Skip Navigation


Human Molecular Genetics Advance Access originally published online on May 3, 2007
Human Molecular Genetics 2007 16(12):1423-1436; doi:10.1093/hmg/ddm093
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/12/1423    most recent
ddm093v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Van de Putte, T.
Right arrow Articles by Huylebroeck, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Van de Putte, T.
Right arrow Articles by Huylebroeck, D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Neural crest-specific removal of Zfhx1b in mouse leads to a wide range of neurocristopathies reminiscent of Mowat–Wilson syndrome

Tom Van de Putte1,2,*, Annick Francis1,2, Luc Nelles1,2,{dagger}, Leo A. van Grunsven1,2,{ddagger} and Danny Huylebroeck1,2

1 Laboratory of Molecular Biology (Celgen), Department of Molecular and Developmental Genetics (VIB11), |nFlanders Institute of Biotechnology (VIB) and 2 Department of Human Genetics, KULeuven, Herestraat 49, B-3000 Leuven, Belgium

* To whom correspondence should be addressed at: Laboratory of Molecular Biology (Celgen), KULeuven, VIB Department of Molecular and Developmental Genetics, Flanders Institute of Biotechnology (VIB), Campus Gasthuisberg, Building Ond&Nav1, PO Box 812, Herestraat 49, B-3000 Leuven, Belgium. Tel: +32 16345916; Fax: +32 16345933; Email: tom.vandeputte{at}med.kuleuven.be


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Mowat–Wilson syndrome is a recently delineated autosomal dominant developmental anomaly, whereby heterozygous mutations in the ZFHX1B gene cause mental retardation, delayed motor development, epilepsy and a wide spectrum of clinically heterogeneous features, suggestive of neurocristopathies at the cephalic, cardiac and vagal levels. However, our understanding of the etiology of this condition at the cellular level remains vague. This study presents the Zfhx1b protein expression domain in mouse embryos and correlates this with a novel mouse model involving a conditional mutation in the Zfhx1b gene in neural crest precursor cells. These mutant mice display craniofacial and gastrointestinal malformations that show resemblance to those found in human patients with Mowat–Wilson syndrome. In addition to these clinically recognized alterations, we document developmental defects in the heart, melanoblasts and sympathetic and parasympathetic anlagen. The latter observations in our mouse model for Mowat–Wilson suggest a hitherto unknown role for Zfhx1b in the development of these particular neural crest derivatives, which is a set of observations that should be acknowledged in the clinical management of this genetic disorder.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The neural crest is a transient multipotent cell population specific to vertebrates. Initially formed in the neural ridge of the late gastrulating embryo (E8.5 in mouse, around the fourth week in humans), these cells detach from the neural tube epithelium and migrate through the embryo. As such, they participate in the formation of many cell types and tissues (1). The site of origin along the anterior–posterior axis of the embryo determines the fate of this pluripotent cell population: cranial neural crest cells give rise to connective tissue and skeleto-muscular components of the head, while vagal neural crest cells (formed at the level of the first seven somites) populate the gastrointestinal tract and the cardiovasculature. Most peripheral sensory neurons, glial cells, some neuro-endocrine cells and melanocytes are trunk-specific neural crest derivatives. Owing to their broad developmental potential and contribution to several tissues and organs, a defect in the induction, (de)adhesion and migration, and/or differentiation of neural crest cells has far-reaching consequences in numerous organs. Neurocristopathies therefore comprise pigmentation defects, Hirschsprung disease and all kinds of craniofacial and cardiac abnormality.

Mutation analysis in humans has demonstrated that mutations in ZFHX1B are implicated in the etiology of a particular neurocristopathy, the Mowat–Wilson syndrome (2–4). This gene encodes Zfhx1b (also named Sip1, Zeb2), a two-handed zinc finger DNA-binding protein, initially characterized as a transcriptional repressor. Zfhx1b has been shown to interact with several proteins including activated Smads (intracellular mediators of transforming growth factor-ß signaling), the co-repressor CtBP and the co-activators p300 and pCAF (5–7). Binding with its zinc finger clusters to a tandem of spaced CACCT(G) (in 50% of the cases) or CACANNT(G) (30%) sequences (8,9) in regulatory elements of genes causes Zfhx1b to act as transcriptional repressor both in cell culture and in vivo (4,10,11), but there is evidence that this factor also activates gene transcription (9,12,13).

Mowat–Wilson syndrome is clinically characterized by a heterogeneous spectrum of defects in the central nervous system (CNS) (mental retardation, delayed development of motoric abilities, agenesis or hypoplasia of the corpus callosum, and seizures), together with defects that are a sign of developmental disturbances in neural crest cells (deviant facial characteristics, Hirschsprung disease and heart defects) (2,3,14). Most of the clinically characterized mutations are either whole gene deletions or nonsense and frameshift mutations in a single allele. The latter type of mutation leads to a truncated protein product or by analogy with observations in the Zfhx1b-deficient mouse, to absence of the given truncated protein (15), which may be caused by instability of the mutant protein. Most clinical evidence (no obvious genotype–phenotype correlation), together with the Zfhx1b mRNA expression pattern during embryonic development (transiently high expression levels in the affected tissues during crucial stages of their development), suggests the loss of one functional ZFHX1B allele to be sufficient to cause the syndrome. However, a few missense mutations have been identified which, if not causing drastic conformational changes, may modify and therefore disclose important interaction sites for protein partners (16–22).

As a first step to understand the disease phenotype, we have reported the generation of a conventional knockout mouse with a defect in the Zfhx1b gene that is conceptually comparable to mutations found in human patients (15). Like human carriers, such heterozygous mutant mice are viable and the vast majority lacks a corpus callosum. However, they do not develop additional morphological phenotypes analogous to the majority of the Mowat–Wilson syndrome patients. There are several explanations why heterozygous mutations in Zfhx1b in the mouse may not be sufficient to produce a phenotype equivalent to human patients. One possibility is that human neural crest development, though most likely similar to that of the mouse, is more sensitive to the loss of one allele of genes that are involved in neural crest development. Indeed, mutations that are completely recessive in mice but cause haploinsufficiency in humans have been observed in mouse models for neurocristopathies like DiGeorge syndrome (23), Shah-Waardenburg syndrome [EDNRB and EDN3 (24)] and Hirschsprung disease [Ret, (25)]. A second possibility is that, whereas in our engineered mouse model the truncated protein is undetectable (15), mutated Zfhx1b protein in human carriers persists in the cell and thereby interferes with the activity of the remaining wild-type protein e.g. by dimerizing with partner proteins. Gene expression modulation via the wild-type protein is thereby slowed or impaired. Thirdly, the clinical variability among Mowat–Wilson patients suggests the presence of modifier genes, which remain to be discovered. If so, the difference between mouse and human in expressivity of the disease may be dependent on species-specific modifying factors.

In the homozygous state, a condition never described in humans, Zfhx1b-deficient mouse embryos are developmentally compromised from post-gastrulation stage E8.5 onwards. They show defective neuro-epithelium development, failure of neural tube closure and defects in neural crest induction (at the vagal level) and delamination (at the cranial level) (4). In addition, somitogenesis is arrested (26) and homozygous mutant embryos die at E9.5 likely from cardiovascular dysfunction. These studies have provided molecular clues for the importance of Zfhx1b in the development of somites, neural plate and neural crest cells.

Induction of neural crest precursor cells on the neural ridge and their correct specification relies on the interplay between signals (like BMP4 and – 7) originating from the non-neural epithelium and the competence of the CNS primordium to respond to them (27). Seen the documented molecular deviations in the mutant neural plate, neural crest cell defects attributable to mutations in Zfhx1b may indirectly be caused by the incompetence of the CNS primordium to correctly interpret the neural crest inducing signals. By applying neural crest precursor-specific gene ablation (28), we anal- yze here the cell-autonomous role of Zfhx1b in the intrinsic differentiation program of particular neural crest cell precursors apart from the indirect influences of surrounding tissue. Because this tissue-specific mutagenesis approach may overcome the early embryonic lethality, we anticipate to create a mouse model for Mowat–Wilson syndrome which would unmask the later and neural crest-specific functions of Zfhx1b in the etiology of the human disease. Here, we show that homozygous mutant mice that have lost Zfhx1b protein in their neural crest cells display specific abnormalities in craniofacial, heart and melanocyte development, as well as defects in the peripheral nervous system of the gastrointestinal tract and the sympatho-adrenal lineage. These various abnormalities suggest that the dependence on Zfhx1b activity differs in different neural crest precursors.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Neural crest-specific Zfhx1b gene deletion causes embryonic lethality
Our strategy relied on the use of a floxed allele of Zfhx1b, in which exon 7 is flanked by loxP sites (15), in concert with the established Wnt1–Cre transgenic mouse (28), enabling recombination in neural crest precursor cells on the dorsal neural folds from four-somites stage on (29). Such deletion of the Zfhx1b gene results in death at two different stages. Up to E11.5, neural crest-specific knockout embryos were fully represented, had normal gross morphology and normal intra-embryonic and yolk sac blood circulation. A total of 40% (40 of 98) of the knockout embryos recovered at E12.5 or later were however necrotic (Fig. 1A), whereas the majority of the knockouts develop further and is stillborn (Fig. 1A). Two morphologically distinct types of knockout are found at birth. The majority of mutant fetuses are, although normal with respect to their birth weight, easily recognizable by their deviant facial appearance: wide open eyes and a shorter snout with reduced muzzle mesenchyme that lacks whisker follicles (Fig. 1B and C). Some embryos are however smaller, pale and have an open brain, which mainly encompasses the midbrain (Fig. 1D). This can be observed at very early developmental stages (E10.5–E11.0). It may be attributed to a local defect in neural tube closure due to the transient Wnt1-driven Cre expression in the midbrain at E8.5 (29,30), similar to what was observed along the whole anterior–posterior axis in the conventional knockout (4). This phenotype has a penetrance of approximately 13% (21 of 161) among the knockout embryos. The incomplete penetrance of this phenotype may be related to the mixed genetic background in which the mice are bred. Indeed, mouse strains differ in the initiation site and timing of the closure of the cranial neural tube (31). A variable time lag between the onset of Cre synthesis and the initiation of the closure of the neural tube may be a reason for the differential sensitivity to neural tube closure defects among the different pups. Since we believe that this phenotype finds its origin in a deficiency of the neural tube rather than in a neural crest defect caused by Zfhx1b, we did not include a phenotypic analysis of it in this paper. In fact, we focused here on the dissection of the role of Zfhx1b in neural crest development.


Figure 1
View larger version (59K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Neural crest-specific Zfhx1b gene deletion causes embryonic lethality. (A) Genotype distribution of embryos from intercrosses between a compound heterozygous Wnt1–Cre+/– SipKO/flox male with a Sip1flox/flox female reveals that the Wnt1–cre+/– Sip1KO/flox embryos are fully represented (Mendelian expectation is 25%) up to E10.5. At E11.5, 40% of the knockout embryos die abruptly due to cardiovascular dysfunction. Remnant of the latter embryos can be found in increasingly resorbing state throughout gestation. About 13% of the mutant embryos show an open midbrain. Both phenotypes can occur independently within the same knockout. No knockout embryos survive birth (asterisk). (B–D) Appearance of a control littermate (B) and stillborn Wnt1–cre+/– Sip1KO/flox embryos at P0 without (C) and with (D) an open midbrain phenotype. Note the deviant facial appearance in the latter embryos: wide open eyes and a shorter snout with reduced muzzle mesenchyme that lack whisker follicles.

 
Zfhx1b expression in the cranium and its role in craniofacial development
Mowat–Wilson patients are diagnosed by their facial features including cupped ears with upturned lobules, wide nasal bridge, deep-set eyes, hypertelorism, pointed chin, prominent rounded nasal tip and a high incidence of microcephaly (21,22). Also malpositioning of teeth and/or delayed tooth eruption have recently been reported (32). The cranium has a dual developmental origin (33–35). The neurocranium that protects the brain is of mesodermal origin, while the coronal suture between the frontal and the parietal bones are derived from neural crest that emerges from the dorsal midline at midbrain level. These neural crest cells are also the source of skeletal and muscular elements of the presumptive midface, the palate shelves and the oral region. Hindbrain-born neural crest cells give rise to the mandibular, hyoid and five branchial arches (36,37).

Using an anti-Zfhx1b antibody raised against a polypeptide encoded by mouse exons 2 and 3 (38), we found high Zfhx1b content in migrating cranial neural crest cells originating from the mid- and hindbrain and in condensing neural crest cells in the branchial arches at E8.5 (4,39) (Fig. 2A). At E9.5, Zfhx1b is down-regulated abruptly in the skeletogenic neural crest cells that have completed their migration into the midfacial region and branchial arches (Fig. 2B). Interestingly, the strong nuclear Zfhx1b staining re-emerges at E11.5 in the neural crest-derived mesenchyme of the maxillary and mandibular process, except for the prospective nasal capsule (Fig. 2C). This particular expression pattern is maintained at least until E15.5 at which the connective tissue of the papilla in the follicles of the vibrissa showed particularly high expression levels (Fig. 2D). At E12.5, a regionally confined Zfhx1b-positive domain was seen in the dental mesenchyme surrounding the dental epithelium of the upper and lower molar tooth primordia (Fig. 2E).


Figure 2
View larger version (61K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Zfhx1b expression in the developing head and its role in craniofacial development. (A–E) Immunohistochemistry on cranial sections of wild-type embryos with an anti-Zfhx1b antibody. (A) Zfhx1b protein (brown) is found in E8.5 embryos in migratory cranial neural crest cells and those populating the mesenchyme of the branchial arches (BA). (B) At E9.5, the first branchial arch, which becomes the maxilla, the palate and the mandibular, is devoid of Zfhx1b protein. The only Zfhx1b-positive cells are Schwann cell precursors associated with the nerves that extend from the trigeminal (V) ganglia into the first branchial arch. (C) At E11.5, strong nuclear Zfhx1b staining re-emerges in the neural crest-derived mesenchyme of the maxillary (max) and mandibular (man) process. Cells in the prospective nasal capsule (nc) do not express Zfhx1b. (D) At E15.5, the connective tissue of the papilla (p) in the follicles (f) of the vibrissa as well as the mesenchyme surrounding them showed particularly high Zfhx1b expression levels. (E) At E12.5 a local increase of Zfhx1b expression could be seen in the dental mesenchyme surrounding the dental epithelium (de) of the molar tooth primordia. (F,G) Whole-mount staining of Wnt1–Cre-dependent ß-galactosidase expression in control (F) and mutant (G) E14.5 embryos shows a distinct facial morphology: reduced muzzle mesenchyme that lack whisker follicles, a shorter snout and wide open eyes. No disturbance of cranial neural crest distribution is apparent in the mutant. (H–Q) Alizarin red (bone)/Alcian blue (cartilage) preparations of skulls of control (H,J,L,N,P) and Zfhx1b-null (I,K,M,O,Q) newborns. (H,I) The mutant (I) nasal (n) and premaxilla (pm) bones are hypoplastic (compare length of green arrow) and the ossification is incomplete leading to a fragmented appearance. The cartilaginous nasal capsule appears normal. (J,K) The mutant mandible (K) is hypoplastic. Its molar socket (ms) is absent whereas the incisors are normal. (L,M) Ventral view on the mandibles and the skull base reveals that the mandibles (man) are not curved in the mutant (M). The hyoid (hy) bone and the cartilage of the larynx (la) (P,Q), both derived from the second and third arch mesenchyme, were unaffected. (N,O) Dorsal view on the vault of the skull: The squamous parts of frontal (fr) and parietal (pa) bones lack ossification towards the metopic region in the absence of Zfhx1b (O), resulting in broadened sagital (ss) and metopic (ms) sutures, whereas the coronal sutures (cs) are not affected. (P,Q) Lateral view on the ear region of the skull of control (P) and mutant (Q) newborns. The retrotympanic process of the squamous bone (sq) is severely reduced in size. In the inner ear, the tympanic ring (tr) and gonium (go) is truncated. nt: neural tube, bo: basi-oocipital, i: interparietal bone. Size bar: 100 µm.

 
Lineage labeling by Wnt1–Cre-mediated ß-galactosidase expression from the R26R locus indicates no apparent difference in migration, distribution and proliferative (as measured by Ki67 staining) capacities of mutant cranial neural crest cells within the first branchial arch and the frontonasal prominence till E11.5 (data not shown). Knockout fetuses can be recognized from E12.5 by their reduced muzzle mesenchyme that lacks whisker follicles. A shortened snout (the nasal bone is 25 ± 6% reduced in size (compare green arrows in Fig. 2H–I,N,O)) and wide open eyes were apparent from E14.5 (Fig. 2F and G). This craniofacial phenotype was fully penetrant among the mutants that survive beyond E11.5. Alizarin red/Alcian blue preparations of newborns revealed the underlying skeletal deformities. The nasal and premaxilla bone ossification was incomplete, whereas the cartilaginous nasal capsule was normal (Fig. 2H and I). The mandible was hypoplastic and its rotation altered (Fig. 2J–M). Its molar socket was absent but the lower incisors were normal. The squamous parts of frontal bones lacked ossification towards the metopic region, resulting in broadened sagital and metopic (frontal) sutures, whereas the coronal (lateral) sutures were not affected. The squamal bones (in particular the retrotympanic process of the squamous bone) were misshaped and reduced in size. In the inner ear region, the otic capsule and the semicircular canals appeared normal but the skeletal elements associated with it (the tympanic ring and gonial bone) showed hypoplasia (Fig. 2P and Q). The palate was not affected. Hyoid skeleton and the cartilage of the larynx (both derived from the second and third arch mesenchyme) forms normally as well. More posterior skull structures like the occipital bones (derived from paraxial mesoderm) and the parietal and interparietal bones (from cephalic and somatic mesoderm origin) were normal.

Zfhx1b expression in the gastrointestinal tract and intestinal aganglionosis in Wnt1–Cre Zfhx1b knockout mice
Congenital megacolon (Hirschsprung disease) occurs in half to two-thirds of Mowat–Wilson patients. Another 10% has chronic constipation (21). Neural crest precursors bound for the gastroenteric nervous system seem to originate from two regions along the neural tube, the post-otic hindbrain adjacent to somites 1–7 and possibly the sacral neural crest cells that emerge caudal to somite 24. The former innervates the esophagus, stomach, and small and large intestine and the latter is proposed to colonize the hindgut (1,40–42).

At E10.5 and later, Zfhx1b protein was seen in single cells found at increasingly more caudal levels within the mesenchyme of the foregut and midgut wall (Fig. 3A and B). From E12.5, high Zfhx1b levels were detected in the submucosal and myenteric plexus of the stomach (Fig. 3C) and along the entire length of the intestinal tract (Fig. 3D).


Figure 3
View larger version (144K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. Zfhx1b expression in the gastrointestinal tract and intestinal aganglionosis in ‘Wnt1cre–Zfhx1b’ knockout mice. (A–D) Immunohistochemistry of wild-type embryos with the antibody to Zfhx1b. (A) At E10.5, Zfhx1b-expressing cells were found closely apposed to the lateral and ventral parts of the left and right dorsal aorta and were found entering the primitive gut (red arrowhead). This staining pattern reflects the migratory pathway of enteric nervous system precursors originating from the vagal region. Note the Zfhx1b expression in the neural tube (nt), the DRG and the mesenchyme of the developing lung buds (lb). (B) Section through the foregut of an E10.5 embryo reveals a strong staining for Zfhx1b (red) in the nuclei of undifferentiated vagal crest derived enteric precursors. Neurofilament immunoreactivity is visualized in green and nuclei are blue. (C,D) Later during the development of the gut, the enteric nervous system is heavily expressing Zfhx1b. On transversal section through the E12.5 stomach (C) and gut (D), the outer myenteric (red arrowhead) and inner submucosal ganglia (green arrowhead) are clearly positive. The black arrow marks a distal segment of the gut where the front of migrating neural crest cells has just arrived. (E,F) In situ hybridization with cRET cRNA probe in transverse sections through the trunk of control (E) and mutant (F) embryos at E10.5 reveals that the migrating vagal-derived enteric nervous system precursors take the correct migration route but are arrested in the esophagal region of the foregut (fg). (G,H) In situ hybridization with Sox10 cRNA probe in transverse sections through the trunk of control (G) embryos at E11.5 visualizes migrating enteric nervous system precursors in the stomach (st) and the gut (g) (yellow arrowheads). In mutant embryos (H) the enteric nervous system precursors are not found more distal than the proximal half of the stomach. (I–N) Whole mount human dopamine ß-hydroxylase (DBH) promoter-dependent expression of ß-galactosidase in control (I,K,M) and mutant (J,L,N) gastrointestinal tracts at E18.5. Whereas in control fetuses, the whole gut contains ß-galactosidase labeled cells which are organized into interconnected ganglia, such an enteric nervous system is absent from the most of the gut. Large parts of the stomach (I,J) and, in some embryos, the rostral part of the duodenum are innervated. Caecum (K,L) and distal colon (M,N) are aganglionic. Size bars: 100 µm, except for (A) and (E) 200 µm.

 
To understand further the role of Zfhx1b in the development of the enteric nervous system (ENS), we analyzed the gastroenteric neurogenesis in our neural crest-specific mutants. First, we examined the myenteric and submucosal plexus by means of the human dopamine ß-hydroxylase (DBH) promoter-LacZ transgene (43) crossed into the Zfhx1b mutant background. Analysis at E18.5 (Fig. 3I–N) revealed severe intestinal aganglionosis that involved the distal colon (indicative for an absence of sacral neural crest derivatives) and extended past the ileocaecal junction into the small intestine. The stomach and the rostral part of the duodenum were innervated. Subsequent cell tracing using the R26R reporter allele and visualizing Sox10 (Fig. 3G and H), ErbB3 (data not shown) and cRET (Fig. 3E and F) expressing enteroblasts at E10.5 revealed neural crest cell derivatives entering the foregut where their migration is stalled around that stage. Hence, enteric precursors fail to populate massively the gastrointestinal tract beyond the distal half of the stomach (Fig. 3G and H).

The expression of Zfhx1b during heart and cardio-vascular development in neural crest-specific Zfhx1b mutants
About half of Mowat–Wilson patients display congenital heart defects. Pulmonary artery sling and stenosis, congenital tracheal stenosis, patent ductus arteriosus and atrial septal defect have been reported (22,32,44). A subpopulation of vagal neural crest cells contributes to normal cardiovascular development by supplying cells that exert the complex remodeling of the conotruncus outflow tract to create the separate pulmonic and systemic circulations (35). They also take part in the formation of the neurons and ganglia of sympathetic and parasympathetic cardiac innervation (45,46). In addition, neural crest-derived cells have been shown to populate the nascent epicardium and subsequently invade the underlying ventricular myocardium from E10.5 onwards (22).

At E10.5, Zfhx1b-positive cells are found in the cardiac outflow tract (Fig. 4A). The pattern resembles that of Wnt1–Cre-mediated R26R fate-mapped neural crest derivatives at this stage (data not shown). At E11.5 and E12.5, no Zfhx1b was found in the heart (Fig. 4B).


Figure 4
View larger version (58K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4. The expression of Zfhx1b during heart development and the cardiovascular development in Wnt1–cre, Zfhx1b mutants. (A,B) Immunohistochemistry of wild-type hearts with the antibody to Zfhx1b. (A) At E10.5, a stream of Zfhx1b-positive cells is found in the cardiac outflow tract. This location is conform the expected location of cardiac neural cells at this developmental stage. (B) At E12.5, no protein was found in the developmental heart. Note the positive cells in the mesenchyme surrounding the bronchi (green arrowhead in B). (C,D) Yolk sac appearance upon dissection at E11.5. Yolk sacs of mutants that die of cardiac failure (D) show an apparent lack of blood flow despite the presence of yolk sac blood vessels (blue arrowheads). (E,F) Mutant embryos (F) show pooling of blood in the peripheral vessels at E11.5, whereas in wild-type embryos (E), the pumping heart has cleared the blood from the body during dissection. (G,H) Wnt1–Cre-dependent ß-galactosidase synthesis from the Rosa-reporter locus (R26R) in an E11.5 wild-type (G) and a heart from a mutant embryo (H) that just died from acute heart failure. Note the large hemorrhage on the ventricle (white arrowhead in H). The migration of the neural crest cells into the mutant heart (H) outflow tract (*) is clearly visible and comparable to the control littermate’s heart (G). However, whereas neural crest-derived epicardial cells (green arrowheads) were found in the control embryos by virtue of the Wnt1–cre-mediated ß-galactosidase expression, they are absent from the mutant hearts at the moment of ventricular dysfunction. (I,J) Hearts of P0 control (I) and mutant (J) embryos. No difference in outflow tract septation is seen. fg: foregut, baa: branchial arch artery, as: aortic sac. Size bars: 50 µm (A) and 200 µm (B).

 
At E10.5–11.0, all mutants had normal gross morphology and normal intra-embryonic and yolk sac circulation. However, from E11.5, one-third of the mutant embryos revealed an anemic yolk sac despite the presence of blood vessels (Fig. 4C and D), and displayed pooling of blood in peripheral vessels, heart and liver (Fig. 4E and F). The heart was not beating suggesting that the cardiovascular defects observed in the embryo and extra-embryonic tissues are secondary to an inadequate blood flow from the heart. Since these are all symptoms of acute heart failure at E11.5–12.0, we analyzed the hearts of the affected mutants at this stage. Cell tracing experiments with the R26R reporter mice revealed an apparently normal migration of Zfhx1b-null neural crest cells into the outflow tract and no persistent truncus arteriosis was observed in any of the embryos that survive beyond E12.5 (Fig. 4I and J). Therefore, and because defects in septation of the cardiac outflow tract, like persistent truncus arteriosis, are compatible with development to term (47–49), it seems unlikely that subtle defects in this septation would cause the observed sudden death at mid-gestation. In search of the primary cause of this phenotype, we found that the hearts of E11.5 mutants with cardiovascular dysfunction often showed hemorrhage on their ventricles (Fig. 4G and H). This most likely represents a rupture of the myocardial tissue, whereby the blood is contained by the epicardium that covers the ventricular surface. Experimental interference with the development of the epicardium (50) and absence of extra-cardiac neural crest cells from the epicardium (29) leads to anomalies in the ventricular myocardium. These include reduction of the thickness of the otherwise compact and trabeculated myocardium, and small interruptions in the myocardium that sometimes result in subepicardial bleeding. Neural crest-derived epicardial cells were found in the control embryos by virtue of the Wnt1–Cre-mediated R26R recombination, but appeared absent from our Zfhx1b-null hearts at the moment of their ventricular dysfunction (Fig. 4G and H). However, histological analysis of the ventricular myocardium at E11.5 did not show significant differences in the ventricular or compact myocardium in hearts of those embryos that died of acute heart failure.

Zfhx1b is expressed in melanocytes and its absence leads to an early arrest in melanocyte development
Pigmentation defects have been reported to be associated with Mowat–Wilson syndrome in two patients so far (3,21). Melanocytes develop from a neural crest cells subpopulation that is generated late (E10.0–10.5) and whereby the migrating cells follow a dorsolateral pathway. These cells differentiate into melanoblasts after they have moved a short distance from the neural tube and start to express genes typical for the melanocytic lineage.

Zfhx1b staining in trunk sections of E11.5 embryos revealed positive cells dispersed underneath the surface ectoderm (Fig. 5A), and later (e.g. E16.5, Fig. 5B) numerous Zfhx1b-positive cells were found at the dermal–epidermal junction. This expression pattern is consistent with the location of melanoblasts at these stages. In our conditional knockouts, the early development of the melanocytic lineage was addressed by in situ hybridization for tyrosine-related protein (Trp)-2 mRNA (51) and immunostaining for the bHLH transcription factor Mitf (52). In wild-type E11.5 embryos, Trp-2 and Mitf-positive cells can be found in presumptive melanoblasts in the head and the rostral trunk region. In knockout embryos, such precursor cells were absent (Fig. 5C–F). The observation of an early block in differentiation of Zfhx1b-deficient neural crest precursors into melanoblasts was supported by the absence of ß-galactosidase-positive subcutaneous cells lateral to the neural tube when tracing recombined cells in the mutant mice at E11.5 and E12.5 (data not shown).


Figure 5
View larger version (65K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5. Zfhx1b is a melanocyte marker and its absence leads to an early arrest in melanocyte development. (A,B) Immunohistochemistry of wild-type skin with the antibody to Zfhx1b identifies positive cells dispersed underneath the surface ectoderm on rostral trunk sections of E11.5 embryos (A). At E16.5 (B) Zfhx1b-positive cells are confined at the dermal–epidermal junction, the location of melanoblasts at this developmental stage. (C,D) Immunostaining for Mitf on sections through the rostral trunk region of E11.5 embryos reveals the absence of melanoblasts in the mutant (D) embryos. (E,F) In situ hybridization with a Trp-2 cRNA probe confirms the absence of presumptive melanoblasts in the rostal trunk region of E11.5 mutant (F) embryos. Size bars: 50 µm.

 
Zfhx1b is found in the sympathetic nervous system and is essential for the development of autonomic neural crest derivatives
We also observed a developmental defect in the adrenosympathetic lineage. This phenotype is not reminiscent of clinical features observed in Mowat–Wilson patients. The autonomic nervous system is entirely derived from neural crest cells and consists of the catecholaminergic and non-catecholaminergic cells of the sympathetic nervous system (including both sympathetic ganglia and adrenal chromaffin cells), and of the cholinergic and non-cholinergic parasympathetic ganglia (see below). The sympathetic ganglia are derivatives of thoracic neural crest cells that migrate to the dorsolateral side of the dorsal aorta. In this originally continuous rostrocaudal allocation of neural crest cells, distinct swellings, the primary sympathetic ganglia, form due to locally restricted rostral migration along the dorsal aorta (53). Adrenal chromaffin cells arise from a population of primary sympathoblasts that initially localizes in the primary sympathetic chains of the embryo, and migrate ventrally towards the developing adrenal gland (54–56).

At E10.5, Zfhx1b was found associated with condensing sympathetic ganglia. Cells at the periphery of the ganglia (most likely glial cells) were displaying a strong signal (Fig. 6A and B). To address developmental abnormalities in the autonomic nervous system, we examined the interlimb region of E10.5 embryos where the formation of the sympathetic chain has just begun. In the mutant primary sympathetic chain, we found a regular pattern of clustered ß-galactosidase-positive cells, but compared to those in (conditional) heterozygotes, they were lower in number and more loosely organized (Fig. 6C–F). Neural differentiation capacities of the sympathoblasts appeared to be maintained since the early neural markers SCG10 (57) and Isl1 (58) were present in aggregating sympathoblasts (data not shown), as well as tyrosine hydroxylase (TH; data not shown) and dopamine-ß-hydroxylase (Fig. 6G and H), which control noradrenaline synthesis in the sympathetic neurons. Development of the glial lineage in the sympathetic ganglia anlage was addressed by staining for ErbB3, Sox10 and Cadherin6 (59). Some cells in the loosely aggregated sympathetic clusters appeared positive for these early Schwann cell precursor markers (Fig. 6I and J, and data not shown).


Figure 6
View larger version (89K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 6. Zfhx1b is expressed in the sympathetic nervous system and is essential for the development of autonomic neural crest derivatives. (A) Immunohistochemistry at E10.5 reveals Zfhx1b expression associated with condensing sympathetic ganglia (arrows) at the dorsolateral side of the dorsal aorta (da). (B) Enlargement of nascent sympathetic ganglia at E10.5. Double fluorescent immunostaining reveals a strong nuclear staining for Zfhx1b (red) in cells, likely glial cells, which are in close association with the tyrosine hydroxylase (TH)-positive neurons, visualized in green. Nuclei are stained blue by DAPI. (C–F) Visualization of Wnt1–Cre-dependent ß-galactosidase synthesis from the R26R in E10.5 embryos. Sections along the midline dorsal aorta in the interlimb region of E10.5 embryo's (C,D) and a higher magnification of a transverse section through the dorsal aorta in the rostral part of the trunk (E,F). The primary sympathetic ganglionic condensations (green arrowheads) along the midline dorsal aorta are, though present, greatly reduced in size and more loosely organized in the mutant (D,F) embryos when compared with the control sections (C,E). (G,H) In situ hybridization for dopamine-ß-hydroxylase reveals the neural differentiation capacity of the sympathoblasts is maintained in the mutant (H) sympathetic anlage. (I,J) In situ hybridization with anti-Sox10 cRNA in aggregating sympathoblasts in the interlimb region of E10.5 embryos reveals glial cells enveloping the primary loosely aggregated sympathetic ganglia in mutant (J) embryos. (K,L) Immunostaining for PNMT (phenylethanol amine-N-methyl transferase) on E17.5 adrenal glands. Within the mutant adrenal medulla (L), the number of chromaffin cells is significantly reduced when compared with control adrenal glands. Nt: neural tube. Size bars: 50 µm (A,E,G) and 100 µm (K) and 200 µm (C).

 
The number of phenylethanol amine-N-methyl transferase (PNMT) (60), and TH-positive cells in the E17.5 adrenal gland were significantly reduced (Fig. 6K and L; data not shown), but—similar to the sympathetic chain—some chromaffin cells remained in the adrenal medulla of the neural crest-specific knockouts as if their correctly positioned precursors failed to proliferate.

Zfhx1b is found in the cranial sensory ganglia but is redundant for their initial development
In contrast to the glial component of cranial ganglia, which originate purely from neural crest cells, neurons in cranial ganglia have a dual origin. Neurons in the VI, IX and X nerve ganglion are of neural crest origin, whereas the neurons from the trigeminal (V) nerve originate both from neural crest cells and placodal ectoderm (61,62). The neural crest cells that participate in the formation of the cranial ganglia arise from a level encompassing the caudal midbrain and hindbrain rhombomeres 1–3 (63).

We found abundant quantities of Zfhx1b in the cranial sensory ganglia from the moment they condense at E9.5. Similar as to what has been described by in situ hybridization (64), nuclear staining was found in all cells of the trigeminal (V), the facio-acoustic (VII,VIII) ganglia and in the ganglia of the glossopharyngeal (IX) and vagal (X) nerves from E10.5 (Fig. 7A).


Figure 7
View larger version (58K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 7. Zfhx1b is expressed in cranial sensory ganglia and not essential for their initial development. (A) Transversal section through the head of an E10.5 embryo reveals high nuclear staining in the trigeminal ganglion. (B,C) Immunostaining for Brn3A on a transversal section through an E10.5 trigeminal ganglion identifies early sensory neuron precursors in the control embryos (B) as well as in the mutant (C). (D,E) In situ hybridization for ErbB3 reveals glial cell committed precursors scattered within and around the mutant trigeminal ganglion (E), similar to what is seen in the control littermate. (F,G) Immunostaining for B-FABP on a transversal section through an E11.5 trigeminal ganglion identifies satellite glial cells in the control embryos (F) as well as in the mutant (G). (H–K) Whole mount immunostainings for neurofilament-160. (H,I) The outgrowth of neurofilament-160-positive cranial nerves was apparently normal at E10.5. (J,K) At E11.5, the central trigeminal axon terminals showed a reduced innervation at the vibrissal (whisker) follicles in the mystacial pad of the mutants. nt: neural tube, g: geniculate ganglion, n: nodosal ganglion, p: petrosal ganglion, t: trigeminal ganglion, v: vestibulocochlear ganglion.

 
The Zfhx1b-deficient parasympathetic ganglia appeared morphologically normal at E10.5 and contained both neuronal and glial precursor cells, as analyzed by staining of Brn3A (Fig. 7B and C), and respectively, ErbB3 (Fig. 7D and E) and B-FABP (Fig. 7F and G), and despite the second wave of NeuroD expression in these ganglia at E10.5 was found to be reduced significantly (Van de Putte, Francis, Seuntjens, Desmet, Higashi, Kondoh and Huylebroeck, manuscript in preparation). The outgrowth of neurofilament-160-positive cranial nerves was apparently normal at E10.5 (Fig. 7H and I). At E11.5, however, central trigeminal axon terminals showed a reduced innervation at the vibrissal (whisker) follicles in the mystacial pad of the mutants (Fig. 7J and K).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
The aim of this study was to document and understand the relationships between the presence of Zfhx1b in neural crest cells and their derivatives, and the clinical heterogeneous features associated with Mowat–Wilson syndrome. Indeed, in the 5 years that have passed since the identification of mutations in ZFHX1B as cause of Mowat–Wilson syndrome, several clinical reports have added to the most prevalent clinical features (mental retardation, microcephaly, distinct facial dysmorfisms) several associated congenital anomalies (Hirschsprung disease, cardiac defects, anomalies of the hair, skeleton and urogenital systems) (2,3,14,16,18,65–68). No animal model existed so far to study experimentally the etiology of the neurocristopathy associated with ZFHX1B mutations. A comprehensive morphological phenotyping of adult heterozygous mouse mutants (our unpublished data) did not reveal apparent defects in the peripheral tissues shown to be affected in Mowat–Wilson patients. Studies in homozygous knockout embryos however have revealed, despite their lethality at early post-gastrulation (E8.5, which roughly corresponds to 4–5 weeks in human), the importance of Zfhx1b for the migration of cranial and induction of vagal neural crest cells as well as for early neurulation (4) and somitogenesis (26). In order to obtain insight of how far neural crest cells are responsible for the observed clinical features in human patients, we removed Zfhx1b specifically from neural crest cell precursors and analyzed in vivo the developmental consequences for the tissues in which they participate.

Patients bearing ZFHX1B mutations typically have microcephaly, hypertelorism, medially flared and broad eyebrows, prominent columella, a rounded nasal tip, pointed chin and large uplifted earlobes (66,67). The cranial phenotype in our conditional mouse model that lacks Zfhx1b in the cephalic neural crest derivatives is consistent with this facial dysmorfism. In particular the sagital and metopic sutures, the nasal bone, the nasal mesenchyme and the mandibles were affected in our mutant mice, as well as bones around the ears. The eyelids were open. This may indirectly be caused by developmental defects in the eye itself, since neural crest-derived cells contribute to the corneal endothelial cells, the collagen-synthesizing keratocytes and the melanocytes of the iris (39). Alternatively, this may result from a differential implantation of the eye in the skull due to subtle alterations of the facial bones. More likely however, it reflects a defect in periocular mesenchyme that is derived from the neural crest as well (39,69).

Whereas, Zfhx1b is present in all skeletogenic neural crest cells when they leave their site of origin, the defects in the mutant head are confined to specific neural crest cell derivatives and no apparent defect is observed within 2 days after the first recombined cranial neural crest cells start migrating into the frontonasal prominences and the branchial arches. This suggests that the earliest events such as neural crest cell generation and migrations are not affected upon removing Zfhx1b from cranial neural crest precursors using the Wnt1–Cre line. Once arrived at their destination, Zfhx1b synthesis ceases in all cranial neural crest derivatives but re-emerges 2 days later in the frontonasal region at E11.5. The latter tissue-specific expression is highly correlated with the naso-maxillary bones that are affected in the conditional mutant embryos. We therefore suppose that the deformities in the mouse facial bones are due to the requirement for Zfhx1b activity in the facial primordia from E11.5 onwards and not because of defective migration capabilities of the conditional Zfhx1b-null neural crest cells before that stage.

During facial development, a number of organs like teeth and whisker follicles develop as a result of sequential and reciprocal interactions between the oral ectoderm and cranial neural crest-derived mesenchyme (70–72). Interestingly, Zfhx1b was never detected in epithelial tissue like the facial epithelium or the invaginating molar and whisker follicle epithelium but was always confined to mesenchymal cells. High Zfhx1b levels were seen in the condensing mesenchyme that is in contact with the epithelium like those of the molar epithelium and the whisker follicle invaginations. Interestingly, several studies have shown that ligands of the TGF-ß family are involved in these morphogenetic processes and/or are abundantly expressed in the epithelial component of, for example, the whisker follicles and tooth bud (73,74). Their receptors are often confined to the mesenchyme during these critical interactions. Given the established interaction between Zfhx1b and receptor-activated Smads (6), the particular expression pattern described here, the absence of molars and whisker follicles in the conditional knockout mouse and the teeth defects in some individuals with Mowat–Wilson syndrome supports a role of Zfhx1b in TGF-ß family signaling during these morphogenetic events.

Conditional removal of Zfhx1b from the vagal and sacral enteric neural crest precursors leads to an aganglionic small and large intestine, thereby underlining the importance of Zfhx1b as a Hirschsprung disease susceptibility gene. Despite the absence of enteric ganglia from the midgut and hindgut of the knockout embryos, significant numbers of neurons and glial cells develop in the proximal stomach of these animals. This contests against the observations made in the conventional knockout in which vagal neural crest precursors are not induced at all. One explanation may be that Zfhx1b’s role in the induction of vagal neural crest cells is non-cell-autonomous and is caused by the malfunctioning of the adjacent neural tube in the total knockout. Alternatively, vagal neural crest cells in the conditional knockout may already have been induced before the onset of Wnt1-promoter-controlled Cre synthesis and thereby retain enough Zfhx1b protein to populate the foregut where a disturbed Zfhx1b-controlled equilibrium between proliferation and differentiation would subsequently prevent them from colonizing the entire intestinal tract. However, the phenotype we observe has a striking similarity with the phenotypes in GDNF, Ret, Gfr{alpha}1 and Pax3 knockout mice, which show a reduced neuron and glial cell number in the esophagus and proximal stomach and have no enteric ganglia in the midgut and hindgut (41,75,76). Surgical ablation of the vagal neural crest cells in chick embryos similarly results in the elimination of the mid- and hindgut ENS but has minimal effects on the development of the foregut ENS (77). In line with this, cell tracing experiments have shown that the anterior vagal neural crest cells participate in the innervation along the whole gastrointestinal tract, whereas the derivatives of the most caudal vagal neural crest cells are only found in the foregut derivatives (75,78). It is therefore possible that Zfhx1b is essential for the correct differentiation of the anterior vagal neural crest cells, whereas a Zfhx1b-independent lineage of neural crest cells contributes to the formation of the foregut ENS in mouse embryos.

A considerable percentage of mutant embryos die at midgestation from acute heart failure, whereas the majority develops to term, with no obvious cardiac outflow septation defects. This suggests that Zfhx1b is dispensable for neural crest cell colonization of the septating truncus arteriosus in the mouse, a process in which BMP signaling has been shown necessary (79,80). A small neural crest cell-derived population of cells in the epicardium has recently been identified and shown to be involved in myocardial development by promoting ventricular proliferation. A Wnt1–Cre-mediated ablation has shown these cells to depend on signaling via the receptor Bmpr1a (Alk3). The fact that the Zfhx1b-deficient mutants die of acute heart failure at the same moment as the Alk3 knockouts, and epicardial cells appear absent from our mutants, suggests Zfhx1b to act in the same cascade. In contrast to the study in the Alk3 mutants (29), we could not observe a significant difference in thickness of the ventricular myocardium in our mutants. Yet, the presence of blood-filled blisters on the mutant hearts suggests a defect that impairs the intactness of the myocardium. Further studies are expected to reveal the role of these epicardial cells in the development of the heart and their role in the etiology of heart anomalies in human patients, in particular Mowat–Wilson patients.

Zfhx1b is present in melanoblasts and persists in melanocytes. Although pigmentation defects have only been associated with two Mowat–Wilson patients, the complete removal of Zfhx1b from neural crest precursors appears deleterious for melanocyte development from the earliest stages of their development. Other trunk derivatives, the sympathetic ganglia and chromaffin cells in the adrenal ganglia, appear to be affected as well upon complete depletion of Zfhx1b from neural crest. Our observations suggest that Zfhx1b-null sympatho-adrenal precursor cells are able to follow the correct migratory routes and do not loose their commitment to aggregate in ganglionic anlagen that consist of glial cells and functional neural precursors. However, Zfhx1b absence leads to abnormal proliferation of the primary sympathetic ganglia that aggregate at the dorsal aorta. Interestingly, sympatho-adrenal precursors contain Bmpr1a and Bmpr1b (81), and BMP-2/4/7 have been shown to increase the number of catecholaminergic and neuronal cells in cultures of avian neural crest cells (82–84). Therefore, if the same is true for mammalian embryos, cellular interpretation of this BMP signal by the sympatho-adrenal precursors may depend on the transcriptional activity of Zfhx1b.

Inspite of the strong nuclear Zfhx1b staining from the onset of ganglia aggregation, the initial development of the neuronal and glial compartment in the cranial ganglia as well as the initial axonal outgrowth was apparently normal. This contrasts with the Zfhx1b-deficient dorsal root ganglia (DRG), which show a more severe phenotype (Van de Putte, Francis, Seuntjens, Desmet, Higashi, Kondoh and Huylebroeck, manuscript in preparation). In this respect, it is striking that its paralogue Zfhx1a ({delta}EF1) is co-expressed in the cranial ganglia at this developmental stage, whereas the onset of expression of Zfhx1a in the DRG is only at E11.5 (64), which is one day after we observe defects like the axonal outgrowth defect. Given the significant similarity (6) and biological activity (85), it may well be that the presence of Zfhx1a in cranial ganglia compensates for the loss of Zfhx1b.

Taken together, Zfhx1b is expressed throughout the neural crest cells and its removal from the whole embryo leads to an early arrest of neural crest cell development (4). Our observations suggest this transcription factor to be dispensable for many cell autonomous aspects of the consecutive development of particular neural crest cell derivatives, like many cranial bones, the neural crests involved in stomach innervation and heart septation, and the initial development of sympathetic and parasympathetic nervous system. At first sight, this discrepancy between the dramatic neural crest phenotype in the conventional and the milder phenotype in the conditional knockout may reflect a necessity for this transcription factor during the earliest phases of neural crest induction on the neural ridge, an event that relays on the reciprocal signaling between the surface ectoderm and the neural plate (27). Our current experiments involving conditional Zfhx1b ablation are unable to assess the necessity for Zfhx1b prior to 5–6 somite stage, when Wnt1–Cre-mediated recombination becomes prevalent in the formation of neural crest cells. However, given the considerable time lag between the first synthesis of Wnt1 promoter-driven Cre in the neural crest precursors and the appearance of the developmental abnormalities, it seems very unlikely that the differential consequences could simply be brought about by a variable persistence of the Zfhx1b protein after the gene has been ablated in particular neural crest cell precursors. In addition, a cell type-specific role for Zfhx1b in different neural crest precursors in vivo is also supported by the fact that the cranial and vagal neural crests in the conventional knockout show an arrest in their development at two different developmental stages (neural crest induction and delamination, respectively).

The phenotypes of this conditional mutant mouse line provide strong clues to refine clinical examination of Mowat–Wilson patients. Indeed, in addition to the clinically recognized alteration in craniofacial appearance and the developmental defects in the enteric nervous system, we propose to consider the performance of the heart myocardium, the development and the activity of the adrenomedullary system as well as the functioning of the sensory system (Van de Putte, Francis, Seuntjens, Desmet, Higashi, Kondoh and Huylebroeck, manuscript in preparation) and the sympathetic chain in the clinical management of this genetic disorder.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Mice
The constitutive knockout allele for Zfhx1b was Zfhx1b{Delta}ex7 and the floxed allele Zfhx1bflox(ex7) (15) was back-crossed (greater than eight generations) and maintained in the CD1 background. The (conditional) deletion leads to a frameshift and premature termination of the putative protein at the level of the fourth zinc finger in the N-terminal zinc finger cluster (15). Neural crest-specific knockout of the Zfhx1b gene was obtained by mating mice homozygous for the floxed Zfhx1b allele (Zfhx1bflox(ex7)/flox(ex7)) with compound heterozygotes for Wnt1–Cre and the conventional knockout Zfhx1b allele (Wnt1–Cre+/–, Zfhx1b{Delta}ex7/+), the latter mice back-crossed once or twice in CD1. Crosses generated four types of genotypes: among them [Wnt1–Cre+/–, Zfhx1bflox(ex7)/KO] animals, which were homozygous null for the Zfhx1b gene in neural crest cells and their derivatives, and heterozygous in other tissues. Littermate embryos, heterozygous for Zfhx1b in neural crest cells only [Wnt1–Cre+/–, Zfhx1bflox(ex7)/+] or heterozygous in all cells [Zfhx1bflox(ex7)/KO] were used as control. Wnt1–Cre mice (28) were kindly provided by A. McMahon (Harvard University). The R26R reporter (86) and DBH–lacZ mice (43) were obtained from the Jackson Laboratories. Embryos were fixed overnight in 4% paraformaldehyde in phosphate-buffered saline at 4°C. Mice were handled according to the guidelines of the Committee for Animal Experiments of the University of Leuven (Belgium).

Genotyping of mice
The genotype of mice and embryos was determined by PCR analysis of genomic DNA from a tail tip fragment or yolk sac, respectively. Fragments were lysed in 10 mM Tris–HCl (pH 8.4), 50 mM KCl, 2.5 mM MgCl2, 0.5% NP40, 0.5% Tween 20, 0.02% SDS, 0.5 mg/ml proteinase K at 55°C for 3–12 h. Lysates were treated at 99°C for 15 min to inactivate proteinase K and subjected to PCR analysis with denaturation at 94°C for 3 min, amplification for 35 cycles at 96°C for 30 s, 64.5°C for 30 s and 72°C for 1 min using the following primers: Intron6-Fwd primer, 5'-GAACTAGTTGAATTGGTAGAATCAATGGGG-3'; Intron6-Rev primer, 5'-GTAAAGGCTCTCTACGCCTTTTTCAGTTAG-3'; Intron7-Rev primer, 5'-AAGCATGTCGGTAAGCTGACCAACTACTAG- 3'. The PCR product diagnostic for the wild-type Zfhx1b and Zfhx1bflox(ex7)/+ allele was 150 and 256 bp, respectively (between intron6-Fwd and intron6-Rev primers), and 288 bp for the Zfhx1b{Delta}ex7 allele (between intron6-Fwd and intron7-Rev). Primers used to reveal the presence of the Cre-encoding cassette were Cre-Fwd 5'-TGCCACGACC-AAGTGACAGCAATG-3' and Cre-Rev 5'-ACCAGAGACG-GAAATCCATCGCTC-3'. Identical PCR conditions were used apart from an annealing temperature of 58°C, yielding a PCR product of about 400 bp. Genotyping for the presence of the lacZ gene (DBH–lacZ and R26R mice) was done using the primers and conditions described previously (87).

Plasmids and in situ hybridization
Chromogenic ISH on paraffin sections was done with antisense riboprobes labeled with digoxigenin-UTP (Roche) on the Ventana DiscoveryTM automated in situ hybridization instrument (Ventana Medical Systems, Tucson, AZ, USA) using the Ribomap and Bluemap kit (Ventana Medical Systems). Pre-treatment, hybridization temperature, probe concentration, post-hybridization wash stringency and color development were dependent on the probes and embryonic stages analyzed. Detailed procedures are available on simple request. Probe templates were gifts from D.J. Anderson (NeuroD), C. Birchmeier (ErbB3), M. Takeichi (Cadherin-6), J.Golden (cRET), B. Wehrle-Haller (TRP2) and J.F. Brunet (DBH). A PCR fragment containing 1194–2064 bp of MMU66141 (NCBI) was used as template for the cRNA probe of mSox10.

Immunostaining
For IHC on paraffin sections, tissue samples were processed using the Ventana DiscoveryTM instrument using the DAB map kit (Ventana Medical Systems). Antigen retrieval conditions and antibody dilutions are available on request. For fluorescent immunohistochemistry, Alexa488-conjugated goat anti-rabbit (Molecular Probes) and Cy3-conjugated donkey anti-rabbit IgG's (Jackson Immunoresearch Laboratories) were used as secondary antibodies. Sections were preserved and stained for nuclei in Mowiol containing DAPI. Whole-mount immunostaining: Endogenous peroxidase activity of fixed embryos was inactivated with 3% H2O2 in Dent's fixative [80% MeOH and 20% dimethylsulfoxide (DMSO)] for 3 h at 4°C. After washing with Tris-buffered saline containing 1% Tween-20 (TBST) for 2 h, the embryos were blocked for at least 1 h in TBST containing 5% (w/v) skimmed milk. Next, embryos were incubated for 3 days at 24°C with a 1:100 dilution of concentrated 2H3 hybridoma culture supernatant in 5% skim milk–TBST containing 5% DMSO, and 0.1% sodium azide. Embryos were then washed (eight times, 1 h) in TBST, blocked again for 1 h in 5% skim milk–TBST and incubated for 2 days at 4°C with peroxidase-conjugated goat-anti-mouse IgG antibody (Jackson Immunoresearch Laboratories, 1:200 dilution in 5% skim milk–TBST, 5% DMSO). Embryos were washed five times for 30 min in TBST. Then, embryos were pre-equilibrated in staining solution for 30 min without H2O2 after which peroxidase deposits were visualized with 4-chloro-1-Naphtol solution (Sigma–Aldrich) according to the manufacturer's instructions. Embryos were cleared in 80% glycerol. The anti-NCL-TH and NCL-L-MITF antibodies were obtained from Novocastra Laboratories Ltd. (Newcastle-upon-Tyne, UK) and the anti-PNMT antibody from Acris antibodies GmbH (Hiddenhausen, Germany). The antibodies against Brn3A and B-FABP (Kurtz et al., 1994) were gifts from E. Turner and T. Müller, respectively. The 2H3 neurofilament-specific antibody, developed by T. Jessell and J. Dodd, was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA, USA.

Skeletal preparations
Alcian blue/alizarin red staining of cartilage and bone was done as described (88).


   ACKNOWLEDGEMENTS
 
The authors are grateful to L. Desmet and M. Missoul for animal care. We appreciate the stimulating discussions with Drs H. Kondoh and Y. Higashi. We thank Drs A. McMahon, D. J. Anderson, C. Birchmeier, J. Golden, M. Takeichi, B. Wehrle-Haller, J. F. Brunet, E. Turner and T. Müller for providing the Wnt1–Cre mouse line and sharing reagents. Grant sponsors for the D.H. Laboratory were FWO-V (G.0279.04 and G.0288.07), Interuniversity Attraction Pole Network 5/35 and 6/20 and the University of Leuven (0T/00/41).

Conflict of Interest statement. The authors declare that they have no financial conflict of interests.


   FOOTNOTES
 
{dagger} Present address: Galapagos S.A., Generaal De Wittelaan L11, B-2800 Mechelen, Belgium. Back

{ddagger} Present address: Cell Biology and Histology (CYTO), Faculty of Medicine and Pharmacy, Vrije Universiteit Brussel, Laarbeeklaan 103, B-1090 Jette, Belgium. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Le Douarin N.M., Kalcheim C. The Neural Crest (1999) 2nd ed. London: Cambridge University Press.

  2. Wakamatsu N., Yamada Y., Yamada K., Ono T., Nomura N., Taniguchi H., Kitoh H., Mutoh N., Yamanaka T., Mushiake K., et al. Mutations in SIP1, encoding Smad interacting protein-1, cause a form of Hirschsprung disease. Nat. Genet. (2001) 27:369–370.[CrossRef][ISI][Medline]

  3. Cacheux V., Dastot-Le Moal F., Kaariainen H., Bondurand N., Rintala R., Boissier B., Wilson M., Mowat D., Goossens M. Loss-of-function mutations in SIP1 Smad interacting protein 1 result in a syndromic Hirschsprung disease. Hum. Mol. Genet. (2001) 10:1503–1510.[Abstract/Free Full Text]

  4. Van de Putte T., Maruhashi M., Francis A., Nelles L., Kondoh H., Huylebroeck D., Higashi Y. Mice lacking ZFHX1B, the gene that codes for Smad-interacting protein-1, reveal a role for multiple neural crest cell defects in the etiology of Hirschsprung disease-mental retardation syndrome. Am. J. Hum. Genet. (2003) 72:465–470.[CrossRef][ISI][Medline]

  5. van Grunsven L.A., Michiels C., Van de Putte T., Nelles L., Wuytens G., Verschueren K., Huylebroeck D. Interaction between Smad-interacting protein-1 and the corepressor C-terminal binding protein is dispensable for transcriptional repression of E-cadherin. J. Biol. Chem. (2003) 278:26135–26145.[Abstract/Free Full Text]

  6. Verschueren K., Remacle J.E., Collart C., Kraft H., Baker B.S., Tylzanowski P., Nelles L., Wuytens G., Su M.T., Bodmer R., et al. SIP1, a novel zinc finger/homeodomain repressor, interacts with Smad proteins and binds to 5'-CACCT sequences in candidate target genes. J. Biol. Chem. (1999) 274:20489–20498.[Abstract/Free Full Text]

  7. van Grunsven L.A., Taelman V., Michiels C., Opdecamp K., Huylebroeck D., Bellefroid E.J. deltaEF1 and SIP1 are differentially expressed and have overlapping activities during Xenopus embryogenesis. Dev. Dyn. (2006) 235:1491–1500.[CrossRef][ISI][Medline]

  8. Sekido R., Murai K., Kamachi Y., Kondoh H. Two mechanisms in the action of repressor deltaEF1: binding site competition with an activator and active repression. Genes Cells (1997) 2:771–783.[Abstract]

  9. Yoshimoto A., Saigou Y., Higashi Y., Kondoh H. Regulation of ocular lens development by Smad-interacting protein 1 involving Foxe3 activation. Development (2005) 132:4437–4448.[Abstract/Free Full Text]

  10. Comijn J., Berx G., Vermassen P., Verschueren K., van Grunsven L., Bruyneel E., Mareel M., Huylebroeck D., van Roy F. The two-handed E box binding zinc finger protein SIP1 downregulates E-cadherin and induces invasion. Mol. Cell (2001) 7:1267–1278.[CrossRef][ISI][Medline]

  11. Remacle J.E., Kraft H., Lerchner W., Wuytens G., Collart C., Verschueren K., Smith J.C., Huylebroeck D. New mode of DNA binding of multi-zinc finger transcription factors: deltaEF1 family members bind with two hands to two target sites. EMBO J. (1999) 18:5073–5084.[CrossRef][ISI][Medline]

  12. Dillner N.B., Sanders M.M. Transcriptional activation by the zinc-finger homeodomain protein delta EF1 in estrogen signaling cascades. DNA Cell Biol. (2004) 23:25–34.[CrossRef][ISI][Medline]

  13. Lazarova D.L., Bordonaro M., Sartorelli A.C. Transcriptional regulation of the vitamin D(3) receptor gene by ZEB. Cell Growth Differ. (2001) 12:319–326.[Abstract/Free Full Text]

  14. Amiel J., Espinosa-Parrilla Y., Steffann J., Gosset P., Pelet A., Prieur M., Boute O., Choiset A., Lacombe D., Philip N., et al. Large-scale deletions and SMADIP1 truncating mutations in syndromic Hirschsprung disease with involvement of midline structures. Am. J. Hum. Genet. (2001) 69:1370–1377.[CrossRef][ISI][Medline]

  15. Higashi Y., Maruhashi M., Nelles L., Van De Putte T., Verschueren K., Miyoshi T., Yoshimoto A., Kondoh H., Huylebroeck D. Generation of the floxed allele of the SIP1 (Smad-interacting protein 1) gene for Cre-mediated conditional knockout in the mouse. Genesis (2002) 32:82–84.[CrossRef][ISI][Medline]

  16. Wilson M., Mowat D., Dastot-Le Moal F., Cacheux V., Kaariainen H., Cass D., Donnai D., Clayton-Smith J., Townshend S., Curry C., et al. Further delineation of the phenotype associated with heterozygous mutations in ZFHX1B. Am. J. Med. Genet. (2003) 119A:257–265.

  17. Zweier C., Temple I.K., Beemer F., Zackai E., Lerman-Sagie T., Weschke B., Anderson C.E., Rauch A. Characterisation of deletions of the ZFHX1B region and genotype-phenotype analysis in Mowat–Wilson syndrome. J. Med. Genet. (2003) 40:601–605.[Free Full Text]

  18. Yoneda M., Fujita T., Yamada Y., Yamada K., Fujii A., Inagaki T., Nakagawa H., Shimada A., Kishikawa M., Nagaya M., et al. Late infantile Hirschsprung disease-mental retardation syndrome with a 3-bp deletion in ZFHX1B. Neurology (2002) 59:1637–1640.[Abstract/Free Full Text]

  19. Heinritz W., Zweier C., Froster U.G., Strenge S., Kujat A., Syrbe S., Rauch A., Schuster V. A missense mutation in the ZFHX1B gene associated with an atypical Mowat–Wilson syndrome phenotype. Am. J. Med. Genet. A (2006) 140:1223–1227.[Medline]

  20. Gregory-Evans C.Y., Vieira H., Dalton R., Adams G.G., Salt A., Gregory-Evans K. Ocular coloboma and high myopia with Hirschsprung disease associated with a novel ZFHX1B missense mutation and trisomy 21. Am. J. Med. Genet. A (2004) 131:86–90.[Medline]

  21. Dastot-Le Moal F., Wilson M., Mowat D., Collot N., Niel F., Goossens M. ZFHX1B mutations in patients with Mowat–Wilson syndrome. Hum. Mutat. (2007) 28:313–321.[CrossRef][ISI][Medline]

  22. Zweier C., Thiel C.T., Dufke A., Crow Y.J., Meinecke P., Suri M., Ala-Mello S., Beemer F., Bernasconi S., Bianchi P., et al. Clinical and mutational spectrum of Mowat–Wilson syndrome. Eur. J. Med. Genet. (2005) 48:97–111.[CrossRef][ISI][Medline]

  23. Lindsay E.A., Botta A., Jurecic V., Carattini-Rivera S., Cheah Y.C., Rosenblatt H.M., Bradley A., Baldini A. Congenital heart disease in mice deficient for the DiGeorge syndrome region. Nature (1999) 401:379–383.[CrossRef][Medline]

  24. Jackson I.J. Homologous pigmentation mutations in human, mouse and other model organisms. Hum. Mol. Genet. (1997) 6:1613–1624.[Abstract/Free Full Text]

  25. Schuchardt A., D’Agati V., Larsson-Blomberg L., Costantini F., Pachnis V. Defects in the kidney and enteric nervous system of mice lacking the tyrosine kinase receptor Ret. Nature (1994) 367:380–383.[CrossRef][Medline]

  26. Maruhashi M., Van De Putte T., Huylebroeck D., Kondoh H., Higashi Y. Involvement of SIP1 in positioning of somite boundaries in the mouse embryo. Dev. Dyn. (2005) 234:332–338.[CrossRef]