- Email: [email protected]
Temporal and spatial expression patterns of FoxD2 during the early development of Xenopus laevis
Mechanisms of Development 111 (2002) 181–184 www.elsevier.com/locate/modo
Gene expression pattern
Temporal and spatial expression patterns of FoxD2 during the early development of Xenopus laevis Barbara S. Pohl, Walter Kno¨chel* Abteilung Biochemie, Universita¨t Ulm, Albert-Einstein-Allee 11, D-89081 Ulm, Germany Received 12 September 2001; received in revised form 1 November 2001; accepted 2 November 2001
Abstract We have investigated the sequence and expression pattern of the Xenopus laevis FoxD2 gene, a member of the fork head/winged helix multigene family. The derived protein sequence is most closely related to FoxD2 factors known from other species. Maternal FoxD2 transcripts are degraded during early cleavage stages. Zygotic transcription is activated after the midblastula transition followed by a pronounced increase during neurulation. Whole mount in situ hybridisations reveal that FoxD2 is predominantly expressed in the paraxial mesoderm, but not within the myotome. In addition, FoxD2 transcripts are found within the migrating ventral abdominal muscle precursors, in cranial neural crest cells surrounding the eye and populating the second and third visceral arches as well as within restricted areas of the diencephalon. In hatched tadpoles, FoxD2 expression is also observed within the terminal part of the gut. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Fork head; Winged helix; Transcription factor; Xenopus laevis; Expression pattern; Somites; Abdominal muscles
1. Results and discussion Transcription factors of the fork head/winged helix family are known to serve as important regulators in vertebrate embryogenesis (Kaufmann and Kno¨chel, 1996). We here report on the sequence and expression of the Xenopus FoxD2 gene, which previously had been identified as an incomplete genomic fragment termed XFD-9 (Scheucher et al., 1995). Based upon sequence alignments of winged helix domains of proteins from different species, this gene has recently been classified according to a new nomenclature as FoxD1 (Kaestner et al., 2000). However, a comparison of the full-length protein with Fox genes from other species now clearly suggests that it represents the Xenopus orthologue to FoxD2. Fig. 1A shows the FoxD2 gene sequence including 600 bp of the 5 0 flanking region and the derived amino acid sequence. The transcription start site (11) was determined by 5 0 -RACE. It is preceded by a canonical TATA box 30 nucleotides upstream. A corresponding cDNA, which exhibits over its entire sequence length complete identity with the gene sequence has been isolated from a neurula stage library. Although the isolated cDNA ends at the A-stretch following the stop codon, the * Corresponding author. Tel.: 149-731-502-3280; fax: 149-731-5023277. E-mail address: [email protected] (W. Kno¨chel).
transcribed region might presumably extend over the polyadenylation signal at 11940. The gene we present in this work can clearly be identified as FoxD2 based on amino acid alignments of the derived protein with other members of the FoxD subfamily including those from other species and the evaluation of the phylogenetic tree (Fig. 1B,C). Accordingly, Xenopus FoxD2 shares 70% identity with its chicken ortholog (CWH1; Freyaldenhoven et al., 1997), whereas it exhibits only 55% identity with FoxD3 (Pohl and Kno¨chel, 2001) and 50% identity with Xenopus BF-2 (Mariani and Harland, 1998; Gomez-Skarmeta et al., 1999), the latter representing the true FoxD1 ortholog in this species. This conclusion is evident from a comparison of Xenopus BF-2 with BF-2 factors from other species, among them the human BF-2 (FREAC4; Pierrou et al., 1994) (see also Fig. 1C). Despite the fact that FoxD1, 2 and 3 share extensive sequence homologies not only inside but also outside their fork head domains, they show very different expression patterns during early embryogenesis (Fig. 1D). While FoxD1 expression is mainly restricted to the anterior neural plate and to forebrain structures, FoxD3 is predominantly expressed within the cranial neural crest. In contrast, FoxD2 is found to be transcribed in two paraxial segmented stripes along the trunk, and later, within the somites (see below). The temporal expression pattern of FoxD2 was deter-
0925-4773/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(01)00617-7
182
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
Fig. 1. Sequence of X. laevis FoxD2. (A) Nucleotide and derived amino acid sequences. The sequence has been deposited under EMBL accession number AJ344435. The TATA box, the polyadenylation signal and the fork head/winged helix domain are underlined. (B) Sequence alignment of X. laevis (x) FoxD2 with chicken (g) FoxD2, X. laevis FoxD1 and FoxD3. Identical aa residues are shown in red, deletions are indicated by dashes. (C) Phylogenetic tree for X. laevis FoxD1 (xBF-2; AJ011652), FoxD2, FoxD3 (XFD-6; AJ298865), human FOXD1 (FREAC4; U59831), chicken FoxD2 (CWH1; U37272) and FoxD3 (CWH3; U37274) (DNAstar program using J. Hein method with PAM 250). (D) Comparative whole mount in situ hybridisation for FoxD1, FoxD2 and FoxD3 during (upper; anterior view) or after completion of neurulation (lower; lateral view), respectively.
mined by reverse transcriptase-polymerase chain reaction (RT-PCR) using RNAs of different developmental stages (Fig. 2). Low amounts of maternal transcripts were detected, but are degraded during the early cleavage stages. Zygotic transcription of FoxD2 is activated following midblastula transition, and the amount of transcripts gradually rises throughout development with the strongest increase during neurulation (stages 14–19).
The spatial expression was analysed by whole mount in situ hybridisation. A faint staining is visible (after staining overnight) in gastrula stage embryos with some enrichment at the dorsal side (Fig. 3A). These transcripts are localised within the mesoderm but excluded from epithelial ectoderm (Fig. 3B). However, the dorsal lip itself, which is clearly stained for Xbra expression (see insert) is devoid of FoxD2 transcripts. With the beginning of neurulation (Fig. 3C), a
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
Fig. 2. Temporal expression pattern of FoxD2. RT-PCR was performed with RNA of different developmental stages. The first lane is a negative control lacking RNA, in the second lane mature oocytes were used for RNA isolation and cDNA synthesis. Stages are classified according to Nieuwkoop and Faber (1975).
183
distinct expression is observed within the paraxial mesoderm, lateral to the involuting neural tube. At stage 17, FoxD2 transcripts are concentrated in two segmented stripes located lateral to the dorsal midline (Fig. 3D). These stripes are duplicated within the posterior part of the embryo, but appear to fuse at a posterior pole of more intense staining (Fig. 3D). The expression of FoxD2 is restricted to the dorsal most part of the somitogenic mesoderm (Fig. 3E), which is not involved in formation of the future myotome. Also, the
Fig. 3. Whole mount in situ hybridisation with FoxD2. (A) Stage 11, lateral view. (B) Inner view of an embryo (stage 11) cut into two halves. The insert shows a sagittal section through the dorsal blastopore lip of an embryo after whole mount in situ hybridisation for Xbra and DAPI staining of nuclei. (C) Stage 15. (D) Stage 17. (E) Transverse section of D. (F) Stage 21. (G) Stage 25. (H) Transverse section of G. (I) Stage 27. (J) Stage 29. (K) Transverse section of J. (L) Horizontal section of J. (M) Stage 36. (N) Higher magnification of the lateral region at stage 36. (O) Stage 41. (P) Transverse section of stage 41. All embryos, except for O, are orientated with anterior to the right. (C, D, F, G): dorsal view. (I, J, M, N, O): lateral view. dl: dorsal lip, ey: eye, me: mesencephalon, no: notochord, np: neural plate, nt: neural tube, II. and III.: second and third visceral pouch. Red dotted lines denote the planes of sections.
184
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
mesodermal localisation of these cells at this early stage precludes them to originate from the neural crest. At later stages of development, expression of FoxD2 is found along the somites showing the highest intensities at the most dorsal and most ventral positions (Fig. 3F,G,I). Staining appears to be strongest in the most newly formed somite. Interestingly, a transverse section shows that staining in the ventral part at later stages is no longer restricted to the outer lateral surface of somites (Fig. 3H), a localisation, which would be expected for the future dermatome. Instead, FoxD2 expressing cells appear to migrate ventro/medially as expected for future sclerotome cells. However, it should be noted that the formation of this tissue in Xenopus laevis is histologically not very well understood to date (Keller, 2000). Beginning at stage 25, additional expression is observed in cranial neural crest cells (Fig. 3G,I). FoxD2 transcripts in the cranial neural crest become restricted to narrow stripes migrating on the second and third visceral arches (Fig. 3J,L), but are also visible in crest cells surrounding the eyes (Fig. 3K) and within the ethmoidal and mandibular processes. Additional expression of FoxD2 is found in a diencephalic region between the eyes. Another prominent feature of FoxD2 expression becomes apparent at stage 31, when the precursors of the abdominal muscles of the trunk (Lynch, 1990) are detached from the myotomes and start to migrate in a row of seven spots across the ventrolateral body wall (Fig. 3M, see enlargement Fig. 3N). At the elongation of the tail, we also observe a diffuse staining of the proctodeum. When the gut is curled, this diffuse staining of the proctodeum becomes concentrated exclusively on the left side of the embryo within the terminal part of the gut (Fig. 3O). Expression of FoxD2 in the head demarcates facial mesenchyme and is most prominent at the site where the two halves of the lower jar fuse (Fig. 3P). In summary, through the use of more sensitive methods like RT-PCR and a full-length cDNA to generate probes for in situ hybridisation, we here report on many important additional features of FoxD2 expression during Xenopus embryogenesis which were previously not detected (Scheucher et al., 1995). A comparison of the complete protein sequence with other members of the FoxD subfamily leads to the conclusion that it represents the FoxD2 orthologue. Remarkably, the expression patterns reported for zebrafish fkd9 and the mouse MF-2 (Foxd2) are strikingly similar (Odenthal and Nu¨ sslein-Volhard, 1998; Wu et al., 1998). In both cases, it was concluded that the expression in the developing somites might be correlated with a function in sclerotome formation. Furthermore, transcripts in the diencephalon of both species and in the posterior gut of the zebrafish have also been described. Therefore, besides the evolutionary sequence conservation, the similar expression patterns strongly support the orthologous relationship between all these genes.
2. Materials and methods RT-PCR and whole mount in situ hybridisation was carried out as previously described (Pohl and Kno¨ chel, 2001). FoxD2 was amplified for RT-PCR using the following primers: forward 5 0 -GAGTCCTGTTAACCAAGTGG3 0 ; reverse 5 0 -GGTGTCCTCCGAGAGCAGCG-3 0 .
Acknowledgements We thank Karin Dillinger for the skilful technical assistance in doing whole mount in situ hybridisations. The XBF-2 in situ probe was kindly provided by R. Harland. This investigation was supported by grants from the Deutsche Forschungsgemeinschaft (SFB 497/TP A3) and Fonds der Chemischen Industrie.
References Freyaldenhoven, B.S., Freyaldenhoven, M.P., Iacovoni, J.S., Vogt, P.K., 1997. Aberrant cell growth induced by avian winged helix proteins. Cancer Res. 57, 123–129. Gomez-Skarmeta, J.L., de la Calle-Mustienes, E., Modolell, J., Mayor, R., 1999. Xenopus brain factor-2 controls mesoderm, forebrain and neural crest development. Mech. Dev. 80, 15–27. Kaestner, K.H., Kno¨ chel, W., Martinez, D.E., 2000. Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev. 14, 142–146. Kaufmann, E., Kno¨ chel, W., 1996. Five years on the wings of fork head. Mech. Dev. 57, 3–20. Keller, R., 2000. The origin and morphogenesis of amphibian somites. Curr. Top. Dev. Biol. 47, 183–246. Lynch, K., 1990. Development and innervation of the abdominal muscle in embryonic Xenopus laevis. Am. J. Anat. 187, 374–392. Mariani, F.V., Harland, R.M., 1998. XBF-2 is a transcriptional repressor that converts ectoderm into neural tissue. Development 125, 5019– 5031. Nieuwkoop, P.D., Faber, J., 1975. Normal Table of Xenopus laevis (Daudin), 2nd ed. Elsevier/North-Holland, Amsterdam. Odenthal, J., Nu¨ sslein-Volhard, C., 1998. Fork head domain genes in zebrafish. Dev. Genes Evol. 208, 245–258. Pierrou, S., Hellqvist, M., Samuelsson, L., Enerback, S., Carlsson, P., 1994. Cloning and characterization of seven human forkhead proteins: binding site specificity and DNA bending. EMBO J. 13, 5002–5012. Pohl, B.S., Kno¨ chel, W., 2001. Overexpression of the transcriptional repressor FoxD3 prevents neural crest formation in Xenopus embryos. Mech. Dev. 103, 93–106. Scheucher, M., Dege, P., Lef, J., Hille, S., Kno¨ chel, W., 1995. Transcription patterns of four different fork head/HNF-3 related genes (XFD-4, 6, 9 and 10) in Xenopus laevis embryos. Roux’s Arch. Dev. Biol. 204, 203– 211. Wu, S.C.-Y., Grindley, J., Winnier, G.E., Hargett, L., Hogan, B.L.H., 1998. Mouse mesenchyme forkhead 2 (Mf2): expression, DNA binding and induction by sonic hedgehog during somitogenesis. Mech. Dev. 70, 3– 13.
Gene expression pattern
Temporal and spatial expression patterns of FoxD2 during the early development of Xenopus laevis Barbara S. Pohl, Walter Kno¨chel* Abteilung Biochemie, Universita¨t Ulm, Albert-Einstein-Allee 11, D-89081 Ulm, Germany Received 12 September 2001; received in revised form 1 November 2001; accepted 2 November 2001
Abstract We have investigated the sequence and expression pattern of the Xenopus laevis FoxD2 gene, a member of the fork head/winged helix multigene family. The derived protein sequence is most closely related to FoxD2 factors known from other species. Maternal FoxD2 transcripts are degraded during early cleavage stages. Zygotic transcription is activated after the midblastula transition followed by a pronounced increase during neurulation. Whole mount in situ hybridisations reveal that FoxD2 is predominantly expressed in the paraxial mesoderm, but not within the myotome. In addition, FoxD2 transcripts are found within the migrating ventral abdominal muscle precursors, in cranial neural crest cells surrounding the eye and populating the second and third visceral arches as well as within restricted areas of the diencephalon. In hatched tadpoles, FoxD2 expression is also observed within the terminal part of the gut. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Fork head; Winged helix; Transcription factor; Xenopus laevis; Expression pattern; Somites; Abdominal muscles
1. Results and discussion Transcription factors of the fork head/winged helix family are known to serve as important regulators in vertebrate embryogenesis (Kaufmann and Kno¨chel, 1996). We here report on the sequence and expression of the Xenopus FoxD2 gene, which previously had been identified as an incomplete genomic fragment termed XFD-9 (Scheucher et al., 1995). Based upon sequence alignments of winged helix domains of proteins from different species, this gene has recently been classified according to a new nomenclature as FoxD1 (Kaestner et al., 2000). However, a comparison of the full-length protein with Fox genes from other species now clearly suggests that it represents the Xenopus orthologue to FoxD2. Fig. 1A shows the FoxD2 gene sequence including 600 bp of the 5 0 flanking region and the derived amino acid sequence. The transcription start site (11) was determined by 5 0 -RACE. It is preceded by a canonical TATA box 30 nucleotides upstream. A corresponding cDNA, which exhibits over its entire sequence length complete identity with the gene sequence has been isolated from a neurula stage library. Although the isolated cDNA ends at the A-stretch following the stop codon, the * Corresponding author. Tel.: 149-731-502-3280; fax: 149-731-5023277. E-mail address: [email protected] (W. Kno¨chel).
transcribed region might presumably extend over the polyadenylation signal at 11940. The gene we present in this work can clearly be identified as FoxD2 based on amino acid alignments of the derived protein with other members of the FoxD subfamily including those from other species and the evaluation of the phylogenetic tree (Fig. 1B,C). Accordingly, Xenopus FoxD2 shares 70% identity with its chicken ortholog (CWH1; Freyaldenhoven et al., 1997), whereas it exhibits only 55% identity with FoxD3 (Pohl and Kno¨chel, 2001) and 50% identity with Xenopus BF-2 (Mariani and Harland, 1998; Gomez-Skarmeta et al., 1999), the latter representing the true FoxD1 ortholog in this species. This conclusion is evident from a comparison of Xenopus BF-2 with BF-2 factors from other species, among them the human BF-2 (FREAC4; Pierrou et al., 1994) (see also Fig. 1C). Despite the fact that FoxD1, 2 and 3 share extensive sequence homologies not only inside but also outside their fork head domains, they show very different expression patterns during early embryogenesis (Fig. 1D). While FoxD1 expression is mainly restricted to the anterior neural plate and to forebrain structures, FoxD3 is predominantly expressed within the cranial neural crest. In contrast, FoxD2 is found to be transcribed in two paraxial segmented stripes along the trunk, and later, within the somites (see below). The temporal expression pattern of FoxD2 was deter-
0925-4773/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(01)00617-7
182
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
Fig. 1. Sequence of X. laevis FoxD2. (A) Nucleotide and derived amino acid sequences. The sequence has been deposited under EMBL accession number AJ344435. The TATA box, the polyadenylation signal and the fork head/winged helix domain are underlined. (B) Sequence alignment of X. laevis (x) FoxD2 with chicken (g) FoxD2, X. laevis FoxD1 and FoxD3. Identical aa residues are shown in red, deletions are indicated by dashes. (C) Phylogenetic tree for X. laevis FoxD1 (xBF-2; AJ011652), FoxD2, FoxD3 (XFD-6; AJ298865), human FOXD1 (FREAC4; U59831), chicken FoxD2 (CWH1; U37272) and FoxD3 (CWH3; U37274) (DNAstar program using J. Hein method with PAM 250). (D) Comparative whole mount in situ hybridisation for FoxD1, FoxD2 and FoxD3 during (upper; anterior view) or after completion of neurulation (lower; lateral view), respectively.
mined by reverse transcriptase-polymerase chain reaction (RT-PCR) using RNAs of different developmental stages (Fig. 2). Low amounts of maternal transcripts were detected, but are degraded during the early cleavage stages. Zygotic transcription of FoxD2 is activated following midblastula transition, and the amount of transcripts gradually rises throughout development with the strongest increase during neurulation (stages 14–19).
The spatial expression was analysed by whole mount in situ hybridisation. A faint staining is visible (after staining overnight) in gastrula stage embryos with some enrichment at the dorsal side (Fig. 3A). These transcripts are localised within the mesoderm but excluded from epithelial ectoderm (Fig. 3B). However, the dorsal lip itself, which is clearly stained for Xbra expression (see insert) is devoid of FoxD2 transcripts. With the beginning of neurulation (Fig. 3C), a
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
Fig. 2. Temporal expression pattern of FoxD2. RT-PCR was performed with RNA of different developmental stages. The first lane is a negative control lacking RNA, in the second lane mature oocytes were used for RNA isolation and cDNA synthesis. Stages are classified according to Nieuwkoop and Faber (1975).
183
distinct expression is observed within the paraxial mesoderm, lateral to the involuting neural tube. At stage 17, FoxD2 transcripts are concentrated in two segmented stripes located lateral to the dorsal midline (Fig. 3D). These stripes are duplicated within the posterior part of the embryo, but appear to fuse at a posterior pole of more intense staining (Fig. 3D). The expression of FoxD2 is restricted to the dorsal most part of the somitogenic mesoderm (Fig. 3E), which is not involved in formation of the future myotome. Also, the
Fig. 3. Whole mount in situ hybridisation with FoxD2. (A) Stage 11, lateral view. (B) Inner view of an embryo (stage 11) cut into two halves. The insert shows a sagittal section through the dorsal blastopore lip of an embryo after whole mount in situ hybridisation for Xbra and DAPI staining of nuclei. (C) Stage 15. (D) Stage 17. (E) Transverse section of D. (F) Stage 21. (G) Stage 25. (H) Transverse section of G. (I) Stage 27. (J) Stage 29. (K) Transverse section of J. (L) Horizontal section of J. (M) Stage 36. (N) Higher magnification of the lateral region at stage 36. (O) Stage 41. (P) Transverse section of stage 41. All embryos, except for O, are orientated with anterior to the right. (C, D, F, G): dorsal view. (I, J, M, N, O): lateral view. dl: dorsal lip, ey: eye, me: mesencephalon, no: notochord, np: neural plate, nt: neural tube, II. and III.: second and third visceral pouch. Red dotted lines denote the planes of sections.
184
B.S. Pohl, W. Kno¨ chel / Mechanisms of Development 111 (2002) 181–184
mesodermal localisation of these cells at this early stage precludes them to originate from the neural crest. At later stages of development, expression of FoxD2 is found along the somites showing the highest intensities at the most dorsal and most ventral positions (Fig. 3F,G,I). Staining appears to be strongest in the most newly formed somite. Interestingly, a transverse section shows that staining in the ventral part at later stages is no longer restricted to the outer lateral surface of somites (Fig. 3H), a localisation, which would be expected for the future dermatome. Instead, FoxD2 expressing cells appear to migrate ventro/medially as expected for future sclerotome cells. However, it should be noted that the formation of this tissue in Xenopus laevis is histologically not very well understood to date (Keller, 2000). Beginning at stage 25, additional expression is observed in cranial neural crest cells (Fig. 3G,I). FoxD2 transcripts in the cranial neural crest become restricted to narrow stripes migrating on the second and third visceral arches (Fig. 3J,L), but are also visible in crest cells surrounding the eyes (Fig. 3K) and within the ethmoidal and mandibular processes. Additional expression of FoxD2 is found in a diencephalic region between the eyes. Another prominent feature of FoxD2 expression becomes apparent at stage 31, when the precursors of the abdominal muscles of the trunk (Lynch, 1990) are detached from the myotomes and start to migrate in a row of seven spots across the ventrolateral body wall (Fig. 3M, see enlargement Fig. 3N). At the elongation of the tail, we also observe a diffuse staining of the proctodeum. When the gut is curled, this diffuse staining of the proctodeum becomes concentrated exclusively on the left side of the embryo within the terminal part of the gut (Fig. 3O). Expression of FoxD2 in the head demarcates facial mesenchyme and is most prominent at the site where the two halves of the lower jar fuse (Fig. 3P). In summary, through the use of more sensitive methods like RT-PCR and a full-length cDNA to generate probes for in situ hybridisation, we here report on many important additional features of FoxD2 expression during Xenopus embryogenesis which were previously not detected (Scheucher et al., 1995). A comparison of the complete protein sequence with other members of the FoxD subfamily leads to the conclusion that it represents the FoxD2 orthologue. Remarkably, the expression patterns reported for zebrafish fkd9 and the mouse MF-2 (Foxd2) are strikingly similar (Odenthal and Nu¨ sslein-Volhard, 1998; Wu et al., 1998). In both cases, it was concluded that the expression in the developing somites might be correlated with a function in sclerotome formation. Furthermore, transcripts in the diencephalon of both species and in the posterior gut of the zebrafish have also been described. Therefore, besides the evolutionary sequence conservation, the similar expression patterns strongly support the orthologous relationship between all these genes.
2. Materials and methods RT-PCR and whole mount in situ hybridisation was carried out as previously described (Pohl and Kno¨ chel, 2001). FoxD2 was amplified for RT-PCR using the following primers: forward 5 0 -GAGTCCTGTTAACCAAGTGG3 0 ; reverse 5 0 -GGTGTCCTCCGAGAGCAGCG-3 0 .
Acknowledgements We thank Karin Dillinger for the skilful technical assistance in doing whole mount in situ hybridisations. The XBF-2 in situ probe was kindly provided by R. Harland. This investigation was supported by grants from the Deutsche Forschungsgemeinschaft (SFB 497/TP A3) and Fonds der Chemischen Industrie.
References Freyaldenhoven, B.S., Freyaldenhoven, M.P., Iacovoni, J.S., Vogt, P.K., 1997. Aberrant cell growth induced by avian winged helix proteins. Cancer Res. 57, 123–129. Gomez-Skarmeta, J.L., de la Calle-Mustienes, E., Modolell, J., Mayor, R., 1999. Xenopus brain factor-2 controls mesoderm, forebrain and neural crest development. Mech. Dev. 80, 15–27. Kaestner, K.H., Kno¨ chel, W., Martinez, D.E., 2000. Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev. 14, 142–146. Kaufmann, E., Kno¨ chel, W., 1996. Five years on the wings of fork head. Mech. Dev. 57, 3–20. Keller, R., 2000. The origin and morphogenesis of amphibian somites. Curr. Top. Dev. Biol. 47, 183–246. Lynch, K., 1990. Development and innervation of the abdominal muscle in embryonic Xenopus laevis. Am. J. Anat. 187, 374–392. Mariani, F.V., Harland, R.M., 1998. XBF-2 is a transcriptional repressor that converts ectoderm into neural tissue. Development 125, 5019– 5031. Nieuwkoop, P.D., Faber, J., 1975. Normal Table of Xenopus laevis (Daudin), 2nd ed. Elsevier/North-Holland, Amsterdam. Odenthal, J., Nu¨ sslein-Volhard, C., 1998. Fork head domain genes in zebrafish. Dev. Genes Evol. 208, 245–258. Pierrou, S., Hellqvist, M., Samuelsson, L., Enerback, S., Carlsson, P., 1994. Cloning and characterization of seven human forkhead proteins: binding site specificity and DNA bending. EMBO J. 13, 5002–5012. Pohl, B.S., Kno¨ chel, W., 2001. Overexpression of the transcriptional repressor FoxD3 prevents neural crest formation in Xenopus embryos. Mech. Dev. 103, 93–106. Scheucher, M., Dege, P., Lef, J., Hille, S., Kno¨ chel, W., 1995. Transcription patterns of four different fork head/HNF-3 related genes (XFD-4, 6, 9 and 10) in Xenopus laevis embryos. Roux’s Arch. Dev. Biol. 204, 203– 211. Wu, S.C.-Y., Grindley, J., Winnier, G.E., Hargett, L., Hogan, B.L.H., 1998. Mouse mesenchyme forkhead 2 (Mf2): expression, DNA binding and induction by sonic hedgehog during somitogenesis. Mech. Dev. 70, 3– 13.