stage and scoring their fates after morphological definition of the brain (Woo and Fraser, 1995). On the fate map, positions marked with dots indicate contribution to the following tissues in the 19-36 hour (20 somite-prim25) stage embryo. The specification map was drawn by summarizing the results of specification assays (Fig. 5). The overlap between ‘specification domains’ (circled regions) indicates that none of the markers used was specified exclusively in one explant type. This may be due to lack of sharp boundaries between specification domains or to an inability to explant precisely the same group of cells from different embryos. ve, ventral non-neural ectoderm; tel, telencephalon; di, diencephalon. (B) A model of neural induction and patterning showing the degree of specification of the ectodermal A/P pattern and the signals responsible for this pattern. Different fates are represented by different colors, as in A. At late blastula, specification for both ventral (strongly expressed) and dorsal (weakly expressed) ectodermal markers is observed in the animal cap. This might be supplied by cell autonomous factors, by signals from the yolk cell (lateral red arrows) or by signals from the marginal zone (dorsal and ventral colored arrows). By early gastrula, ventral ectoderm is strongly specified as non-neural, and dorsal ectoderm as the forebrain. Furthermore, the subdivision of the forebrain into prospective telencephalon and diencephalon is already specified. This strong specification is, at least in part, due to signals originating from the organizer (dorsal margin/shield) (red arrow) and its signaling antagonist, the ventral margin (blue arrow) (Sagerström et al., 1996) (see text for more discussion).

Fig. 9.

A model of early neural induction and patterning in zebrafish. (A) Comparison of the early gastrula specification map obtained in this study with a fate map constructed by labeling cells at early gastrula (shield) stage and scoring their fates after morphological definition of the brain (Woo and Fraser, 1995). On the fate map, positions marked with dots indicate contribution to the following tissues in the 19-36 hour (20 somite-prim25) stage embryo. The specification map was drawn by summarizing the results of specification assays (Fig. 5). The overlap between ‘specification domains’ (circled regions) indicates that none of the markers used was specified exclusively in one explant type. This may be due to lack of sharp boundaries between specification domains or to an inability to explant precisely the same group of cells from different embryos. ve, ventral non-neural ectoderm; tel, telencephalon; di, diencephalon. (B) A model of neural induction and patterning showing the degree of specification of the ectodermal A/P pattern and the signals responsible for this pattern. Different fates are represented by different colors, as in A. At late blastula, specification for both ventral (strongly expressed) and dorsal (weakly expressed) ectodermal markers is observed in the animal cap. This might be supplied by cell autonomous factors, by signals from the yolk cell (lateral red arrows) or by signals from the marginal zone (dorsal and ventral colored arrows). By early gastrula, ventral ectoderm is strongly specified as non-neural, and dorsal ectoderm as the forebrain. Furthermore, the subdivision of the forebrain into prospective telencephalon and diencephalon is already specified. This strong specification is, at least in part, due to signals originating from the organizer (dorsal margin/shield) (red arrow) and its signaling antagonist, the ventral margin (blue arrow) (Sagerström et al., 1996) (see text for more discussion).

Our data indicate the presence at early gastrula of a ventral domain, specified as ventral, primarily non-neural, ectoderm, an anterior dorsal domain, specified primarily as telencephalon, and a posterior dorsal domain, specified primarily as diencephalon. We compared our specification map of the early gastrula ectoderm to a fate map made by dye marking cells of early gastrula embryos and scoring their fates after the brain had morphologically formed (Woo and Fraser, 1995). This comparison (Fig. 9A) shows a striking overlap between the domains we defined by specification assays and the domains defined by fate mapping. Since a fate map defines the regions of the embryo which will form given organs, but contains no information about when the specification to form such organs is first acquired, this overlap was not necessarily expected. Early gastrula is the earliest stage at which a fate map of the zebrafish embryo can be drawn, probably due to slowing down of cell mixing, and the coincidence of this with activation of neural specification and patterning suggests that inductive centers are in place prior to early gastrula.

The region of dorsal ectoderm which is fated to give rise to rhombencephalon is specified as anterior neural tissue at early gastrula, since after culture it expresses anterior markers opl and fkh5, but not the more posterior hindbrain marker krx20. This finding has recently been confimed using several novel markers of posterior neurectoderm isolated in our laboratory (M. E. Lane, C. G. Sagerström, and H. S.). These data support the idea that in zebrafish the tissue fated to become rhombencephalon is first induced as anterior and later reprogrammed into more posterior neural tissue. Recent studies suggest that the zebrafish germ ring is a source of posteriorizing signals (Woo and Fraser, 1997; M. E. Lane and H. S., unpublished data) which may convert the neurectoderm, initially specified as anterior, into more posterior derivatives (Kolm and Sive, 1997).

Tissues responsible for forebrain specification and patterning

Previous studies using transplant (Shih and Fraser, 1996) and explant (Sagerström et al., 1996) analyses have shown that the zebrafish embryonic shield is a potent source of neural inducing signals. The shield encompasses the gsc expression domain, which has previously been shown to mark the prospective prechordal plate (Stachel et al., 1993). However, the zebrafish organizer is larger than the gsc domain and may correspond to the chd expression domain (Miller-Bertoglio et al., 1997). Since the organizer field has not been carefully mapped in fish, we relied on the evidence from Xenopus, where the functional organizer correlates well with the domain of chd expression at late blastula and early gastrula stages (Sasai et al., 1994; Stewart and Gerhart, 1990; Vodicka and Gerhart, 1995). By removing the entire chd expression domain we have shown that dorsal mesendoderm is required for anterior neural patterning marked by opl expression. This effect is not due simply to the removal of chd activity, since genetic deletion of the chd locus does not completely prevent the formation and patterning of neural tissue (Fisher et al., 1997; Hammerschmidt et al., 1996b; Schulte-Merker et al., 1997).

We used two methods to demonstrate that the presumptive prechordal plate is dispensible for opl activation and patterning but may contribute to the correct levels of gene expression: extirpation of the presumptive prechordal plate and analysis of oep mutant embryos, which fail to form the prechordal plate. Both methods allowed expression of opl in the correct position in the extreme anterior part of the neural plate. In oep mutant embryos opl expression levels were significantly reduced relative to wild type controls, but this was not the case in embryos from which the presumptive prechordal plate was removed microsurgically. One possible explanation for this difference is that the domain removed microsurgically also included posterodorsal mesendoderm. Posterodorsal mesendoderm, including prospective notochord, may suppress extreme anterior neural determination and removal of this inhibitory region may compensate for the loss of positive signals coming from the anterior mesendoderm. This idea is supported by observations in Xenopus, where posterodorsal mesendoderm can inhibit anterior neural determination (Sive et al., 1990, 1989). Alternatively, since the gene identified by the oep mutation is highly expressed in the prospective forebrain itself by late gastrula (Zhang et al., 1998), it may play a direct role in maintaining high levels of anterior neural gene expression.

The early gastrula organizer is functionally heterogeneous (Oppenheimer, 1959; Zoltewicz and Gerhart, 1997). In amphibian and chick embryos, the presumptive prechordal plate (the gsc-expressing region) is correlated with head organizer activity (Foley et al., 1997; Holtfreter, 1938a,b; Pera and Kessel, 1997; Zoltewicz and Gerhart, 1997). However, consistent with our results, removal of this region from chick embryos does not prevent the correct patterning of the neural anteroposterior axis (Pera and Kessel, 1997) and organizers from which this region is deleted at early gastrula are fully functional in secondary axis induction (Psychoyos and Stern, 1996).

While the prechordal plate is not required for opl induction, the early gastrula shield, which includes the presumptive prechordal plate, is able to induce strong opl expression in blastula stage ectoderm. This was not necessarily expected, since isolated and cultured shield did not go on to express opl, and since transplanted shields do not activate head formation in secondary axes (Shih and Fraser, 1996). The ability of the shield to induce opl expression indicates that it contains molecules which activate determination of the telencephalon. The failure of isolated shield to express opl, together with the specification in intact embryos of opl at a distance from the presumptive shield, may suggest that telencephalon-inducing signals must act at a minimum distance from their source in the presumptive shield, perhaps after formation of a gradient. This idea is consistent with the observation that different doses of BMPs, or anti-BMP molecules such as noggin, lead to different readouts along the anteroposterior axis (Knecht and Harland, 1997; Wilson et al., 1997).

Although the blastula stage organizer (dorsal mesendoderm) is required for expression of opl by late gastrula, we do not know whether this is the tissue that initiates opl specification by early gastrula. Some inducing signals may be produced by the yolk cell, which contacts anterior ectoderm by the onset of gastrulation and which contains mesoderm-inducing and patterning capacity (Mizuno et al., 1996). It is possible, but not demonstrated, that this tissue is equivalent to the dorsoanterior endoderm in Xenopus and the extraembryonic endoderm of mice. However, since organizer removal prevents opl expression, the yolk cell is clearly not sufficient for opl activation. A recent interesting report indicates that during mid-to late gastrula stages in zebrafish a row of anterior ectodermal cells is required for forebrain patterning (Houart et al., 1998). The relationship between the forebrain patterning function of these cells and that of the late blastula organizer identified here is not clear since the two studies analyse events occurring at different times of gastrulation. In particular, it is not clear whether the neurectodermal region (Houart et al., 1998) is active at earlier stages, or whether it is activated subsequent to the action of the late blastula organizer and initial neural patterning events we describe here.

A model for induction and patterning of anterior neurectoderm in zebrafish

A model for early neural patterning in the zebrafish embryo which summarizes our findings is shown in Fig. 9B. In a late blastula embryo cells at the animal pole are specified primarily as ventral (non-neural) ectoderm, including prospective epidermis, with weak dorsal character. Signals that activate this specification may be cell autonomous, or may be derived from the yolk cell or from the blastoderm margin. Two hours later, at early gastrula, dorsal ectoderm is already specified as neural with an anteroposterior axis. The prospective telencephalon lies anteriorly and the prospective diencephalon posteriorly. We conclude that the initial induction and patterning of the forebrain in zebrafish takes place just prior to, or coincident with, the onset of gastrulation. Expression of the forebrain markers we have used does not begin until mid-gastrula, and it is not known whether a wave of patterned gene expression precedes that of opl and fkh5. Our data suggest that the signals responsible for forebrain patterning arise from the late blastula organizer (dorsal mesendoderm). Inducing signals may also originate from the yolk cell that underlies the future telencephalon, although these signals are not sufficient to activate expression of forebrain markers. A later source of inducing signals may also arise from the anterior neurectoderm itself (Houart et al., 1998).

Future directions include analysis of the mechanisms of patterned gene expression within the forming forebrain, both with regard to the cis-acting promoter elements and the trans-acting factors responsible for generating this pattern.

We thank J. Rosenfield and H. Guo for help with cloning zebrafish opl and fkh5, our collegues S. Fisher and M. Halpern for zebrafish chd and G. Conway for α-tubulin prior to publication, and A. Schier for oep mutant embryos. Special thanks to members of the Sive lab and S. Sokol for critical reading of the manuscript, and to Vladimir Apekin for expert care of the fish. We also thank Liuda Ziaugra for DNA sequence analysis and Jeanne Reis for sectioning embryos. This work was supported by the Whitehall Foundation and the NSF and by postdoctoral fellowship awards from NIH and the Medical Foundation/Charles A. King Trust to Y. G. H. S. is the Latham Family Career Development Professor at MIT.

Ang
,
S.-L.
,
Conlon
,
R. A.
,
Jin
,
O.
and
Rossant
,
J.
(
1994
).
Positive and negative signals from mesoderm regulate the expression of mouse Otx2 in ectoderm explants
.
Development
120
,
2979
2989
.
Aruga
,
J.
,
Yokota
,
N.
,
Hashimoto
,
M.
,
Furuichi
,
T.
,
Fukuda
,
M.
and
Mikoshiba
,
K.
(
1994
).
A novel zinc finger protein, Zic, is involved in neurogenesis, especially in the cell lineage for cerebellar granule cells
.
J. Neurochem
.
63
,
1880
1890
.
Bally-Cuif
,
L.
and
Boncinelli
,
E.
(
1997
).
Transcription factors and head formation in vertebrates
.
BioEssays
19
,
127
135
.
Beddington
,
R. S. P.
(
1994
).
Induction of a second neural axis by the mouse node
.
Development
120
,
613
620
.
Benedyk
,
M. J.
,
Mullen
,
J. R.
,
DiNardo
,
S.
(
1994
).
Odd-paired: a zinc finger pair-rule protein required for the timely activation of engrailed and wingless in Drosophila embryos
.
Genes Dev
.
8
,
105
117
.
Blitz
,
I. L.
and
Cho
,
K. W. Y.
(
1995
).
Anterior neuroectoderm is progressively induced during gastrulation: the role of the Xenopus homeobox gene orthodenticle
.
Development
121
,
993
1004
.
Bouwmeester
,
T.
,
Kim
,
S.
,
Sasai
,
Y.
,
Lu
,
B.
and
De Robertis
,
E.
(
1996
).
Cerberus is a head-inducing secreted factor expressed in the anterior endoderm of Spemann’s organizer
.
Nature
382
,
595
601
.
Bradley
,
L.
,
Wainstock
,
D.
and
Sive
,
H. L.
(
1996
).
Positive and negative signals modulate formation of the Xenopus cement gland
.
Development
122
,
2739
2750
.
Church
,
G. M.
and
Gilbert
,
W.
(
1984
).
Genomic sequencing
.
Proc. Natl. Acad. Sci. USA
81
,
1991
1995
.
Driever
,
W.
,
Stemple
,
D.
,
Schier
,
A.
and
Solnica-Krezel
,
L.
(
1994
).
Zebrafish: genetic tools for studying vertebrate development
.
Trends Genet
.
10
,
152
159
.
Eakin
,
R. M.
(
1939
).
Regional determination in the development of the trout
.
Wilhelm Roux Arch. EntwMech. Org
.
139
,
274
281
.
Eyal-Giladi
,
H.
(
1954
).
Dynamic aspects of neural induction
.
Arch. Biol
.
65
,
180
259
.
Fisher
,
S.
,
Amacher
,
S.
and
Halpern
,
M.
(
1997
).
Loss of cerebum function ventralizes the zebrafish embryo
.
Development
124
,
1301
1311
.
Foley
,
A.
,
Storey
,
K.
and
Stern
,
C.
(
1997
).
The prechordal region lacks neural inducing ability, but can confer anterior character to more posterior neuroepithelium
.
Development
124
,
2983
2996
.
Furutani-Seiki
,
M.
,
Jiang
,
Y.
,
Brand
,
M.
,
Heisenberg
,
C.
,
Houart
,
C.
,
Beuchle
,
D.
,
van Eeden
,
F.
,
Granato
,
M.
,
Haffter
,
P.
,
Hammerschmidt
,
M.
,
Kane
,
D.
,
Kelsh
,
R.
,
Mullins
,
M.
,
Odenthal
,
J.
and
Nusslein-Volhard
,
C.
(
1996
).
Neural degeneration mutants in the zebrafish, Danio rerio
.
Development
123
,
229
239
.
Gerhart
,
J.
(
1996
).
Johannes Holtfreter’s contributions to ongoing studies of the organizer
.
Dev. Dyn
.
205
,
245
256
.
Glinka
,
A.
,
Wu
,
W.
,
Delius
,
H.
,
Monaghan
,
A.
,
Blumenstock
,
C.
and
Niehrs
,
C.
(
1998
).
Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction
.
Nature
391
,
357
362
.
Glinka
,
A.
,
Wu
,
W.
,
Onichtchouk
,
D.
,
Blumenstock
,
C.
and
Niehrs
,
C.
(
1997
).
Head induction by simultaneous repression of Bmp and Wnt signalling in Xenopus
.
Nature
389
,
517
519
.
Grinblat
,
Y.
,
Lane
,
M. E.
,
Sagerström
,
C. G.
, and
Sive
,
H. L.
(
1998
). Analysis of Zebrafish development using explant culture assays. In
The Zebrafish: Methods in Cell Biology
. vol. 59 (ed.
H. W.
Dietrich
,
M.
Westerfield
and
L.
Zon
), pp.
127
-
159
. San Diego:
Academic Press
.
Hammerschmidt
,
M.
,
Pelegri
,
F.
,
Mullins
,
M.
,
Kane
,
D.
,
Brand
,
M.
,
van Eeden
,
F.
,
Furutani-Seiki
,
M.
,
Granato
,
M.
,
Haffter
,
P.
,
Heisenberg
,
C.
,
Jiang
,
Y.
,
Kelsh
,
R.
,
Odenthal
,
J.
,
Warga
,
R.
and
Nusslein-Volhard
,
C.
(
1996a
).
Mutations affecting morphogenesis during gastrulation and tail formation in the zebrafish, Danio rerio
.
Development
123
,
143
151
.
Hammerschmidt
,
M.
,
Pelegri
,
F.
,
Mullins
,
M.
,
Kane
,
D.
,
van Eeden
,
F.
,
Granato
,
M.
,
Brand
,
M.
,
Furutani-Seiki
,
M.
,
Haffter
,
P.
,
Heisenberg
,
C.
,
Jiang
,
Y.
,
Kelsh
,
R.
,
Odenthal
,
J.
,
Warga
,
R.
and
Nusslein-Volhard
,
C.
(
1996b
).
dino and mercedes, two genes regulating dorsal development in the zebrafish embryo
.
Development
123
,
95
102
.
Hammerschmidt
,
M.
,
Serbedzija
,
G.
and
McMahon
,
A.
(
1996
).
Genetic analysis of dorsoventral pattern formation in the zebrafish: requirement of a BMP-like ventralizing activity and its dorsal repressor
.
Genes Dev
.
10
,
2452
2461
.
Hawley
,
S. H.
,
Wunnenberg-Stapleton
,
K.
,
Hashimoto
,
C.
,
Laurent
,
M. N.
,
Watabe
,
T.
,
Blumberg
,
B. W.
and
Cho
,
K. W.
. (
1995
).
Disruption of BMP signals in embryonic Xenopus ectoderm leads to direct neural induction
.
Genes Dev
.
9
,
2923
2935
.
Holtfreter
,
J.
(
1938a
).
Differenzungspotenzen isolierter Teile der Anurengastrula. Wilhelm Roux Arch. EntwMech. Org
..
138
,
657
738
.
Holtfreter
,
J.
(
1938b
).
Differenzungspotenzen isolierter Teile der Urodelengastrula
.
Wilhelm Roux Arch. EntwMech. Org
.
138
,
522
656
.
Holtfreter
,
J.
and
Hamburger
,
V.
(
1955
). Embryogenesis: progressive differentiation. In
Analysis of Development
. (ed.
B. H.
Willier
,
P.
Weiss
and
V.
Hamburger
). pp.
230
-
296
. Philadelphia PA: Saunders.
Houart
,
C.
,
Westerfield
,
M.
and
Wilson
,
S. W.
(
1998
).
A small population of anterior cells patterns the forebrain during zebrafish gastrulation
.
Nature
391
,
788
792
.
Jones
,
E.
and
Woodland
,
H.
(
1989
).
Spatial aspects of neural induction in Xenopus laevis
.
Development
107
,
785
791
.
Kaestner
,
K.
,
Schutz
,
G.
and
Monaghan
,
A.
(
1996
).
Expression of the winged helix genes fkh-4 and fkh-5 defines domains in the central nervous system
.
Mech Dev
.
55
,
221
230
.
Kimmel
,
C. B.
(
1989
).
Genetics and early development of zebrafish
.
Trends Genet
.
5
,
283
288
.
Kimmel
,
C. B.
(
1993
).
Patterning the brain of the zebrafish embryo
.
Annu. Rev. Neurosci
.
16
,
707
732
.
Kimmel
,
C. B.
and
Warga
,
R. M.
(
1987
).
Indeterminate cell lineage of the zebrafish embryo
.
Dev. Biol
.
124
,
269
280
.
Kimmel
,
C. B.
,
Warga
,
R. M.
and
Schilling
,
T. F.
(
1990
).
Origin and organization of the zebrafish fate map
.
Development
108
,
581
594
.
Kishimoto
,
Y.
,
Lee
,
K.-H.
,
Zon
,
L.
,
Hammerschmidt
,
M.
and
Schulte-Merker
,
S.
(
1997
).
The molecular nature of zebrafish swirl: BMP2 function is essential during early dorsoventral patterning
.
Development
124
,
4457
4466
.
Knecht
,
A. K.
and
Harland
,
R. M.
(
1997
).
Mechanisms of dorsal-ventral patterning in noggin-induced neural tissue
.
Development
124
,
2477
2488
.
Kolm
,
P. J.
and
Sive
,
H. L.
(
1997
).
Retinoids and posterior neural induction: a reevaluation of Nieuwkoop’s two-step hypothesis
.
Cold Spring Harbor Symp. Quant. Biol
.
62
,
511
521
.
Krauss
,
S.
,
Johansen
,
T.
,
Korzh
,
V.
,
Moens
,
U.
,
Ericson
,
J. U.
and
Fjose
,
A.
(
1991
).
Zebrafish pax [zf-a]: A paired box-containing gene expressed in the neural tube
.
EMBO J
.
10
,
3609
3619
.
Kuo
,
J.
,
Patel
,
M.
,
Gamse
,
J.
,
Merzdorf
,
C.
,
Apekin
,
V.
,
Liu
,
X.
and
Sive
,
H. L.
(
1998
).
Opl: a zinc finger protein that regulates neural determination and patterning in Xenopus
.
Development
.
125
,
2867
2882
.
Labosky
,
P. A.
,
Winnier
,
G. E.
,
Jetton
,
T. L.
,
Hargett
,
L.
,
Ryan
,
A. K.
,
Rosenfeld
,
M. G.
,
Parlow
,
A. F.
and
Hogan
,
B. L. M.
(
1997
).
The winged helix gene, Mf3, is required for normal development of the diencephalon and midbrain, postnatal growth and the milk-ejection reflex
.
Development
124
,
1263
1274
.
Li
,
Y.
,
Allende
,
M. L.
,
Finkelstein
,
R.
and
Weinberg
,
E. S.
(
1994
).
Expression of two zebrafish orthodenticle-related genes in the embryonic brain
.
Mech. Dev
.
48
,
229
244
.
Macdonald
,
R.
,
Xu
,
Q.
,
Barth
,
K. A.
,
Mikkola
,
I.
,
Holder
,
N.
,
Fjose
,
A.
,
Krauss
,
S.
and
Wilson
,
S. W.
(
1994
).
Regulatory gene expression boundaries demarcate sites of neuronal differentiation in the embryonic zebrafish forebrain
.
Neuron
.
13
,
1039
1053
.
Mangold
,
O.
and
Spemann
,
H.
(
1927
).
Uber Induktion von Medullarplatte durch Medullarplatte imjungeren
.
Wilhelm Roux Arch. EntwMech. Org
.
111
,
341
422
.
McGrew
,
L.
,
Hoppler
,
S.
and
Moon
,
R.
(
1997
).
Wnt and FGF pathways cooperatively pattern anteroposterior neural ectoderm in Xenopus
.
Mech Dev
.
69
,
105
114
.
Miller-Bertoglio
,
V.
,
Fisher
,
S.
,
Sanchez
,
A.
,
Mullins
,
M.
and
Halpern
,
M.
(
1997
).
Differential regulation of chordin expression domains in mutant zebrafish
.
Dev Biol
.
192
,
537
550
.
Mizuno
,
T.
,
Yamaha
,
E.
,
Wakahara
,
M.
,
Kurolwa
,
A.
and
Takeda
,
H.
(
1996
).
Mesoderm induction in zebrafish
.
Nature
383
,
131
132
.
Nagai
,
T.
,
Aruga
,
J.
,
Takada
,
S.
,
Gunther
,
T.
,
Sporle
,
R.
,
Schughart
,
K.
and
Mikoshiba
,
K.
(
1997
).
The expression of the mouse Zic1, Zic2, and Zic3 gene suggests an esssential role for Zic genes in body pattern formation
.
Dev. Biol
.
182
,
299
313
.
Nakata
,
K.
,
Nagai
,
T.
,
Aruga
,
J.
and
Mikoshiba
,
K.
(
1997
).
Xenopus Zic3, a primary regulator both in neural and neural crest development
.
Proc Natl Acad Sci. USA
94
,
11980
11985
.
Neave
,
B.
,
Rodaway
,
A.
,
Wilson
,
S. W.
,
Patient
,
R.
and
Holder
,
N.
(
1995
).
Expression of zebrafish GATA3 (gta3) during gastrulation and neurulation suggests a role in the specification of cell fate
.
Mech. Dev
.
51
,
169
182
.
Oppenheimer
,
J. M.
(
1959
).
Extraembryonic transplantations of sections of the Fundulus embryonic shield
.
J. Exp. Zool
.
140
,
247
268
.
Oppenheimer
,
J. M.
(
1936
).
Transplantation experiments on developing teleosts (Fundulus and Perca
).
J. Exp. Zool
.
72
,
409
437
.
Oxtoby
,
E.
and
Jowett
,
T.
(
1993
).
Cloning of the zebrafish Krox-20 gene (krx20) and its expression during hindbrain development
.
Nucl. Acids Res
.
21
,
1087
1095
.
Pannese
,
M.
,
Polo
,
C.
,
Andreazzoli
,
M.
,
Vignali
,
R.
,
Kablar
,
B.
,
Barsacchi
,
B.
and
Boncinelli
,
B.
(
1995
).
The Xenopus homologue of Otx2 is a maternal homeobox gene that demarcates and specifies anterior body regions
.
Development
121
,
707
720
.
Patel
,
M.
and
Sive
,
H. L.
(
1996
). PCR-based subtractive cDNA cloning. In
Current Protocols in Molecular Biology
. (ed.
F. M.
Ausubel
,
R.
Brent
,
R. E.
Kingston
,
D. D.
Moore
,
J. G.
Seidman
,
J. A.
Smith
and
K.
Struhl
).
New York: John Wiley & Sons
.
Pera
,
E. M.
and
Kessel
,
M.
(
1997
).
Patterning of the chick forebrain anlage by the prechordal plate
.
Development
124
,
4153
4162
.
Piccolo
,
S.
,
Sasai
,
Y.
,
Lu
,
B.
and
De Robertis
,
E. M.
(
1996
).
Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of chordin to BMP-4
.
Cell
86
,
589
598
.
Psychoyos
,
D.
and
Stern
,
C.
(
1996
).
Restoration of the organizer after radical ablation of Hensen’s node and the anterior primitive streak in the chick embryo
.
Development
122
,
3263
3273
.
Puelles
,
L.
and
Rubenstein
,
J.
(
1993
).
Expression patterns of homeobox and other putative regulatory genes in the embryonic mouse forebrain suggest a neuromeric organization
.
Trends Neurosci
.
16
,
472
479
.
Roberts
,
C.
,
Platt
,
N.
,
Streit
,
A.
,
Schachne
,
r. M.
and
Stern
,
C.
(
1991
).
The L5 epitope: an early marker for neural induction in the chick embryo and its involvement in inductive interactions
.
Development
112
,
959
970
.
Rossant
,
J.
and
Hopkins
,
N.
(
1992
).
Of fin and fur: mutational analysis of vertebrate embryonic development
.
Genes Dev
.
6
,
1
13
.
Rubenstein
,
J. L. R.
,
Nartinez
,
S.
,
Shimamura
,
K.
and
Puelles
,
L.
(
1994
).
The embryonic vertebrate forebrain: the prosomeric model
.
Science
266
,
578
580
.
Sagerström
,
C. G.
,
Grinblat
,
Y.
and
Sive
,
H.
(
1996
).
Anteroposterior patterning in the zebrafish, Danio rerio: an explant assay reveals inductive and suppressive cell interactions
.
Development
122
,
1873
1883
.
Sasai
,
Y.
and
De Robertis
,
E. M.
(
1997
).
Ectodermal patterning in vertebrate embryos
.
Dev. Biol
.
182
,
5
20
.
Sasai
,
Y.
,
Lu
,
B.
,
Steinbeisser
,
H.
,
Geissert
,
D.
,
Gont
,
L.
and
De Robertis
,
E.
(
1994
).
Xenopus chordin: a novel dorsalizing factor activated by organizer-specific homeobox genes
.
Cell
79
,
779
790
.
Schier
,
A.
,
Neuhauss
,
S.
,
Helde
,
K.
,
Talbot
,
W.
and
Driever
,
W.
(
1997
).
The one-eyed pinhead gene functions in mesoderm and endoderm formation in zebrafish and interacts with no tail
.
Development
124
,
327
342
.
Schmidt
,
J. E.
,
Suzuki
,
A.
,
Ueno
,
N.
,
Kimelman
,
D.
(
1995
).
Localized BMP- 4 mediates dorsal/ventral patterning in the early Xenopus embryo
.
Dev. Biol
.
169
,
37
50
.
Schmitz
,
B.
and
Campos-Ortega
,
J. A.
(
1994
).
Dorso-ventral polarity of the zebrafish embryo is distinguishable prior to the onset of gastrulation. Roux’ Arch. Dev. Biol
.
203
,
374
380
.
Schulte-Merker
,
S.
,
Lee
,
K. J.
,
McMahon
,
A. P.
and
Hammerschmidt
,
M.
(
1997
).
The zebrafish organizer requires chordino
.
Nature
387
,
862
863
.
Schulte-Merker
,
S.
,
van Eeden
,
F.
,
Halpern
,
M. E.
,
Kimmel
,
C. B.
and
Nüsslein-Volhard
,
C.
(
1994
).
no tail (ntl) is the zebrafish homologue of the mouse T (brachyury) gene
.
Development
120
,
1009
1015
.
Shih
,
J.
and
Fraser
,
S. E.
(
1996
).
Characterizing the zebrafish organizer: microsurgical analysis at the early-shield stage
.
Development
122
,
1313
1322
.
Shih
,
J.
and
Fraser
,
S. E.
(
1995
).
Distribution of tissue progenitors within the shield region of the zebrafish gastrula
.
Development
121
,
2755
2765
.
Simeone
,
A.
,
Acompora
,
D.
,
Mallamaci
,
A.
,
Stornaiuolo
,
A.
,
D’Apice
,
M. R.
,
Nigro
,
V.
and
Boncinelli
,
E.
(
1993
).
A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm of the gastrulating mouse embryo
.
EMBO J
.
12
,
2735
2747
.
Sive
,
H.
,
Draper
,
B.
,
Harland
,
R.
and
Weintraub
,
H.
(
1990
).
Identification of a retinoic acid sensitive period during primary axis formation in Xenopus laevis
.
Genes Dev
.
4
,
932
942
.
Sive
,
H. L.
,
Hattori
,
K.
and
Weintraub
,
H.
(
1989
).
Progressive determination during formation of the anteroposterior axis in Xenopus laevis
.
Cell
58
,
171
80
.
Spemann
,
H.
and
Mangold
,
H.
(
1964
). Part Ten: 1924 – Induction of embryonic primordia by implantation of organizers from a different species. In
Foundations of Experimental Embryology
(ed.
B. H.
Willier
and
J. M.
Oppenheimer
). pp.
145
-
184
. Englewood Cliffs, N. J. : Prentice-Hall, Inc.
Stachel
,
S. E.
,
Grunwald
,
D. J.
and
Myers
,
P. Z.
(
1993
).
Lithium perturbation and goosecoid expression identify a dorsal specification pathway in the pregastrula zebrafish
.
Development
117
,
1261
1274
.
Stewart
,
R. M.
and
Gerhart
,
J. C.
(
1990
).
The anterior extent of dorsal development of the Xenopus embryonic axis depends on the quantity of organizer in the late blastula
.
Development
109
,
363
372
.
Storey
,
K. G.
,
Crossley
,
J. M.
,
De Robertis
,
E. M.
,
Norris
,
W. F.
and
Stern
,
C. D.
(
1992
).
Neural induction and regionalisation in the chick embryo
.
Development
114
,
729
741
.
Suzuki
,
A.
,
Kaneko
,
E.
,
Ueno
,
N.
and
Hemmati-Brivanlou
,
A.
(
1997
).
Regulation of epidermal induction by BMP2 and BMP7 signalling
.
Dev. Biol
.
189
,
112
122
.
Thisse
,
C.
,
Thisse
,
B.
,
Halpern
,
M. E.
and
Postlethwait
,
J. H.
(
1994
).
goosecoid expression in neurectoderm and mesendoderm is disrupted in zebrafish cyclops gastrula
.
Dev. Biol
.
164
,
420
429
.
Thomas
,
P.
and
Beddington
,
R.
(
1996
).
Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo
.
Curr. Biol
.
6
,
1487
1496
.
Toyama
,
R.
,
O’Connel
,
M. L.
,
Wright
,
C. V. E.
,
Kuehn
,
M. R.
and
Dawid
,
I. B.
(
1995
).
Nodal induces ectopic goosecoid and lim1 expression and axis duplication in zebrafish
.
Development
121
,
383
391
.
Varlet
,
I.
,
Collignon
,
J.
and
Robertson
,
E. J.
(
1997
).
Nodal expression in the primitive endoderm is required for specification of the anterior axis during mouse gastrulation
.
Development
124
,
1033
1044
.
Vodicka
,
M.
and
Gerhart
,
J.
(
1995
).
Blastomere derivation and domains of gene expression in the Spemann Organizer of Xenopus laevis
.
Development
121
,
3505
3518
.
Waddington
,
C. H.
(
1932
).
Experiments on the development of the chick and the duck embryo cultivated in vitro
.
Proc. Trans. R. Soc. Lond
.
211
,
179
230
.
Wehr
,
R.
,
Mansouri
,
A.
,
de Maeyer
,
T.
and
Gruss
,
P.
(
1997
).
Fkh5-deficient mice show dysgenesis in the caudal midbrain and hypothalamic mammillary body
.
Development
124
,
4447
4456
.
Westerfield
,
M.
(
1995
).
The Zebrafish Book
.
Third Edition
.
Eugene; Univ. Oregon Press
.
Wilson
,
P. A.
,
Lagna
,
G.
,
Suzuki
,
A.
and
Hemmati-Brivanlou
,
A.
(
1997
).
Concentration-dependent patterning of the Xenopus ectoderm by BMP4 and its signal transducer Smad1
.
Development
124
,
3177
3184
.
Woo
,
K.
and
Fraser
,
S.
(
1997
).
Specification of the zebrafish nervous system by nonaxial signals
.
Science
277
,
254
257
.
Woo
,
K.
and
Fraser
,
S. E.
(
1995
).
Order and coherence in the fate map of the zebrafish nervous system
.
Development
121
,
2595
2609
.
Zhang
,
J.
,
Talbot
,
W.
and
Schier
,
A.
(
1998
).
Positional cloning identifies zebrafish one-eyed pinhead as a permissive EGF-related ligand required during gastrulation
.
Cell
92
,
241
251
.
Zimmerman
,
L. B.
,
De Jesus-Escobar
,
J. M.
and
Harland
,
R. M.
(
1996
).
The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4
.
Cell
86
,
599
606
.
Zoltewicz
,
J. S.
and
Gerhart
,
J. C.
(
1997
).
The Spemann organizer of Xenopus is patterned along its anteroposterior axis at the earliest gastrula stage
.
Dev. Biol
.
192
,
482
491
.