Twist Function Is Required for the Morphogenesis of the Cephalic Neural Tube and the Differentiation of the Cranial Neural Crest Cells in the Mouse Embryo

Developmental Biology 247, 251–270 (2002) doi:10.1006/dbio.2002.0699

Twist Function Is Required for the Morphogenesis of the Cephalic Neural Tube and the Differentiation of the Cranial Neural Crest Cells in the Mouse Embryo Kenneth Soo,* Meredith P. O’Rourke,* Poh-Lynn Khoo,* Kirsten A. Steiner,* Nicole Wong,* Richard R. Behringer,† and Patrick P. L. Tam* ,1 *Embryology Unit, Children’s Medical Research Institute, Locked Bag 23, Wentworthville, NSW 2145, Australia; and †Department of Molecular Genetics, MD Anderson Cancer Center, University of Texas, Houston, Texas 77030

Loss of Twist function in the cranial mesenchyme of the mouse embryo causes failure of closure of the cephalic neural tube and malformation of the branchial arches. In the Twist ⴚ/ⴚ embryo, the expression of molecular markers that signify dorsal forebrain tissues is either absent or reduced, but those associated with ventral tissues display expanded domains of expression. Dorsoventral organization of the mid- and hindbrain and the anterior–posterior pattern of the neural tube are not affected. In the Twist ⴚ/ⴚ embryo, neural crest cells stray from the subectodermal migratory path and the late-migrating subpopulation invades the cell-free zone separating streams of cells going to the first and second branchial arches. Cell transplantation studies reveal that Twist activity is required in the cranial mesenchyme for directing the migration of the neural crest cells, as well as in the neural crest cells within the first branchial arch to achieve correct localization. Twist is also required for the proper differentiation of the first arch tissues into bone, muscle, and teeth. © 2002 Elsevier Science (USA) Key Words: Twist; dorsoventral patterning; neural tube; neural crest cells; tissue potency; cell migration.

INTRODUCTION The Twist gene, which encodes a basic helix–loop– helix DNA-binding transcription factor, was originally identified in Drosophila as a recessive-lethal mutation resulting in deficient mesoderm formation (Nusslein-Volhard et al., 1984; Simpson, 1983). A similar requirement of Twist for mesoderm formation has been proposed for other animal species, based on the pattern of expression of Twist orthologues in lancelets (Yasui et al., 1998), leech (Soto et al., 1997), Caenorhabditis elegans (Harfe et al., 1998), Xenopus (Hopwood et al., 1989), mouse (Wolf et al., 1991), and human (Wang et al., 1997). In the mouse embryo, the expression of the Twist gene is detected first at gastrulation in the primitive streak and the adjacent mesoderm (Stoetzel et al., 1995; Fuchtbauer, 1995). At early organogenesis, To whom correspondence should be addressed. Fax: ⫹61 2 9687 2120. E-mail: [email protected]. 1

0012-1606/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

Twist transcripts are expressed in the mesenchyme of the branchial arches, and in the paraxial and lateral mesoderm (Figs. 1A and 1B). However, TWIST protein expression lags behind the appearance of transcript and is first detected at E8.25 in the crest region of the neural folds and the head mesenchyme that is populated by cranial neural crest cells (Gitelman, 1997). Unlike their Drosophila counterparts, Twist ⫺/⫺ mouse embryos survive gastrulation, but failure of fusion of the cranial neural folds becomes apparent by E9.5. Therefore, the timing of expression and the distribution of protein, but not those of mRNA transcripts in the head mesenchyme, correlate well with the craniofacial phenotype of the Twist ⫺/⫺ embryo (Chen and Behringer, 1995). Twist has been implicated to play a role in the differentiation of the mesodermal tissue. A high level of Twist activity enhances the differentiation of somatic muscles but blocks that of other mesodermal tissues in Drosophila (Baylies and Bate, 1996). However, constitutive overexpression of Twist can inhibit muscle differentiation and down-regulate Myf5

251

252

Soo et al.

(Chen and Behringer, 1995). However, a more detailed analysis of the organization of the neural tube has not been carried out. Although the expression of a neural crest cell marker (AP2) indicates that the Twist ⫺/⫺ neural crest cells are found in the branchial arch, a study of Twist ⫺/⫺ 7 wild type chimeras showed that mutant cells are excluded from tissues in the peripheral region of the branchial arch (Chen and Behringer, 1995), where neural crest cells normally reside (Trainor et al., 1994). This finding raises the possibility that there may be defects in the differentiation of the Twist-deficient neural crest cells, which have not been revealed by the preliminary analysis of the mutant phenotype. The aims of the present study were to characterize more fully the tissue patterning in the malformed cephalic neural tube, to seek potential candidates of downstream genes, and to examine the migratory behavior of cranial neural crest cells and the differentiation potency of tissues in the first branchial arch. Our results show that Twist activity is essential for the patterning of the forebrain and the differentiation of the craniofacial mesenchyme.

MATERIALS AND METHODS Transgenic and Mutant Mice

FIG. 1. Twist expression in (A) the head mesenchyme, early branchial arches (ba), somites (sm), lateral mesoderm (lm), and allantoic stalk (al) of an early-somite (E8.5)-stage embryo, and (B) the mesenchyme of the head, face, and branchial arches of a E10.5 embryo. (C) The craniofacial region of a wild type E10.5 embryo showing formation of the major brain parts and the facial primordia (arrow points to the first branchial arch). (D) Twist ⫺/⫺ E10.5 embryo showing exposed cranial neural tissues (*) and the accumulation of blood in the head mesenchyme. The arrow points to the tip of the truncated first branchial arch. Abbreviations: fb, forebrain; mb, midbrain; hb, hindbrain; op, optic vesicle.

activity in mouse C 2C 12 cells in vitro, and suppresses myocyte and myotube differentiation in BLC6 embryonic stem cells (Hebrok et al., 1994, 1997; Rohwedel et al., 1995). In Twist ⫺/⫺ mouse embryos, the anterior neural tube fails to close, craniofacial primordia are malformed, and limb bud development is retarded (Chen and Behringer, 1995). Twist heterozygotes survive to reproductive age, exhibiting malformations of skull and limbs, as well as defects in middle ear formation (Bourgeois et al., 1998), which is reminiscent of Saethre– Chotzen syndrome in humans (El Ghouzzi et al., 1997). Analysis of the impact of Twist-null mutation so far has focused mainly on the developmental stages from gastrulation to about E10.5 (forelimb bud stage). Tissue marker analysis of Twist ⫺/⫺ mouse embryos revealed that, at E9.5, when the cranial neural tube defect can be recognized, the midbrain has been appropriately specified as revealed by En1 expression

Twist mutant mice were generated by replacing the proteincoding region of the gene with a neomycin-resistance cassette that results in a loss-of-function mutation (Chen and Behringer, 1995). Twist ⫹/⫺ mice were interbred to obtain embryos of Twist ⫹/⫹, Twist ⫹/⫺, and Twist ⫺/⫺ genotypes. Twist ⫹/⫺ mice were bred with H253 and B5X transgenic mice to produce Twist heterozygous mice carrying either one or both lacZ and EGFP transgenes. TWX (Twist ⫹/⫺;H253) mice were produced by crossing Twist ⫹/⫺ mice with H253 mice that express the HMGCoA-nls-lacZ transgene [TgN (HMG-lacZ) 253Tan; Tam and Tan, 1992]. Inbreeding of TWX mice gives embryos of the three Twist genotypes whose cells express the HMG-lacZ transgene. TGX (Twist ⫹/⫺;H253;EGFP) mice were produced by mating Twist ⫹/⫺ mice with compound B5X transgenic HMG-lacZ;EGFP mice. These compound transgenic mice express both transgenes widely in the embryonic and extraembryonic tissues. The B5X mice were derived from crosses of H253 mice and B5/EGFP mice which express the ␤-actin-CMV-EGFP transgene [TgN(GFPU)5Nagy; Hadjantonakis et al., 1998]. Inbreeding of TGX mice gives embryos of the three Twist genotypes whose cells express both the HMGlacZ and the EGFP transgene. TWX and TGX embryos provided the transgene-tagged cells for the transplantation experiments. The differentiation of the branchial arch tissue in E9.5–E10.5 embryos of different Twist genotypes was assessed by the expression of a lacZ reporter gene that has been integrated into the Gsc locus (Gsc lacZ; Rivera-Perez et al., 1999) and a randomly integrated RBP-lacZ transgene (Tan, 1991). Gsc lacZ and RBP-lacZ mice were crossed to Twist ⫹/⫺ mice to produce compound transgenic Twist ⫹/⫺; Gsc lacZ and Twist ⫹/⫺; RBP-lacZ mice. Inbreeding of Twist ⫹/⫺;Gsc lacZ or Twist ⫹/⫺;RBP-lacZ, or backcrossing to Twist ⫹/⫺ mice produced embryos of different Twist genotypes, which were used for assessing the impact of the mutation on the expression of Gsc lacZ and RBP-lacZ in the craniofacial tissues. The lacZ activity in these embryos was detected by X-gal histochemistry as outlined later.

© 2002 Elsevier Science (USA). All rights reserved.

FIG. 2. Expression domains of molecular markers revealing appropriate anteroposterior organization of the cranial neural tube. Results of double in situ hybridization for two genes are shown in (A) (Six3, En1 indicated by red arrow), (E) (Hesx1, En1 red arrow), (I) (Fgf8, Hoxb1 red arrow), and J (⫹/⫹: Krox20, white asterisk marks rhombomere 3; Sox10 red asterisks). Others show in situ hybridization of individual genes: (B) BF1, (C), Dmbx1 (arrows point to the diencephalon-midbrain junction), (D) Erbb4 (arrows point to the telencephalon– diencephalon junction), (F) Otx2, (G) En1, (H) Fgf8, (J, ⫺/⫺) Krox20, (K) Erbb4 (top view of hindbrain showing expression in rhombomeres 3 and 5). (A, J) E8.5 embryos; (B, D–F, H, I, K) E9.5 embryos; (C, G) E10.5 embryos. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺. 253

254

Soo et al.

Genotyping the Embryos Twist mutant embryos were genotyped by a multiplex polymerase chain reaction (PCR) amplification of yolk sac DNA using the following Twist-specific primers: 5⬘-ATCTGGCGGCCAGGTACATCG-3⬘, 5⬘-GGCTGTTTTCTATGACCGCT-3⬘, and a neo primer: 5⬘-TTGTGCCCAGTCATAGCCGAAT-3⬘. Cycle parameters were 94°C for 4 min, 1 cycle; 94°C for 1 min, 58°C for 30 s, 72°C for 1 min 30 s, 35 cycles; 72°C for 7 min, 1 cycle. Additional PCR was performed to determine the lacZ genotype in RBP-lacZ; Twist and Gsc lacZ;Twist mutant mice. Primers for the lacZ gene were, 5⬘-GCA TCG AGC TGG GTA ATA AGG GTT GGC CAT-3⬘ and 5⬘-GAC ACC AGA CCA ACT GGT AAT GGT AGC GAC-3⬘. To check the fidelity of the PCR, a housekeeping gene, Rap, was also amplified. Primers for the Rap gene were 5⬘-AGC TTC TCA TTG CTG CGC GCC AGG TTC AGG-3⬘ and 5⬘-AGG ACT GGG TGG CTT CCA ACT CCC AGA CAC-3⬘. Cycle parameters for the lacZ and Rap PCR were 95°C for 5 min, 1 cycle; 94°C for 1 min, 55°C for 30 s, 72°C for 1 min, 32 cycles; and 72°C for 10 min, 1 cycle.

Whole-Mount mRNA in Situ Hybridization The protocol used for in situ hybridization was as described by Wilkinson and Nieto (1993) with the following modifications: the Ampliscribe Kit (Epicentre Technologies) was used in conjunction with Dig-11-UTP (Roche) to synthesize sense and antisense RNA probes, SDS was used instead of CHAPS in the hybridization solution and posthybridization washes, 5⫻ SSC was used in the hybridization solution, and posthybridization washes were carried out at 70°C and excluded formamide. No RNA digestion was performed after hybridization. Genes that were examined in this study are listed in Table 1.

Testing the Developmental Potency of Neural Crest Cells Embryo collection, micromanipulation, and culture. Pregnant TWX and TGX mice were dissected at E8.5 days to obtain 5- to 6-somite-stage embryos. This ensures that the neural crest cells that are studied are among the first population to commence migration from the neural plate at the level of the caudal midbrain and rostral hindbrain (Chan and Tam, 1988; Trainor et al., 1994). From the donor TWX or TGX embryos, a strip of neuroepithelium was dissected from the lateral region of the neural plate by using fine glass needles. This tissue fragment was then broken up into clumps of 5–20 cells for transplantation. Host ARC/s strain embryos at the five- to six-somite stage were used as recipients of cell transplantation. They were dissected free of the decidua and parietal yolk sac, and kept in culture until manipulation. Clumps of donor neural crest cells were grafted to the lateral margin of the caudal midbrain to rostral hindbrain of the host embryo by using a Leitz micromanipulator under a Wild M3Z dissecting microscope (Kinder et al., 2000). After grafting, embryos were cultured for 48 h in vitro (Sturm and Tam, 1993). During the first 24 h, embryos were cultured in DR75 medium in 5% CO 2, 5% O 2, and 90% N 2 atmosphere. During the second 24-h period, the gas phase was changed to 5% CO 2, 20% O 2, and 75% N 2. Detection of lacZ activity. Expression of the lacZ transgene was assessed by X-gal histochemical staining. Embryos from the aforementioned cell transplantation experiment and E9.5–E10.5 embryos isolated from pregnant Twist ⫹/⫺;Gsc lacZ mice and Twist ⫹/⫺;

RBP-lacZ mice were fixed for 5 min in 4% paraformaldehyde (PFA), and then left overnight at 37°C in X-gal staining solution (Trainor et al., 1999). Stained embryos were refixed in 4% PFA, photographed, and processed for wax histology. Serial 8-␮m sections were counterstained with Nuclear Fast Red. Fluorescence microscopy of EGFP-expressing cells. Embryos that contained graft-derived EGFP-expressing cells were examined by fluorescence microscopy (Quinlan et al., 2001). After culturing for 48 h, the embryo was taken out of the culture vessel and placed in a phosphate-buffered medium for fluorescence imaging. The EGFP fluorescence was recorded by using a SPOT 2 digital camera (SciTech, Melbourne, VIC) attached to a Leica MZFLIII fluorescent microscope with a GFPIII filter set (absorption at 495 nm). For each imaging session, a color bright field image and a monochrome high-resolution fluorescent image were taken. EGFP fluorescence was also visualized by using an Olympus Fluoview FV300 confocal microscope using the Ar ⫹ HeNe lasers and GFP filters. The pictures were superimposed to compose an image showing the location of the EGFP-expressing cells in the cranial mesenchyme and the branchial arch. Teratoma induction. The histogenetic potency of mandibular arch tissues was tested by assessing the differentiation of the tissues grown as teratomas under the kidney capsule. E10.5 embryos were harvested from pregnant Twist ⫹/⫺ mice. The mandibular (first) arch and the hyoid (second) arch of embryos with normal head morphology and the first arch of those with open neural tube defects were dissected and trimmed with hypodermic needles. The remaining embryonic fragments were taken for PCR genotyping. Explants of branchial arches were transplanted underneath the kidney capsule of female ARC/s strain adult mice anesthetized by Ketamine (100 mg/kg) and Xylazine (15 mg/kg), following the procedure described by Tam (1993). Tissue fragments from Twist ⫺/⫺ embryos were transplanted to the right kidney and those from Twist ⫹/⫹ or Twist ⫹/⫺ embryos to the left kidney of the same recipient animal. The teratomas were harvested 14 days after transplantation, fixed in Bouin picro formol solution (Gurr product #36087), and processed for wax histology. Serial 8-␮m sections of the specimens were stained with eosin and hematoxylin. Every other 12–15 section in the series was examined under the light and phase contrast microscope, and between 10 and 30 sections were scanned for each teratoma depending on its size. The relative abundance of each type of tissue was scored semiquantitatively by estimating the space occupied by specific tissue types in histological sections of the teratomas.

RESULTS Patterning of the Twist ⴚ/ⴚ Neural Tube The initial formation of the neural plate and the early head folds of the Twist ⫺/⫺ embryo were indistinguishable from those of the wild type. At E8.75–E9.5, although the neural tube closed properly in the trunk region, the closure of the anterior neuropore was noticeably delayed. In the E10.5 mutant embryo, in contrast to the wild type embryo (Fig. 1C), the cephalic neural primordium became everted and formed irregular bulges, and pools of blood accumulated in the head mesenchyme (Fig. 1D). By E11.5, the embryo ceased development and extensive degeneration of the cranial mesenchyme and neuroepithelial tissues was evident.

© 2002 Elsevier Science (USA). All rights reserved.

255

Twist Function in the Mouse Embryo

TABLE 1 The List of Molecular Markers Used and the Number of Wild Type and Twist Mutant Embryos Twist ⫹/⫹ Marker Alx3 Alx4 BF1 Bmp4 Cart1 Cdh11 Dlx5 Dmbx1 En1 Erbb4 Fgf8 Fgf10 Fgfr1 Fgfr2 Gli1 Gli3 Gsc Gsc lacZ Hesx1 Hoxa2 Hoxb1 Krox20 Msx1 Otx2 Pax3 Pax6 Ptch Pitx1 RBP-lacZ Shh Sim1 Six3 Sox9 Sox10 a b

E8.5

E9.5

E10.5

2

3 9 8 2 5 5 3

1

2 3

1 1 4 7

Twist ⫹/⫺

5a 1 ⫹ 6a 6a 4 ⫹ 17 a 3 4

32 a 9a 1 5 5 6 3 9 2

1

2 1

4 (⫹2 E9.0)

7

E8.5

E9.5

Twist ⫺/⫺ E10.5

3

2 3b

E9.5

E10.5

1

3 3 2 1 6 6 1

3

13

8

6 1

1

5a 4 5⫹1 4 5 7 7 6 38 a 5a 1

E8.5

3 1

2 3 5

9 7 ⫹ 3b 3

2

4 3 4 11 4 4

6 7 2 2 2 6

9 5

2

9

5 4 4

4 5

3 6 15 6 8 13 6 2 3

3 3 2

9 2 2 5 3 2 3 6

5 21 4

23

6 3 3

1

3 1

6 (⫹3 E9.0)

13

5 5 5 ⫹ 2b 3 2 8 3 1 11 4 5 2 2b 3 6 3 2 12 4 5 3 11

Samples of mixed Twist ⫹/⫹ and Twist ⫹/⫺ genotypes. Denotes samples of E11.5 embryos.

To test whether the malformed neural tube displayed normal anteroposterior organization, expression of genes that are localized to different regions of the neural tube was examined by whole-mount in situ hybridization. In the brain of the wild type embryo, the expression of Six3, Fgf8 (the anterior neural ridge and ventral forebrain), BF1, Hesx1 (ventral telencephalon), Otx2 (entire forebrain to upper midbrain), Dmbx1 (diencephalon, diencephalon–midbrain junction and midbrain), Erbb4 (diencephalon, midbrain and rhombomeres 3 and 5), En1 (midbrain to mid/hindbrain junction), Fgf8 (midbrain– hindbrain junction), Hoxb1 (rhombomere 4), and Krox20 (rhombomeres 3 and 5) mark specific segments in the anterior–posterior neural axis (Figs.

2A–2K, ⫹/⫹). Most of these markers were expressed in the appropriate anteroposterior order in the malformed neural tube of the E9.5–E10.5 Twist mutant embryos (Figs. 2A–2K, ⫺/⫺), suggesting that loss of Twist does not affect anterior– posterior patterning of the neural tube. There were, however, changes in the relative size of the expression domain of several genes. The En1 (Figs. 2A, 2C, and 2E, ⫺/⫺) expression domain expanded anteriorly, becoming in closer proximity to the Six3- and Hesx1-expressing tissues, while that of BF1 (Fig. 2B ⫺/⫺) expanded posteriorly to the ventral and posterior aspects of the optic evagination. In the exposed neural tissues, Dmbx1 expression was reduced or absent in the tissues anterior to the expression domain that

© 2002 Elsevier Science (USA). All rights reserved.

256

Soo et al.

FIG. 3. (A) Loss of Gli3 expression in the neural plate of the Twist ⫺/⫺ embryo. Gli3 marks the telencephalon in the wild type embryo. (B) Loss of RBP-lacZ expression (revealed by X-gal staining) in the malformed forebrain of the Twist ⫺/⫺ mutant embryo. RBP-lacZ marks the dorsal part of the telencephalon in the wild type embryo. (C) In the Twist ⫺/⫺ embryo (en face view), Pax6 expression is reduced and displaced to the lateral margin of the open neural plate (arrowhead), in contrast to the wild type embryo, where Pax6 is expressed in the dorsal part of the telencephalon and diencephalon, and in the eye primordium. (D) Fgf8 expression that marks the anterior commissure in the floor of the forebrain (white asterisk), is absent from the medial region (black asterisk) of the open neural plate (en face views), but is found in more lateral regions (arrowhead). (E) Dmbx1 expression in the dorsal region of the diencephalon and midbrain (⫹/⫹) is found in the presumptive diencephalon and midbrain regions of the exposed neural plate of the mutant embryo (both embryos show top views of the fore- and midbrain). (F) Pax3 expression is localized to the lateral margin of the Twist ⫺/⫺ neural plate, which corresponds to the Pax3-expressing tissues in the dorsal midbrain of the wild type embryo. (G) Fgf10 expression is found in the ventral neuroepithelium of the diencephalon and upper midbrain in the wild type embryo (lateral view) and in the corresponding medial region in the neural plate of the mutant embryo (top view). (H) In the mutant embryo, Sim1 expression is localized to the medial region of the neural plate, which is equivalent to the ventral midbrain of the wild type embryo. (A–F, H) E10.5 embryos; (G) E9.5 embryo. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺.

resembles the diencephalon–midbrain junction (Fig. 2C, ⫺/⫺). Erbb4 expression, however, still extended to the anatomical restriction of the neural plate that may be equivalent to the border between the diencephalon and telencephalon (Fig. 2D, ⫺/⫺). This may indicate a reduction

or the loss of certain characteristics of prospective diencephalic tissues that are normally intercalated between the domains of BF1 and En1 expression (Figs. 2A and 2B, ⫺/⫺). In the hindbrain of the Twist ⫺/⫺ embryo, expression of Krox20 and Erbb4 was maintained in rhombomeres (r) 3 and

© 2002 Elsevier Science (USA). All rights reserved.

257

Twist Function in the Mouse Embryo

FIG. 4. Expanded or ectopic expression domain of (A–D) Fgfr2, (E–I) Shh, (J–N) Gli1, and (O–S) Ptch that normally mark the ventral tissues of the cranial neural tube and the adjacent mesenchyme in E10.5 wild type embryos. (A, E, J, O) Left-side view of wild type embryos (dashed lines indicate plane of section shown in B, F, K, and P). (C) Oblique-dorsal view, (G, L, Q) left-side view, and (H, M, R) frontal view of E10.5 Twist ⫺/⫺ embryos (dashed line: plane of section of D, I, N and S). Asterisks in (C), (H), (M), and (R) mark the midline of the neural plate. Abbreviations: ba, branchial arch; fl, floor of the brain; ht, heart; ne, neuroepithelium; nt, neural tube; vm, ventral mesenchyme underneath the neural tube. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺.

5 (Figs. 2J and 2K, ⫺/⫺), but in one of four mutant embryos, Erbb4 was expressed only in r5 but was absent in r3 (data not shown). In wild type embryos, tissues in the telencephalon expressed Gli3 (Fig. 3A, ⫹/⫹), and those in the dorsal part of the telencephalon express RBP-lacZ and Pax6 (Figs. 3B and 3C, ⫹/⫹). In the Twist ⫺/⫺ embryo, the neuroepithelium of the anterior neural plate did not display any Gli3 and RBP-lacZ activity (Figs. 3A and 3B, ⫺/⫺). Although Pax6 expression was much reduced, it was localized to tissues in the lateral margin of the open neural plate (Fig. 3C, ⫺/⫺), which corresponds to the prospective dorsal region if the neural tube were closed. Fgf8 expression in tissues in the dorsal region of the diencephalon–telencephalon junction in the forebrain was also missing in the mutant embryo (Fig. 3D, ⫺/⫺). Genes such as Six3, BF1, and Hesx1 that are expressed in the ventral telencephalon of the wild type embryo (Figs. 2A–2C, ⫹/⫹)

were still expressed in tissues in the rostral-most medial (prospective ventral) region of the open anterior neural plate (Figs. 2A–2C, ⫺/⫺). However, Six3 and BF1 were expressed in a slightly expanded domain (Figs. 2A–2C, ⫺/⫺). In the wild type embryo, Dmbx1 and Pax3 were expressed in the dorsal and lateral neuroepithelium of the diencephalon and the midbrain (Figs. 3E and 3F, ⫹/⫹), and Fgf10 and Sim1 were expressed in the ventral neuroepithelium (Figs. 3G and 3H, ⫹/⫹) in the same regions of the brain. The expression of Dmbx1 and Pax3 in the lateral region (Figs. 3E and 3F, ⫺/⫺), and that of Fgf10 and Sim1 in the medial region of the exposed neural plate of the mutant embryo (Figs. 3G and 3H, ⫹/⫹) indicates that these genes were expressed appropriately in the prospective dorsal and ventral domains of the neural primordium. Taken together, results of these marker studies show that dorsoventral patterning of the forebrain is disrupted by the loss of Twist function in the cranial mesenchyme, and

© 2002 Elsevier Science (USA). All rights reserved.

258

Soo et al.

cells in the prospective dorsal telencephalon lose some of their molecular characteristics. The correct expression of Pax3, Fgf8, Hoxb1, Krox20, and Erbb4 in the hindbrain of the Twist ⫺/⫺ embryo (Figs. 2H–2K and 3F) suggests that normal tissue patterning in this region of the neural tube is unaffected. Disruption of dorsoventral patterning was accompanied by changes in the expression pattern of molecular components of FGF and SHH signaling. Fgf8 activity that is normally localized to the anterior commissure in the medial region of the ventral forebrain was displaced to the more lateral regions, leaving a medial region devoid of Fgf8 activity (Fig. 3D, ⫺/⫺). Fgfr1 and -2 expression was absent from the ventral forebrain tissues, and instead, was strongly expressed in the lateral neuroepithelial tissues of the open neural tube (Figs. 4A– 4D; also see Fig. 5D). In the forebrain of the wild type embryo, Shh expression is restricted to the ventral neuroectoderm and adjacent mesendoderm (Figs. 4E and 4F). In Twist ⫺/⫺ embryos, Shh was expressed in a broad medial region of the rostral part of the neural plate (Figs. 4G– 4I). Accompanying the expansion of Shh expression, Gli1 and Ptch, which were weakly expressed in the ventral neural tissues and the adjacent mesenchyme of the wild type forebrain (Figs. 4J, 4K, 4O, and 4P), were up-regulated in the neuroepithelium of the Twist mutant embryos (Figs. 4L, 4M, and 4Q– 4S). Remarkably, the expression domain of Gli1 and Ptch shifted laterally (Figs. 4M and 4R) like that of Fgf8 (Fig. 2D), and seemed to be flanking the enlarged Shh expression domain. In the midbrain of the mutant embryo, Pax3 expression, normally found in the tissues of the roof and alar plates, was found in the lateral (prospective dorsal) region (Fig. 3E), and that of Sim1 and Uncx4.1 was localized in the medial (prospective ventral) region of the neural plate (Fig. 3F; data not shown). The RBP-lacZ transgene, which is expressed in the neuroepithelium of the ventral midbrain of the wild type embryo, was also expressed correctly in the equivalent parts of the neural tube of mutant embryo (Fig. 3A). However, expression of Fgfr1 (Fig. 5D) and Fgfr2 (Figs. 4C and 4D) was localized ectopically in the lateral region of the neural plate, and that of Gli1 (Figs. 4L and 4M) and Ptch (Figs. 4Q– 4S) were also shifted laterally away from the ventral region of the midbrain. In the hindbrain of the mutant embryo, Pax3 and RBP-lacZ were expressed in the dorsal and ventral part of the neural tube as in the wild type embryo (Figs. 3A and 3E). In addition, spatially correct expression of Dmbx1, Pax3, and Shh was found in the midbrain (Figs. 3E, 3F, ⫺/⫺, and 4G), as was the expression of Pax3, Shh, Fgf8, Hoxb1, Krox20, and Erbb4 in the hindbrain of the Twist ⫺/⫺ embryo (Figs. 2D, 2H–2J, 3F, and 4G). The findings of marker analysis thus strongly suggest that normal tissue patterning can be achieved in the midbrain and hindbrain in the absence of Twist. The early lethality of the Twist ⫺/⫺ embryo does not permit examination of markers of anterior–posterior and

dorsoventral patterning in more advanced stages of development. Nevertheless, our results suggest that the loss of Twist function in the head mesenchyme impacts on the dorsoventral, but not the anteroposterior patterning of the cephalic neural tube.

Disrupted Development of the Branchial Arch In the Twist mutant embryo, the failure of neural tube closure is accompanied by the malformation of frontonasal primordia, and stalling in the growth of the branchial arches following initial outgrowth. Development of the maxillary and mandibular processes of the first branchial arch were arrested at the formative stage, and thus, the contralateral arches were unable to fuse to form the rudiment of the upper and lower jaws, even as late as E10.5–E11.5. Bmp4 activity, which is expressed during the growth of maxillary and mandibular components of the first branchial arch, was expressed properly in the maxillary process of the Twist ⫺/⫺ embryo but was markedly reduced in the mutant mandibular process (Fig. 5A). Fgf8 activity was correctly expressed in the Twist ⫺/⫺ maxillary process and the proximal portion of the mandibular process, but was absent from the distal portion (Fig. 5B). Fgf10, Fgfr1, and Dlx5 expression was reduced in the Twist ⫺/⫺ branchial arches (Figs. 5C–5E). Alx3, Alx4, and Cart1 activity in the frontonasal mesenchyme and the first branchial arch (Figs. 5F–5H) was repressed in mutant embryos. Therefore, our findings show that abnormal morphogenesis of the branchial arches and frontonasal structures is accompanied by alteration of BMP and FGF signaling activity and down-regulation of Cart1, Alx3, and Alx4 of the family of Aristaless-like transcription factors.

Loss of Twist Function Affects the Migration of Neural Crest Cells In wild type embryos, Sox10 activity follows closely the formation, migration, and differentiation of the cranial neural crest cells (Britsch et al., 2001). Sox10 was first expressed in the nascent neural crest cells as they migrate from the midbrain at E8.5 (Fig. 6A, ⫹/⫹, 4s), and is progressively found in the migratory neural crest cells originating from the forebrain, midbrain, and upper hindbrain and those from the postotic rhombomeres (Fig. 6A, ⫹/⫹, 10s to 17s). Sox10 expression was down-regulated in the neural crest cells in the frontonasal and maxillary region, but was maintained in the neural crest cells streaming from the lower midbrain and rhombomeres 1–2 to the first branchial arch, and those from rhombomeres 4 and 5 to the second branchial arch (Fig. 6A, ⫹/⫹, 12s to 17s). These two streams of neural crest cells remained separate from one another, leaving a discrete Sox10-free zone (Fig. 6A, ⫹/⫹, 10s to 22s) lateral to rhombomere 3 (Fig. 2H). During subsequent development, Sox10 expression was consolidated to the progenitors of the cranial nerve ganglion (Fig. 6A, ⫹/⫹, 27s to 32s).

© 2002 Elsevier Science (USA). All rights reserved.

259

Twist Function in the Mouse Embryo

In the Twist ⫺/⫺ mutant embryo, Sox10-expressing cells emerged on time and at the correct sites from the neural plate, and were distributed to the facial tissue and branchial arches in a similar manner to the neural crest cells of the wild type (Fig. 6A, ⫺/⫺, 5s to 14s). However, the latemigrating populations of neural crest cells heading for the first and second branchial arches (Fig. 6A, ⫺/⫺, 5s to 8s) merged as they streamed toward the branchial arches (Fig. 6A, ⫺/⫺, 17s to 30s). The progressive loss of the Sox10-free zone is unlikely to be caused by either the absence of rhombomere 3 (Fig. 2H) or by the ectopic production of neural crest cells from rhombomere 3 (Fig. 6A, ⫺/⫺, 11s to 17s), but probably due to intrusion of the neural crest cells into the normally cell-free zone. Histological analysis of the lower midbrain/upper hindbrain region of 17- to 22-somite-stage mutant embryos revealed that the migrating neural crest cells were not restricted to the subectodermal mesenchyme as in the wild type embryo (Fig. 6B) but have invaded the deeper paraxial mesenchyme (Fig. 6C). In the mandibular component of the first arch of the wild type embryo, Sox10-expressing cells intermingled with nonexpressing cells throughout the arch with a slight tendency to localize in the lateral subectodermal region (Fig. 6D). Expression of Sox10 in the branchial arch tissue was weaker and localized to cells in the lateral aspect of the arch of the mutant embryo (Fig. 6E). The merging of streams of neural crest cells destined for the first two branchial arches and their invasion into paraxial mesenchyme raise the possibility that the loss of Twist function affects the migratory behavior of neural crest cells. This hypothesis was tested by cell transplantation experiments. In the first set of experiments, hindbrain neural crest cells of TGX or TWX embryos were transplanted orthotopically to ARC/s hosts. Twist ⫹/⫹ neural crest cells were distributed along a discrete track from the site of transplantation in the upper hindbrain to the maxillary and mandibular components of the first branchial arch (Fig. 7A). In the mandibular process, graft-derived cells were dispersed in the mesenchyme in the lateral half of the arch throughout its proximal– distal length (Fig. 7A). Most Twist ⫹/⫺ neural crest cells were distributed like Twist ⫹/⫹ cells in the craniofacial tissues of wild type host embryos (Table 2). In the first branchial arch, Twist ⫹/⫺ cells tended to congregate more often in the core of the arch (Fig. 7B; Table 2). Twist ⫺/⫺ neural crest cells were scattered widely in the paraxial mesenchyme outside the branchial arches (Table 2). Twist ⫺/⫺ cells that have entered the maxillary and mandibular components of the first branchial arch clumped together in the mesenchymal core of the arch (Fig. 7C; Table 2). In contrast, mutant (Twist ⫹/⫺ or Twist ⫺/⫺) neural crest cells were distributed normally in the second arch of the wild type host (data not shown). The loss of Twist function therefore appears to affect the ability of neural crest cells to follow the correct migratory path and to be properly regionalized within the first branchial arch. The impaired homing ability displayed by the mutant neural crest cells was not entirely due to the cellautonomous effect of the loss of Twist function. When wild

type neural crest cells were transplanted to the Twist ⫺/⫺ embryo, the graft-derived cells scattered widely in the cranial mesenchyme, although some found their way to the branchial arch (Fig. 7D). Interestingly, in the abnormal branchial arch, the graft-derived cells scattered in the mesenchyme and, unlike the mutant cells in the wild type arch (Fig. 7D), were not confined to the core of the arch. These findings suggest that Twist is required in the paraxial mesoderm for proper guidance of neural crest cell migration, in addition to the requirement in the neural crest cells to follow the migratory path correctly. That wild type neural crest cells distribute properly in the Twist ⫺/⫺ arch suggests that Twist function is not required in the arch mesenchyme but acts cell-autonomously in the neural crest cells for their regionalization in the branchial arch. Expression of Twist in the craniofacial mesenchyme (Fig. 1B) overlaps significantly with that of Cdh11 (Fig. 7E), which is reputed to mediate cell– cell adhesion. Cdh11 expression was markedly reduced in the Twist ⫺/⫺ embryo, suggesting that alterations in cell adhesion properties of the mesenchyme may contribute to defective cell migration. However, despite the aberrant pattern of cell migration, neural crest cells that colonized the first and second branchial arches were able to adopt the proper molecular identities, as revealed by the appropriate expression of Pitx1 and Msx1 in the maxillary and mandibular components of the first arch (Figs. 7F and 7G), and Hoxa2 in the second (hyoid) arch (Fig. 7H).

Defective Differentiation of the Neural Crest Cells In the craniofacial tissues of the wild type embryo, Sox10 activity is down-regulated progressively in the postmigratory neural crest cells that populate the mesenchyme surrounding the telencephalon and adjacent to the optic and nasal primordia (Fig. 6A, ⫹/⫹,12s–7s), and in the branchial arches (Fig. 6A, ⫹/⫹, 20s–27s). Sox10 expression was later confined to the precursors of the cranial nerve ganglia (Fig. 6A, ⫹/⫹, 32s). In contrast to their wild type counterparts, Twist ⫺/⫺ neural crest cells in the periocular, maxillary, and mandibular primordia did not down-regulate Sox10 even by E10.5 (Fig. 6A, ⫺/⫺, 22s–30s). The lack of down-regulation of Sox10 activity in the neural crest cell population in the branchial arches suggests that the cells may be arrested at an early phase of differentiation. The expression of differentiation markers was studied to find out whether the differentiation of the branchial arch tissues is affected by the loss of Twist function. The initial phase of osteogenic differentiation of neural crest cells was assessed by Gsc activity. In the wild type E9.5 (n ⫽ 25) and E10.5 (n ⫽ 20) embryo, Gsc lacZ was expressed in the distal region of the first branchial arch (the prospective mandibular rami and symphysis) and the second branchial arch around the first branchial groove (the prospective external auditory meatus) (Fig. 8A). Expression of the LacZ transgene was completely absent from the branchial arches in four E9.5 and eight E10.5 Gsc lacZ;Twist ⫺/⫺ embryos (Fig. 8A). In other E9.5 (n ⫽ 5) and E10.5 (n ⫽ 3) Gsc lacZ;Twist ⫺/⫺

© 2002 Elsevier Science (USA). All rights reserved.

260

Soo et al.

FIG. 5. Altered genetic activity in the craniofacial mesenchyme of E10.5 Twist ⫺/⫺ embryos. (A) Expression of Bmp4 is reduced in the foreshortened mandibular process but is maintained in the maxillary process of the mutant embryo. (B) Expression of Fgf8 in the nasal epithelium, the maxillary process, and in tissues at the anterior margin of the proximal part of the mandibular process persists, but is absent in the distal part. Arrowhead points to the diencephalon–telencephalon junction. (C) Fgf10 expression is markedly reduced in the branchial arches of the mutant embryo. (D) Loss of Fgfr1 expression in the branchial arches and facial mesenchyme in the mutant embryo. (E) Properly localized but diminished activity of Dlx5 in the first and the second arch of the mutant embryo. (F–H) Down-regulation of the expression of (F) Alx3 and (G) Alx4 in the frontonasal mesenchyme and the branchial arches, and (H) Cart1 in the periocular and facial mesenchyme. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺ E10.5 embryos.

embryos, Gsc lacZ was expressed at a much-reduced level (data not shown). However, no Gsc expression was detected by in situ hybridization in the branchial arch tissue of E10.5 Twist ⫺/⫺ embryos (Fig. 8B), suggesting that the mutant neural crest cells that previously expressed Gsc (and were therefore lacZ-positive) were unable to maintain Gsc transcription by E10.5. Sox9 activity, which heralds chondrogenic differentiation, was not expressed in the branchial arch of the mutant embryo (Fig. 8C). The loss of Gsc and

Sox9 activity indicates that the Twist ⫺/⫺ mutant arch tissues may have defects in skeletogenic differentiation. In 10 Twist ⫹/⫹ and Twist ⫹/⫺ embryos, RBP-lacZ activity was detected in the myogenic precursors derived from the cranial mesoderm and the somites. Expression of the RBPlacZ transgene was completely absent in 9 Twist ⫺/⫺ mutant embryos (Fig. 8D), but was apparently normal in the truncated branchial arch and the occipital somites of 2 other Twist ⫺/⫺ embryos (more Twist ⫺/⫺ embryos than Twist ⫹/⫺ or

© 2002 Elsevier Science (USA). All rights reserved.

261

Twist Function in the Mouse Embryo

FIG. 6. (A) Sox10 activity in the cranial neural crest cells during their migration to the branchial arches (arrowheads) and differentiation to progenitors of cranial nerve ganglia in E8.5–E10.5 embryos. Embryos are staged by the somite (s) number. Sox10 is also expressed in the otic placode and subsequently the otic vesicle. Twist ⫺/⫺ embryos show a normal pattern of formation and migration of the neural crest cells in specimens collected at 5s to 17s stages. Subsequently (at 17s–30s), the two streams of the Sox10-expressing hindbrain neural crest cells merge (yellow asterisks). In contrast to progressive down-regulation of Sox10 in the branchial arch (frontier marked by the arrowhead) of the wild type embryo (at 22s to 32s), Sox10 activity persists in the neural crest cells that populate the branchial arches (frontier marked by the arrowhead) of the Twist ⫺/⫺ embryo (at 22s–30s). Sox10 expression also persists in the mesenchyme adjacent to the forebrain (white asterisk) and in the tissue ventral to the telencephalon of the Twist ⫺/⫺ embryo, whereas in the wild type embryos, Sox10 activity is progressively restricted to the neurogenic derivatives of the cranial neural crest cells. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺. (B–E) Histological sections of 17- to 22-somite-stage embryos showing the distribution of Sox10-expressing cells in (B) the subectodermal site in the cranial mesenchyme of the wild type embryo, and (C) the intrusion of the neural crest cells into the paraxial mesenchyme in the Twist ⫺/⫺ embryo. (D, E) Transverse sections of the mandibular component of the first branchial arch showing more Sox10-expressing cells in (D) the wild type than in (E) the Twist ⫺/⫺ embryo. Sox10-expressing cells are distributed more to the lateral region of the mandibular process in the Twist ⫺/⫺ embryo. Abbreviations: ne, neuroepithelium; ec, surface ectoderm; M 7 L, mediolateral axis.

Twist ⫹/⫹embryos did not express the transgene at P ⬍ 0.05 by ␹-squared test). Loss of Twist function therefore also affects the myogenic potency of the mesoderm in the branchial arch.

A full in vivo assessment of the developmental potential of Twist ⫺/⫺ branchial arch tissues is hampered by embryonic lethality at E10.5–E11.5. To overcome this limitation, the histogenetic potential of branchial arch tissues was

© 2002 Elsevier Science (USA). All rights reserved.

262

Soo et al.

TABLE 2 Distribution of Graft-Derived Cells in the Craniofacial Tissues of Host Embryos after Orthotopic Transplantation of Neural Crest Cells to the Upper Hindbrain Types of tissues containing graft-derived cells Arch 1: Maxillary Transplantation

a

Arch 1: Mandibular a

Paraxial mesoderm b

spread

core

spread

core

Arch 2

forebrain

midbrain

hindbrain

14.2 38

0 0

15.0 62

0 0

28.4 77

2.1 15

10.0 23

31.3 73

11.5 57

0 0

27.5 70

0 0

24.3 66

4.1 23

8.6 27

24.8 53

2.0 8

10.4 53

5.8 18

15.4 58

26.6 76

7.8 26

2.8 13

30.0 68

⫹/⫹

Twist cells into wild type host (n ⫽ 30) % of cells c % of host embryos d Twist ⫹/⫺ cells into wild type host (n ⫽ 38) % of cells % of host embryos Twist ⫺/⫺ cells into wild type host (n ⫽ 26) % of cells % of host embryos

Note. Bold type highlights differences in pattern of distribution of Twist ⫺/⫺ mutant cells: sequestration to the core of the first branchial arch and spreading to forebrain mesenchyme. a Maxillary and mandibular components of the first branchial arch: spread, widely distributed in the mesenchyme (lateral and core) of the arch; core, tightly clumped together in the core of the mesenchyme. b Cranial paraxial mesenchyme colonized by the transplanted cells listed in the anterior to posterior order: frontonasal (forebrain), facial (midbrain), and occipital– cervical (hindbrain). c % of total cell population for patterns of cell distribution (branchial arches) or site in the cranial mesenchyme (paraxial mesoderm). d % of host embryos showing different patterns of cell distribution (branchial arches) or site in the cranial mesenchyme containing graft-derived cells (paraxial mesoderm).

tested by analyzing the composition of tissues of teratomas formed by the branchial arches (Figs. 9A–9D; Table 3). The teratomas derived from the Twist ⫺/⫺ first branchial arch were consistently smaller than those from wild type arches. They contained less cartilage (Fig. 9B) and only 3 of 8 teratomas contained a minute amount of trabecular bone (Table 3). Most teratomas derived from the Twist ⫺/⫺ mutant arch tissues did not contain any skeletal muscle (Table 3), and those few that did had mainly myoblasts with few mature muscle fibers. Only one teratoma derived from a Twist ⫺/⫺ arch contained a tooth bud, and this resembled a molar (Fig. 9D), in contrast to the more frequent formation of both incisor- and molar-like tooth buds in the Twist ⫹/⫹ and Twist ⫹/⫺ arch-derived teratomas (Fig. 9C; Table 3). The composition of tissues present in the teratoma derived from the Twist ⫺/⫺ first arch was different from that of the second branchial arch of Twist ⫹/⫹ or Twist ⫹/⫺ embryos (Table 3). This suggests that the change in potency of the mutant first arch tissues is due to the impaired capacity to differentiate and not to the transformation of the first arch to other branchial arches. This is consistent with the normal archspecific expression of Msx1 and Hoxa2 (Figs. 7G and 7H). The outcome of the teratoma study reaffirms the reduced tissue potency of the Twist ⫺/⫺ branchial arch that was revealed by the marker study. It further highlights the impact of the Twist mutation on odontogenic and osteogenic differentiation, which may be connected to a functional deficiency of the neural crest cells.

DISCUSSION Twist Is Involved in Dorsoventral Patterning of the Forebrain Our study has shown that, in the exencephalic brain of Twist ⫺/⫺ embryos, the expression of molecular markers for tissues in the dorsal part of the telencephalon and diencephalon (RBP-lacZ, Pax6, Dmbx1, and Erbb4) is either absent or reduced. The expression domain of some genes that encode signaling activity in the ventral tissues, such as Shh, Ptch, Gli1, Fgfr1, Fgfr2, and Fgf8, expands from the medial (prospective ventral) region of the malformed neural tube to cover a wider area of the prospective forebrain and upper midbrain. However, other genes such as Fgf10, Sim1, and Jag1 (data not shown), which are expressed in the ventral neuroectoderm of the diencephalon and upper midbrain, are expressed normally in the Twist ⫺/⫺ neural tube. The identity of the cells that have lost the dorsal characteristics is presently not known, but we speculate that they become cells that are intermediate between dorsal and ventral types. These findings suggest that dorsoventral patterning of cell types in the forebrain may be disrupted by the loss of Twist activity, and the tissues are ventralized generally. In the midbrain and hindbrain of the mutant embryo, dorsal tissue markers (such as Dmbx1, Erbb4, and Pax3) and ventral markers (RBP-lacZ and Shh) are expressed in the lateral margin and medial region of the neural primordium, respectively. These results suggest that

© 2002 Elsevier Science (USA). All rights reserved.

263

Twist Function in the Mouse Embryo

dorsoventral patterning is normal in the midbrain and the hindbrain. During normal development of the neural tube, SHH signals emanating from the notochord and the floor plate, and Wnt and TGF␤ activity from the dorsal neural tube and the adjacent surface ectoderm are essential for the specification of neuronal cell types in the dorsoventral plane of the neural tube (Millonig et al., 2000; Jessell, 2000; Litingtung and Chiang, 2000; Briscoe et al., 2000) and the induction of neural crest cell formation in the dorsal neural tube (Basler et al., 1993; Lee et al., 2000; Barembaum et al., 2000; Liu and Jessell, 1998). In the Twist ⫺/⫺ forebrain, the expansion of the expression domain of Shh and downstream signaling components, plus the loss of antagonistic Gli3 activity suggests that there may be enhanced SHH signaling activity across the lateral–medial (prospective dorsoventral) plane. This in turn may lead to the ventralization of the neural primordium at the expense of the dorsal tissues. Disruption to SHH signaling has been shown to have a similar impact on dorsoventral patterning of the neural tube in several other mutants. Loss of Shh function leads to the absence of ventral tissue types in the brain, but this can be partially ameliorated by the simultaneous loss of Gli3 activity in Shh ⫺/⫺; Gli3 ⫺/⫺ mice (Litingtung and Chiang, 2000). Loss of Ptch activity leads to ventralization of the neural tube (Goodrich et al., 1997), while conversely, overexpressing human Ptch in mice results in partial dorsalization of the neural tube and the inhibition of endogenous SHH target genes, Gli1 and Ptch (Goodrich et al., 1999). In the Extra toes mice, loss of Gli3 function leads to absence of the dorsal diencephalic–telencephalic junction, abnormal neocortex, and down-regulation of Emx genes in the dorsal telencephalon (Theil et al., 1999). This may be caused by either the loss of dorsalizing activity in the roof plate or the lack of repression of the ventralizing SHH activity due to the lack of Gli3 activity and the defective interaction with other Gli genes (Sasaki et al., 1999). However, unlike findings in the Twist mutant, the Extra toes mutant does not display any change in the Shh expression domain in the ventral forebrain (Theil et al., 1999; Tole et al., 2000), suggesting that Gli3 activity can act independently of SHH signaling to promote dorsal but repress ventral tissue identity in the telencephalon. The down-regulation of Gli3 gene and the concomitant expanded expression of Shh, Gli1, and Ptch in the mutant neural tissue suggests that Twist may impact not only the Gli3 function in dorsal specification but also the SHH signaling by enhancing the ventralizing activity. Analysis of the development of the neural tube in chimaeras containing wild type and Twist ⫺/⫺ cells has shown that the defective morphogenesis of the neural tube is primarily due to the loss of Twist in the head mesenchyme (Chen and Behringer, 1995). In Xenopus, the lateral mesoderm and paraxial mesoderm can induce the formation of neural crest cells (a dorsal cell type) from the neural tube, although such experiment has not eliminated the influence of the adjacent epidermal ectoderm (Bonstein et al., 1998;

Mayor et al., 1995). Twist may therefore influence dorsoventral patterning of the forebrain via the inductive action of the cranial mesoderm.

Twist Is Required in Both the Neural Crest Cells and the Head Mesoderm to Direct Neural Crest Cell Migration During the formative phase of neural crest cell ontogeny, the expression pattern of Sox10 seems to faithfully reflect the formation and migration of the neural crest cells (Southard-Smith et al., 1998; Herbarth et al., 1998). On the basis of Sox10 expression, it is evident that neural crest cells are formed normally at all levels of the brain in the Twist ⫺/⫺ embryos, despite the failure of closure of the neural tube. Analysis of region-specific markers reveals that major brain segments in the anterior–posterior axis are present, and with the exception of the forebrain, dorsoventral tissue pattern are established properly. The loss of dorsal tissue characteristics in the forebrain was detected after neural crest cell formation, raising the possibility that the specification of dorsal tissues is disrupted after the forebrain neural crest cells have left the neural plate. The deficiency of dorsal tissue may be due to the inability to repopulate the dorsal forebrain subsequent to the departure of the neural crest cells by cell proliferation and tissue movement, which are critical for the morphogenesis of the forebrain (Jacobson and Tam, 1982; Tuckett and MorrissKay, 1985). In marked contrast to the separation of streams of Sox10expressing cells extending from the rhombomeres to the branchial arches in the wild type embryo, Sox10-expressing cells in the Twist ⫺/⫺ mutant embryo fail to remain segregated, especially those of the late-migrating population adjacent to the hindbrain of E9.5–E10.5 embryos. The most noticeable consequence is the merging of the neural crest cells emerging from rhombomeres 1/2 and 4 into the region lateral to rhombomere 3, which is normally free of crest cells (Farlie et al., 1999; Trainor and Krumlauf, 2000), and the intrusion of the Sox10-expressing cells into the paraxial mesenchyme. Similar defective migration of hindbrain neural crest cells is found in mutant mice with disrupted Neuregulin–ErbB signaling. Loss of Neuregulin function arrests cell migration and neural crest cells accumulate in the dorsal mesenchyme (Britsch et al., 1998). In Erbb4deficient embryos, misguided cell migration leads to the digression of the neural crest cells into the paraxial mesenchyme adjacent to rhombomere 3, fusion of the Vth and VII/VIII cranial nerves, and displacement of the ganglia complex (Gassmann et al., 1995). The cell transplantation experiment has further revealed that most of the Twist ⫺/⫺ cells transplanted to the hindbrain of the wild type host are able to home in to the branchial arch. However, upon entry to the branchial arch, the mutant cells are excluded from the lateral subectodermal mesenchyme of the mandibular primordium that is normally populated by the neural crest cells (LeDouarin, 1982;

© 2002 Elsevier Science (USA). All rights reserved.

264

Soo et al.

FIG. 7. The impact of Twist mutation on the distribution and identity of hindbrain neural crest cells destined for the branchial arch. Graft-derived cells in the host embryo are revealed by X-gal staining reaction that produces a blue coloration. The genotype combination for each transplantation is shown as graft 3 host. (A) Twist ⫹/⫹ neural crest cells migrate from the hindbrain to the arch and are dispersed in the mesenchyme in the lateral (lat) part of the mandibular process (the histology shows a transverse section through arch I and II) of the wild type host embryo. (B) Twist ⫹/⫺ neural crest cells migrate like Twist ⫹/⫹ cells to the mandibular arch of the wild type host and are dispersed along the long axis of the mandibular arch, but with a tendency to congregate in the core of the arch. (C) Twist ⫺/⫺ neural crest cells are confined to the core of the branchial arch of the wild type host and fail to integrate into the mesenchyme of the lateral part of the arch. (D) Wild type neural crest cells (revealed by lacZ and EGFP expression) migrate to the foreshortened first branchial arch and distribute widely in the head mesenchyme of the Twist ⫺/⫺ host embryo. For (A–C), the dashed line in the left-hand side figure indicates the plane of the corresponding histological section. (E) Down-regulation of Cdh11 expression in the branchial arch and facial mesenchyme of the mutant embryo. Reduced expression of (F) Pitx1 in the first and second arch, and (G) Msx1 in the frontonasal and first arch mesenchyme of Twist ⫺/⫺ embryos. (H) Hoxa2 expression in the hindbrain and second branchial arch (white asterisk) remains unaffected in Twist ⫺/⫺ embryos. (A–D) Embryos harvested on E8.5 and cultured for 24 h. (E–H) E10.5 embryos. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺.

Trainor et al., 1994; Trainor and Tam, 1995; Hacker and Guthrie, 1998). Instead, the graft-derived mutant cells are constrained within the core of the arch. The sequestration

of Twist ⫺/⫺ cells in the arch is reminiscent of that seen in the chimeras, where descendants of Twist-deficient ES cells are also excluded from the lateral mesenchyme of the

© 2002 Elsevier Science (USA). All rights reserved.

265

Twist Function in the Mouse Embryo

FIG. 8. Mesenchyme in the branchial arch, and the facial region of Twist ⫺/⫺ embryo shows (A) diminished Gsc lacZ activity, (B) absent Gsc expression, and (C) down-regulation of Sox9 expression. (D) Expression of RBP-lacZ transgene activity in the presumptive myogenic cells of the paraxial mesoderm (pm) and the branchial arch (ba) is lost in the mutant embryo. (A–D) E10.5 embryo. ⫹/⫹, wild type; ⫺/⫺, Twist ⫺/⫺. FIG. 9. Differentiation of Twist ⫹/⫹ or Twist ⫹/⫺ first branchial arch tissue into (A) bone (bn) and cartilage (ca) and (C) tooth buds of molar and incisor (inset) morphology. Twist ⫺/⫺ branchial arch tissues form (B) cartilage and (D) an isolated molar-like tooth bud (tb). Bar, 100 ␮m.

mandibular arch (Chen and Behringer, 1995). Twist activity therefore is also required in a cell-autonomous manner for the proper regionalization of the neural crest cells in the branchial arch, which may have an important impact on

subsequent differentiation into bone, tooth, and connective tissues. When wild type neural crest cells were transplanted to the hindbrain of the mutant embryo, neural crest cells

© 2002 Elsevier Science (USA). All rights reserved.

266

Soo et al.

TABLE 3 Histogenetic Potency of First and Second Branchial Arches of Normal (Twist ⫹/⫹ or Twist ⫹/⫺) Embryos and the First Branchial Arch of the Twist ⫺/⫺ Mutant E10.5 Embryos Tissue type Ectodermal Stratified epithelium Keratinized epithelium Hair follicles Pigment cells Mesodermal Bone Cartilage Muscle fibers Myoblast Adipose tissues Connective tissues Areolar Dense (fibrous) Endodermal Foregut epithelium Ciliated epithelium Glandular epithelium Organotypic differentiation Tooth Dental lamina/bud Cap/Bell stage Salivary gland Lymphoid nodules Epithelial cyst

Normal First branchial arch (16 specimens)

Normal Second branchial arch (6 specimens)

Twist ⫺/⫺ First branchial arch (8 specimens)

⫹ (8) ⫹⫹⫹ (15) ⫹⫹ (6)

⫹⫹ (5) ⫹⫹⫹ (6) ⫹⫹ (3)

⫹ (2) ⫹⫹ (8)

⫹⫹⫹ (15) ⫹⫹ (15) ⫹⫹⫹ (8) ⫹ (1) ⫹⫹ (5)

⫹ (2) ⫹⫹⫹ (6) ⫹⫹⫹ (6)

⫹ (3) ⫹ (7) ⫹⫹ (2) ⫹ (4)

⫹ (4)

⫹ (1)

⫹⫹ (8) ⫹ (1)

⫹⫹ (5) ⫹ (3)

⫹ (4) ⫹ (1)

⫹⫹ (4)

⫹⫹ (3) ⫹ (2) ⫹ (2)

⫹ (2) ⫹ (3) ⫹ (2)

⫹ (3) ⫹⫹ (8) ⫹⫹ (2) ⫹ (1) ⫹ (6)

⫹ (1) ⫹⫹ (2) ⫹ (1)

Note. Relative amount of tissue types: ⫹, rare; ⫹⫹, moderate; ⫹⫹⫹, abundant. (n), number of teratomas containing each type of tissue.

migrated not only to the first branchial arch, but were also scattered widely in the head mesenchyme. The wandering behavior of the wild type neural crest cells in the Twistdeficient cranial mesenchyme and the merging of the cell streams in the Twist ⫺/⫺ embryo suggest that Twist activity in the head mesenchyme is critical for guiding the migratory neural crest cells and the patterning of neurogenic derivatives. Some potential downstream targets of Twist are several types of molecules that have been identified to influence the pattern and directionality of neural crest cell migration. They include members of the ephrin/Eph family (Helbling et al., 1998; Smith et al., 1997; Holder and Klein, 1999), cadherins (Nakagawa and Takeichi, 1998, 1995), and other components of the extracellular matrix that interact via the integrin pathways with the cytoskeleton and the interaction of the neural crest cells with the secreted signaling molecules (Perris and Perissinotto, 2000). Although interactions of ephrins (e.g., LERK4/ephrin-A4 and RAGS/ephrin-A5) and their receptors are postulated as possible driving forces for the segregation of migrating neural crest cells (Xu et al., 2000; Helbling et al., 1998; Smith et al., 1997; Flenniken et al., 1996), at present there is no evidence of interaction between ephrin/Eph with

Twist. In the Erbb4 mutant embryo, defective migration of the neural crest cells is attributed to the lack of ErbB4Neuregulin signaling in the hindbrain, which in turns impacts the ability of the paraxial mesoderm to provide a guidance cue for migration (Golding et al., 2000). It is unlikely that altered ErbB4 signaling may contribute to the defects in neural crest cell migration in the Twist ⫺/⫺ embryos, since Erbb4 expression in rhombomeres 3 and 5 is maintained in three of four mutant embryos analyzed. In the present study, we have shown that Cdh11, which has an overlapping expression pattern with Twist in the cranial mesenchyme (Kimura et al., 1995), is downregulated in the Twist ⫺/⫺ embryo. This finding is consistent with the notion that one downstream effect of the loss of Twist activity is the disruption of cadherin function. The expression of specific types of cadherin has been shown to influence the migration and segregation of subpopulations of neural crest cells in the chick (Nakagawa and Takeichi, 1995, 1998). However, overexpressing the normal form of Xcad11 or the mutant form that lacks the ␤-catenin binding site blocks neural crest cell migration, whereas overexpressing the Xcad11 that lacks the extracellular domain results in the premature migration of neural crest cells in

© 2002 Elsevier Science (USA). All rights reserved.

267

Twist Function in the Mouse Embryo

Xenopus embryos (Borchers et al., 2001). However, irrespective of the changes in Xcad11 activity, twist expression is always reduced in the neural crest cells. The significance of this finding to the functional relationship between cadherin-11 and Twist is not known and cannot be reconciled with the presumptive upstream effect of Twist on Cdh11 in the mouse. It is possible, however, that Cdh11 and Twist may interact in a feedback regulatory pathway.

Twist Is Involved in the Alx/Cart Pathway for Neural Tube Morphogenesis and Development of Facial Skeletal Structures Loss of Twist function results in the disorganization of tissue architecture, the reduction in the number of cells in the head mesenchyme, and the malformation of the branchial arches (Chen and Behringer, 1995). This is consistent with the expression pattern of Twist in the craniofacial region of the mouse embryo where transcripts are detected in the mesenchyme flanking the neural tube and in the facial primordia (Wolf et al., 1991; Stoetzel et al., 1995; Fuchtbauer, 1995). Analysis of molecular markers reveals that the expression of three members of the gene family that encodes the aristaless-like transcription factors is markedly down-regulated in the head mesenchyme of Twist mutants. The loss of Cart1, Alx4, and Alx3 expression in Twist-null embryos and the identification of potential motifs in the 5⬘ sequence of these genes that may interact with bHLH factors such as TWIST (Loebel et al., 2002), highlights the possibility that members of this family are downstream targets of Twist activity, and that the collective loss of their function may contribute to the abnormal phenotype of the Twist-null mutant (O’Rourke and Tam, 2002). Expression domains of the Alx genes significantly overlap with that of Twist. Cart1 is expressed in the craniofacial mesenchyme and branchial arches (Zhao et al., 1996); Alx3 is expressed in the ectomesenchyme of the nasal process, jaws, the distal part of the tongue, and in other mesodermderived structures (ten Berge et al., 1998); and Alx4 transcripts are detected in the craniofacial mesenchyme of the prospective frontonasal mass, and the first branchial arch (Qu et al., 1997). Thus, the site of expression of these genes lends support to the possibility that they may play a role in mediating the action of Twist in the formation of neural crest- and mesoderm-derived craniofacial structures. In addition to the overlapping expression patterns, the phenotypes of embryos lacking Cart1 and Alx4 appear to represent subsets of the Twist-null mutant phenotype. Cart1 and Twist mutants display a similar exencephalic phenotype, which is also characterized by defects in the forebrain mesenchyme (Zhao et al., 1996; Juriloff and Harris, 2000). The loss of Alx4 alone has no recognizable effect on craniofacial development (Qu et al., 1998). However, embryos that carry mutations in both the Alx4 and Cart1 genes exhibit a midline fusion defect of nasal cartilages and retardation of mandibular outgrowth (Qu et al., 1999).

Based on these observations, it may be predicted that should the Twist ⫺/⫺ embryos survive to birth they would exhibit defects similar to the Cart1 and Cart1/Alx4 mutants, such as reduction of the frontal and parietal bones, and diminished size of branchial neural crest-derived skeletal structures, such as the mandible, maxilla, and zygomatic bones (Zhao et al., 1996; Qu et al., 1999).

Twist Activity Influences the Developmental Potency of Branchial Arch Tissues The malformation of the maxillary processes and the stalled outgrowth of the mandibular processes are accompanied by the loss of Fgf10 and Fgfr2 activity. Genes encoding two other growth factors, Fgf8 and Bmp4, are expressed in the maxillary and mandibular processes of Twist ⫺/⫺ embryos, but not always at the appropriate sites. Expression of Msx1, a putative downstream target of BMP4 signaling, is significantly reduced in the branchial arches. These findings therefore suggest that the defects in the morphogenesis of the branchial arches may be caused by disruption of the FGF and BMP signaling activities that are known to regulate branchial arch morphogenesis (FrancisWest et al., 1998; Trumpp et al., 1999; Kanzler et al., 2000). Fate-mapping studies have shown that the mesenchyme of the branchial arch is derived from both the cranial neural crest cells and the paraxial mesoderm (Trainor et al., 1994; Trainor and Tam, 1995; Trainor and Krumlauf, 2000). Our findings have shown that tissues in the truncated branchial arch of the Twist ⫺/⫺ embryo still express molecular markers that are characteristic of the neural crest cells (Sox10, Gsc lacZ, Dlx5) and the myogenic cells (RBP-lacZ) derived from the paraxial mesoderm. Sox10 expression is not downregulated and the Gsc lacZ and RBP-lacZ transgenes are expressed weakly, indicating that cells in the branchial arch might be arrested in the course of differentiation. Consistent with this is the demonstration of poor osteogenic and myogenic differentiation of the branchial arch tissues in the teratoma. Of particular interest is the reduced odontogenic potency of the mutant tissue and the formation of only the molar tooth bud typical of the proximal arch. Both the bone and the tooth of the jaws contain derivatives of the neural crest cells (Chai et al., 2000). The apparent deficiency in neural crest cell differentiation of Twist ⫺/⫺ cells may be associated with the disruption of signalling by BMP and FGF molecules. Absence of BMP signaling leads to hypomorphic arches and failure of skeletal tissue differentiation (Kanzler et al., 2000; Rice et al., 2000). Loss of the distal domain of Fgf8 expression and arch truncation (Trumpp et al., 1999) might lead to loss of incisors in the Twist ⫺/⫺ arch. The change in the histogenetic potency of the null mutant arch tissues cannot be directly related to the defects associated with Twist haploinsufficiency. In human patients and in Twist ⫹/⫺ mutant mice, the deformity of facial skeleton might be due to defects in tissue patterning rather than abnormal differentiation of arch derivatives (Bourgeois et al., 1998).

© 2002 Elsevier Science (USA). All rights reserved.

268

Soo et al.

ACKNOWLEDGMENTS We thank David Loebel and Peter Rowe for reading the manuscript, Simon Kinder for technical help, Scott Reyes for mouse husbandry, Seong-Seng Tan for the H253 mice, Andras Nagy for the B5/EGFP mice, and Olympus (Australia) Pty Ltd for the use of the Fluoview FV300 confocal microscope. Riboprobes were generously provided by Frits Meijlink (Alx3, Alx4), Eseng Lai (BF1), Brigid Hogan (Bmp4), Masatoshi Takeichi (Cdh11), John Rubenstein (Dlx5), Alex Joyner (En1), Paul Trainor (Erbb4), Gail Martin (Fgf8), Chuxia Deng (Fgfr2, Fgf10 with permission from David Ornitz), C.C. Hui (Gli1, Gli3), Sally Dunwoodie (Hesx1), Filippo Rijli (Hoxa2), Robb Krumlauf (Hoxb1), David Wilkinson (Krox20), Robert Hill (Msx1), Siew-Lan Ang (Otx2), Peter Gruss (Pax3, Pax6), Matthew Scott (Ptch), Jacques Drouin (Pitx1), Andy McMahon (Shh), C. M. Fan (Sim1), Oliver Guillermo (Six3), and Peter Koopman (Sox10). Our work is supported by the National Health and Medical Research Council (NHMRC) of Australia, the Dental Board of NSW and Mr. James Fairfax. K.S. is a Sydney University Medical Foundation Scholar, M.P.O. is a NMHRC Dental Postgraduate Scholar, and P.P.L.T. is a NHMRC Senior Principal Research Fellow.

REFERENCES Barembaum, M., Moreno, T. A., LaBonne, C., Sechrist, J., and Bronner-Fraser, M. (2000). Noelin-1 is a secreted glycoprotein involved in generation of the neural crest. Nat. Cell Biol. 2, 219 –225. Basler, K., Edlund, T., Jessell, T. M., and Yamada, T. (1993). Control of cell pattern in the neural tube: Regulation of cell differentiation by dorsalin-1, a novel TGF beta family member. Cell 73, 687–702. Baylies, M. K., and Bate, M. (1996). twist: A myogenic switch in Drosophila. Science 272, 1481–1484. Bonstein, L., Elias, S., and Frank, D. (1998). Paraxial-fated mesoderm is required for neural crest induction in Xenopus embryos. Dev. Biol. 193, 156 –168. Borchers, A., David, R., and Wedlich, D. (2001). Xenopus cadherin-11 restrains cranial neural crest migration and influences neural crest specification. Development 128, 3049 –3060. Bourgeois, P., Bolcato-Bellemin, A. L., Danse, J. M., Bloch-Zupan, A., Yoshiba, K., Stoetzel, C., and Perrin-Schmitt, F. (1998). The variable expressivity and incomplete penetrance of the twist-null heterozygous mouse phenotype resemble those of human Saethre–Chotzen syndrome. Hum. Mol. Genet. 7, 945–957. Briscoe, J., Pierani, A., Jessell, T. M., and Ericson, J. (2000). A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101, 435– 445. Britsch, S., Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K. A., Birchmeier, C., and Wegner, M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15, 66 –78. Britsch, S., Li, L., Kirchhoff, S., Theuring, F., Brinkmann, V., Birchmeier, C., and Riethmacher, D. (1998). The ErbB2 and ErbB3 receptors and their ligand, neuregulin-1, are essential for development of the sympathetic nervous system. Genes Dev. 12, 1825–1836. Chai, Y., Jiang, X., Ito, Y., Bringas, P. J., Han, J., Rowitch, D. H., Soriano, P., McMahon, A. P., and Sucov, H. M. (2000). Fate of the

mammalian cranial neural crest during tooth and mandibular morphogenesis. Development 127, 1671–1679. Chan, W. Y., and Tam, P. P. L. (1988). A morphological and experimental study of the mesencephalic neural crest cells in the mouse embryo using wheat germ agglutinin-gold conjugate as the cell marker. Development 102, 427– 442. Chen, Z. F., and Behringer, R. R. (1995). twist is required in head mesenchyme for cranial neural tube morphogenesis. Genes Dev. 9, 686 – 699. El Ghouzzi, V., Le Merrer, M., Perrin-Schmitt, F., Lajeunie, E., Benit, P., Renier, D., Bourgeois, P., Bolcato-Bellemin, A. L., Munnich, A., and Bonaventure, J. (1997). Mutations of the TWIST gene in the Saethre–Chotzen syndrome. Nat. Genet. 15, 42– 46. Farlie, P. G., Kerr, R., Thomas, P., Symes, T., Minichiello, J., Hearn, C. J., and Newgreen, D. (1999). A paraxial exclusion zone creates patterned cranial neural crest cell outgrowth adjacent to rhombomeres 3 and 5. Dev. Biol. 213, 70 – 84. Flenniken, A. M., Gale, N. W., Yancopoulos, G. D., and Wilkinson, D. G. (1996). Distinct and overlapping expression patterns of ligands for Eph-related receptor tyrosine kinases during mouse embryogenesis. Dev. Biol. 179, 382– 401. Francis-West, P., Ladher, R., Barlow, A., and Graveson, A. (1998). Signalling interactions during facial development. Mech. Dev. 75, 3–28. Fuchtbauer, E. M. (1995). Expression of M-twist during postimplantation development of the mouse. Dev. Dyn. 204, 316 –322. Gassmann, M., Casagranda, F., Orioli, D., Simon, H., Lai, C., Klein, R., and Lemke, G. (1995). Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390 –394. Gitelman, I. (1997). Twist protein in mouse embryogenesis. Dev. Biol. 189, 205–214. Golding, J. P., Trainor, P., Krumlauf, R., and Gassmann, M. (2000). Defects in pathfinding by cranial neural crest cells in mice lacking the neuregulin receptor ErbB4. Nat. Cell Biol. 2, 103– 109. Goodrich, L. V., Milenkovic, L., Higgins, K. M., and Scott, M. P. (1997). Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 277, 1109 –1113. Goodrich, L. V., Jung, D., Higgins, K. M., and Scott, M. P. (1999). Overexpression of Ptc1 inhibits induction of Shh target genes and prevents normal patterning in the neural tube. Dev. Biol. 211, 323–334. Hacker, A., and Guthrie, S. (1998). A distinct developmental programme for the cranial paraxial mesoderm in the chick embryo. Development 125, 3461–3472. Hadjantonakis, A. K., Gertsenstein, M., Ikawa, M., Okabe, M., and Nagy, A. (1998). Generating green fluorescent mice by germline transmission of green fluorescent ES cells. Mech. Dev. 76, 79 –90. Harfe, B. D., Vaz, G. A., Kenyon, C., Liu, J., Krause, M., and Fire, A. (1998). Analysis of a Caenorhabditis elegans Twist homolog identifies conserved and divergent aspects of mesodermal patterning. Genes Dev. 12, 2623–2635. Hebrok, M., Wertz, K., and Fuchtbauer, E. M. (1994). M-twist is an inhibitor of muscle differentiation. Dev. Biol. 165, 537–544. Hebrok, M., Fuchtbauer, A., and Fuchtbauer, E. M. (1997). Repression of muscle-specific gene activation by the murine Twist protein. Exp. Cell Res. 232, 295–303. Helbling, P. M., Tran, C. T., and Brandli, A. W. (1998). Requirement for EphA receptor signaling in the segregation of Xenopus third and fourth arch neural crest cells. Mech. Dev. 78, 63–79.

© 2002 Elsevier Science (USA). All rights reserved.

269

Twist Function in the Mouse Embryo

Herbarth, B., Pingault, V., Bondurand, N., Kuhlbrodt, K., HermansBorgmeyer, I., Puliti, A., Lemort, N., Goossens, M., and Wegner, M. (1998). Mutation of the Sry-related Sox10 gene in Dominant megacolon, a mouse model for human Hirschsprung disease. Proc. Natl. Acad. Sci. USA 95, 5161–5165. Holder, N., and Klein, R. (1999). Eph receptors and ephrins: effectors of morphogenesis. Development 126, 2033–2044. Hopwood, N. D., Pluck, A., and Gurdon, J. B. (1989). A Xenopus mRNA related to Drosophila twist is expressed in response to induction in the mesoderm and the neural crest. Cell 59, 893– 903. Jacobson, A. G., and Tam, P. P. L. (1982). Cephalic neurulation in the mouse embryo analyzed by SEM and morphometry. Anat. Rec. 203, 375–396. Jessell, T. M. (2000). Neuronal specification in the spinal cord: Inductive signals and transcriptional codes. Nat. Rev. Genet. 1, 20 –29. Juriloff, D. M., and Harris, M. J. (2000). Mouse models for neural tube closure defects. Hum. Mol. Genet. 9, 993–1000. Kanzler, B., Foreman, R. K., Labosky, P. A., and Mallo, M. (2000). BMP signaling is essential for development of skeletogenic and neurogenic cranial neural crest. Development 127, 1095–1104. Kimura, Y., Matsunami, H., Inoue, T., Shimamura, K., Uchida, N., Ueno, T., Miyazaki, T., and Takeichi, M. (1995). Cadherin-11 expressed in association with mesenchymal morphogenesis in the head, somite, and limb bud of early mouse embryos. Dev. Biol. 169, 347–358. Kinder, S. J., Tan, S. S., and Tam, P. P. L. (2000). Cell grafting and fate mapping of the early-somite-stage mouse embryo. Methods Mol. Biol. 135, 425– 437. LeDouarin, N. (1982). “The Neural Crest.” Cambridge Univ. Press, Cambridge. Lee, K. J., Dietrich, P., and Jessell, T. M. (2000). Genetic ablation reveals that the roof plate is essential for dorsal interneuron specification. Nature 403, 734 –740. Litingtung, Y., and Chiang, C. (2000). Specification of ventral neuron types is mediated by an antagonistic interaction between Shh and Gli3. Nat. Neurosci. 3, 979 –985. Liu, J. P., and Jessell, T. M. (1998). A role for rhoB in the delamination of neural crest cells from the dorsal neural tube. Development 125, 5055–5067. Loebel, D. A. F., O’Rourke, M. P., Steiner, K. A., Banyer, J., and Tam, P. P. L. (2002). Isolation of differentially expressed genes from wild-type and Twist mutant mouse limb buds. Genesis, in press. Mayor, R., Morgan, R., and Sargent, M. G. (1995). Induction of the prospective neural crest of Xenopus. Development 121, 767–777. Millonig, J. H., Millen, K. J., and Hatten, M. E. (2000). The mouse Dreher gene Lmx1a controls formation of the roof plate in the vertebrate CNS. Nature 403, 764 –769. Nakagawa, S., and Takeichi, M. (1995). Neural crest cell– cell adhesion controlled by sequential and subpopulation-specific expression of novel cadherins. Development 121, 1321–1332. Nakagawa, S., and Takeichi, M. (1998). Neural crest emigration from the neural tube depends on regulated cadherin expression. Development 125, 2963–2971. Nusslein-Volhard, C., Wieschaus, E., and Kluding, H. (1984). Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. 1. Zygotic loci on the 2nd chromosome. Roux’s Arch. Dev. Biol. 193, 267–282. O’Rourke, M. P., and Tam, P. P. L. (2002). Twist functions in mouse development. Int. J. Dev. Biol., in press.

Perris, R., and Perissinotto, D. (2000). Role of the extracellular matrix during neural crest cell migration. Mech. Dev. 95, 3–21. Qu, S., Li, L., and Wisdom, R. (1997). Alx-4: cDNA cloning and characterization of a novel paired-type homeodomain protein. Gene 203, 217–223. Qu, S., Tucker, S. C., Ehrlich, J. S., Levorse, J. M., Flaherty, L. A., Wisdom, R., and Vogt, T. F. (1998). Mutations in mouse Aristaless-like4 cause Strong’s luxoid polydactyly. Development 125, 2711–2721. Qu, S., Tucker, S. C., Zhao Q., deCrombrugghe, B., and Wisdom, R. (1999). Physical and genetic interactions between Alx4 and Cart1. Development 126, 359 –369. Quinlan, G. A., Davidson, B. P., and Tam, P. P. L. (2001). Lineage allocation during early embryogenesis. Mapping of the neural primordia and application to the analysis of mouse mutants. Methods Mol. Biol. 158, 227–250. Rice, D. P., Aberg, T., Chan, Y., Tang, Z., Kettunen, P. J., Pakarinen, L., Maxson, R. E., and Thesleff, I. (2000). Integration of FGF and TWIST in calvarial bone and suture development. Development 127, 1845–1855. Rivera-Perez, J. A., Wakamiya, M., and Behringer, R. R. (1999). Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development. Development 126, 3811– 3821. Rohwedel, J., Horak, V., Hebrok, M., Fuchtbauer, E. M., and Wobus, A. M. (1995). M-twist expression inhibits mouse embryonic stem cell-derived myogenic differentiation in vitro. Exp. Cell Res. 220, 92–100. Sasaki, H., Nishizaki, Y., Hui, C., Nakafuku, M., and Kondoh, H. (1999). Regulation of Gli2 and Gli3 activities by an aminoterminal repression domain: Implication of Gli2 and Gli3 as primary mediators of Shh signaling. Development 126, 3915– 3924. Simpson, P. (1983). Maternal–zygotic gene interactions during formation of the dorsoventral pattern in Drosophila embryos. Genetics 105, 615– 632. Smith, A., Robinson, V., Patel, K., and Wilkinson, D. G. (1997). The EphA4 and EphB1 receptor tyrosine kinases and ephrin-B2 ligand regulate targeted migration of branchial neural crest cells. Curr. Biol. 7, 561–570. Soto, J. G., Nelson, B. H., and Weisblat, D. A. (1997). A leech homolog of twist: Evidence for its inheritance as a maternal mRNA. Gene 199, 31–37. Southard-Smith, E. M., Kos, L., and Pavan, W. J. (1998). Sox10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nat. Genet. 18, 60 – 64. Stoetzel, C., Weber, B., Bourgeois, P., Bolcato-Bellemin, A. L., and Perrin-Schmitt, F. (1995). Dorso-ventral and rostro-caudal sequential expression of M-twist in the postimplantation murine embryo. Mech. Dev. 51, 251–263. Sturm, K., and Tam, P. P. L. (1993). Isolation and culture of whole postimplantation embryos and germ layer derivatives. Methods Enzymol. 225, 164 –190. Tam, P. P. L., and Tan, S. S. (1992). The somitogenetic potential of cells in the primitive streak and the tail bud of the organogenesisstage mouse embryo. Development 115, 703–715. Tam, P. P. L. (1993). Histogenetic potency of embryonic tissues in ectopic sites. Methods Enzymol. 225, 190 –204. Tan, S. S. (1991). Liver-specific and position-specific expression of a retinol-binding protein-lacZ fusion gene (RBP-lacZ) in transgenic mice. Dev. Biol. 146, 24 –37.

© 2002 Elsevier Science (USA). All rights reserved.

270

Soo et al.

ten Berge, D., Brouwer, A., el Bahi, S., Guenet, J. L., Robert, B., and Meijlink, F. (1998). Mouse Alx3, an aristaless-like homeobox gene expressed during embryogenesis in ectomesenchyme and lateral plate mesoderm. Dev. Biol. 199, 11–25. Theil, T., Alvarez-Bolado, G., Walter, A., and Ruther, U. (1999). Gli3 is required for Emx gene expression during dorsal telencephalon development. Development 126, 3561–3571. Tole, S., Ragsdale, C. W., and Grove, E. A. (2000). Dorsoventral patterning of the telencephalon is disrupted in the mouse mutant extra-toes(J). Dev. Biol. 217, 254 –265. Trainor, P. A., Tan, S. S., and Tam, P. P. L. (1994). Cranial paraxial mesoderm: Regionalisation of cell fate and impact on craniofacial development in mouse embryos. Development 120, 2397– 2408. Trainor, P. A., and Tam, P. P. L. (1995). Cranial paraxial mesoderm and neural crest cells of the mouse embryo: Co-distribution in the craniofacial mesenchyme but distinct segregation in branchial arches. Development 121, 2569 –2582. Trainor, P. A., Zhou, S. X., Parameswaran, M., Quinlan, G. A., Gordon, M., Sturm, K., and Tam, P. P. L. (1999). Application of lacZ transgenic mice to cell lineage studies. Methods Mol. Biol. 97, 183–200. Trainor, P. A., and Krumlauf, R. (2000). Patterning the cranial neural crest: hindbrain segmentation and Hox gene plasticity. Nat. Rev. Neurosci. 1, 116 –124. Trumpp, A., Depew, M. J., Rubenstein, J. L., Bishop, J. M., and Martin, G. R. (1999). Cre-mediated gene inactivation demonstrates that FGF8 is required for cell survival and patterning of the first branchial arch. Genes Dev. 13, 3136 –3148.

Tuckett, F., and Morriss-Kay, G. M. (1985). The kinetic behaviour of the cranial neural epithelium during neurulation in the rat. J. Embryol. Exp. Morphol. 85, 111–119. Wang, S. M., Coljee, V. W., Pignolo, R. J., Rotenberg, M. O., Cristofalo, V. J., and Sierra, F. (1997). Cloning of the human twist gene: Its expression is retained in adult mesodermally-derived tissues. Gene 187, 83–92. Wilkinson, D. G., and Nieto, M. A. (1993). Detection of messenger RNA by in situ hybridization to tissue sections and whole mounts. Methods Enzymol. 225, 361–373. Wolf, C., Thisse, C., Stoetzel, C., Thisse, B., Gerlinger, P., and Perrin-Schmitt, F. (1991). The M-twist gene of Mus is expressed in subsets of mesodermal cells and is closely related to the Xenopus X-twi and the Drosophila twist genes. Dev. Biol. 143, 363–373. Xu, Q., Mellitzer, G., and Wilkinson, D. G. (2000). Roles of Eph receptors and ephrins in segmental patterning. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 993–1002. Yasui, K., Zhang, S. C., Uemura, M., Aizawa, S., and Ueki, T. (1998). Expression of a twist-related gene, Bbtwist, during the development of a lancelet species and its relation to cephalochordate anterior structures. Dev. Biol. 195, 49 –59. Zhao, Q., Behringer, R. R., and de Crombrugghe, B. (1996). Prenatal folic acid treatment suppresses acrania and meroanencephaly in mice mutant for the Cart1 homeobox gene. Nat. Genet. 13, 275–283.

© 2002 Elsevier Science (USA). All rights reserved.

Received for publication January 7, Revised April 12, Accepted April 12, Published online June 7,

2002 2002 2002 2002