During mammalian evolution, expansion of the cerebral hemispheres was accompanied by expansion of the frontal and parietal bones of the skull vault and deployment of the coronal (fronto-parietal) and sagittal (parietal-parietal) sutures as major growth centres. Using a transgenic mouse with a permanent neural crest cell lineage marker, Wnt1-Cre/R26R, we show that both sutures are formed at a neural crest-mesoderm interface: the frontal bones are neural crest-derived and the parietal bones mesodermal, with a tongue of neural crest between the two parietal bones. By detailed analysis of neural crest migration pathways using X-gal staining, and mesodermal tracing by DiI labelling, we show that the neural crest-mesodermal tissue juxtaposition that later forms the coronal suture is established at E9.5 as the caudal boundary of the frontonasal mesenchyme. As the cerebral hemispheres expand, they extend caudally, passing beneath the neural crest-mesodermal interface within the dermis, carrying with them a layer of neural crest cells that forms their meningeal covering. Exposure of embryos to retinoic acid at E10.0 reduces this meningeal neural crest and inhibits parietal ossification, suggesting that intramembranous ossification of this mesodermal bone requires interaction with neural crest-derived meninges, whereas ossification of the neural crest-derived frontal bone is autonomous. These observations provide new perspectives on skull evolution and on human genetic abnormalities of skull growth and ossification.

Neural crest cells are multipotential stem cells that contribute extensively to vertebrate development and give rise to various cell and tissue types. Determination of the fate of mammalian neural crest has been inhibited by the lack of appropriate markers. Here, we make use of a two-component genetic system for indelibly marking the progeny of the cranial neural crest during tooth and mandible development. In the first mouse line, Cre recombinase is expressed under the control of the Wnt1 promoter as a transgene. Significantly, Wnt1 transgene expression is limited to the migrating neural crest cells that are derived from the dorsal CNS. The second mouse line, the ROSA26 conditional reporter (R26R), serves as a substrate for the Cre-mediated recombination. Using this two-component genetic system, we have systematically followed the migration and differentiation of the cranial neural crest (CNC) cells from E9.5 to 6 weeks after birth. Our results demonstrate, for the first time, that CNC cells contribute to the formation of condensed dental mesenchyme, dental papilla, odontoblasts, dentine matrix, pulp, cementum, periodontal ligaments, chondrocytes in Meckel's cartilage, mandible, the articulating disc of temporomandibular joint and branchial arch nerve ganglia. More importantly, there is a dynamic distribution of CNC- and non-CNC-derived cells during tooth and mandibular morphogenesis. These results are a first step towards a comprehensive understanding of neural crest cell migration and differentiation during mammalian craniofacial development. Furthermore, this transgenic model also provides a new tool for cell lineage analysis and genetic manipulation of neural-crest-derived components in normal and abnormal embryogenesis.

Hhex is required for early development of the liver. A null mutation of Hhex results in a failure to form the liver bud and embryonic lethality. Therefore, Hhex null mice are not informative as to whether this gene is required during later stages of hepatobiliary morphogenesis. To address this question, we derived Hhex conditional null mice using the Cre-loxP system and two different Cre transgenics (Foxa3-Cre and Alfp-Cre). Deletion of Hhex in the hepatic diverticulum (Foxa3-Cre;Hhex(d2,3/-)) led to embryonic lethality and resulted in a small and cystic liver with loss of Hnf4alpha and Hnf6 expression in early hepatoblasts. In addition, the gall bladder was absent and the extrahepatic bile duct could not be identified. Loss of Hhex in the embryonic liver (Alfp-Cre;Hhex(d2,3/-)) caused irregular development of intrahepatic bile ducts and an absence of Hnf1beta in many (cystic) biliary epithelial cells, which resulted in a slow, progressive form of polycystic liver disease in adult mice. Thus, we have shown that Hhex is required during multiple stages of hepatobiliary development. The altered expression of Hnf4alpha, Hnf6 and Hnf1beta in Hhex conditional null mice suggests that Hhex is an essential component of the genetic networks regulating hepatoblast differentiation and intrahepatic bile duct morphogenesis.

A subpopulation of neural crest termed the cardiac neural crest is required in avian embryos to initiate reorganization of the outflow tract of the developing cardiovascular system. In mammalian embryos, it has not been previously experimentally possible to study the long-term fate of this population, although there is strong inference that a similar population exists and is perturbed in a number of genetic and teratogenic contexts. We have employed a two-component genetic system based on Cre/lox recombination to label indelibly the entire mouse neural crest population at the time of its formation, and to detect it at any time thereafter. Labeled cells are detected throughout gestation and in postnatal stages in major tissues that are known or predicted to be derived from neural crest. Labeling is highly specific and highly efficient. In the region of the heart, neural-crest-derived cells surround the pharyngeal arch arteries from the time of their formation and undergo an altered distribution coincident with the reorganization of these vessels. Labeled cells populate the aorticopulmonary septum and conotruncal cushions prior to and during overt septation of the outflow tract, and surround the thymus and thyroid as these organs form. Neural-crest-derived mesenchymal cells are abundantly distributed in midgestation (E9.5-12.5), and adult derivatives of the third, fourth and sixth pharyngeal arch arteries retain a substantial contribution of labeled cells. However, the population of neural-crest-derived cells that infiltrates the conotruncus and which surrounds the noncardiac pharyngeal organs is either overgrown or selectively eliminated as development proceeds, resulting for these tissues in a modest to marginal contribution in late fetal and postnatal life.

Most developing structures that express the transcription factor gene AP-2alpha are compromised in AP-2alpha mutant mouse embryos. Since the cardiac neural crest population is one prominent site of AP-2alpha expression, and because the neural crest is known to be required for normal cardiac morphogenesis, we have investigated the involvement of AP-2alpha in cardiac development. All AP-2alpha-deficient embryos examined had malformations of the outflow tract of the developing heart: most had double outlet right ventricle, and a small fraction had persistent truncus arteriosus. To visualize AP-2alpha-expressing cells during the period of cardiac morphogenesis, we established a new mutant germline allele in which an IRES-lacZ sequence was inserted by homologous recombination into the AP-2alpha locus. Positive expression was observed in the cardiac neural crest population during the E9.5-10.5 period (as well as in other known domains of AP-2alpha expression previously noted by in situ hybridization studies), and was mostly extinguished by E11.5 when the cardiac neural crest has migrated into the outflow tract of the developing heart. Importantly, the distribution of AP-2alpha-expressing cardiac neural crest appeared to be identical in normal and mutant embryos. From this analysis, we propose that the AP-2alpha gene functions within the neural crest lineage, that AP-2alpha is not required for neural crest cell migration, and that normal AP-2alpha gene function is required prior to E11.5. AP-2alpha may be involved in an interaction between neural crest and surrounding tissues in the subpharyngeal region, thereby promoting normal outflow tract morphogenesis.

The homeobox gene Hhex has recently been shown to be essential for normal liver, thyroid and forebrain development. Hhex–/– mice die by mid-gestation (E14.5) and the cause of their early demise remains unclear. Because Hhex is expressed in the developing blood islands at E7.0 in the endothelium of the developing vasculature and heart at E9.0-9.5, and in the ventral foregut endoderm at E8.5-9.0, it has been postulated to play a critical role in heart and vascular development. We show here, for the first time, that a null mutation of Hhex results in striking abnormalities of cardiac and vascular development which include: (1) defective vasculogenesis, (2)hypoplasia of the right ventricle, (3) overabundant endocardial cushions accompanied by ventricular septal defects, outflow tract abnormalities and atrio-ventricular (AV) valve dysplasia and (4) aberrant development of the compact myocardium. The dramatic enlargement of the endocardial cushions in the absence of Hhex is due to decreased apoptosis and dysregulated epithelial-mesenchymal transformation (EMT). Interestingly, vascular endothelial growth factor A (Vegfa) levels in the hearts of Hhex–/– mice were elevated as much as three-fold between E9.5 and E11.5, and treatment of cultured Hhex–/– AV explants with truncated soluble Vegfa receptor 1, sFlt-1, an inhibitor of Vegf signaling, completely abolished the excessive epithelial-mesenchymal transformation seen in the absence of Hhex. Therefore, Hhex expression in the ventral foregut endoderm and/or the endothelium is necessary for normal cardiovascular development in vivo, and one function of Hhex is to repress Vegfa levels during development.

Mouse embryos lacking the retinoic acid (RA) receptors RARalpha1 and RARbeta suffer from a failure to properly septate (divide) the early outflow tract of the heart into distinct aortic and pulmonary channels, a phenotype termed persistent truncus arteriosus. This phenotype is associated with a failure in the development of the cardiac neural crest cell lineage, which normally forms the aorticopulmonary septum. In this study, we examined the fate of the neural crest lineage in RARalpha1/RARbeta mutant embryos by crossing with the Wnt1-cre and conditional R26R alleles, which together constitute a genetic lineage marker for the neural crest. We find that the number, migration, and terminal fate of the cardiac neural crest is normal in mutant embryos; however, the specific function of these cells in forming the aorticopulmonary septum is impaired. We furthermore show that the neural crest cells themselves do not utilize retinoid receptors and do not respond to RA during this process, but rather that the phenotype is cell non-autonomous for the neural crest cell lineage. This suggests that an alternative tissue in the vicinity of the outflow tract of the heart responds directly to RA, and thereby induces or permits the neural crest cell lineage to initiate aorticopulmonary septation.

No abstract