A homeotic transformation is generated in the rostral branchial region of the head by disruption of Hoxa-2, which acts as a selector gene

Differential expression of Hoxa-2 protein along the dorsal-ventral axis of the developing and adult mouse spinal cord

We have used synthetic oligopeptides derived from the coding sequence of the murine Hoxa-2 protein to produce polyclonal antibodies that specifically recognize the Hoxa-2 recombinant protein. Immunohistochemical studies reveal a distinct pattern of spatial and temporal expression of Hoxa-2 protein within the mouse spinal cord which is concomitant with the cytoarchitectural changes occurring in the developing cord. Hoxa-2 protein is predominantly detected in the nuclei of cells in the ventral mantle region of 10-day-old mouse embryos. Islet-1, a marker for motor neurons was also shown to be co-localized with Hoxa-2 in nuclei of cells in this region. As development progresses from 10-days to 14-days of gestation, Hoxa-2 protein expression gradually extends to the dorsal regions of the mantle layer. The Hoxa-2 protein expression pattern changes at 16-days of embryonic development with strong expression visible throughout the dorsal mantle layer. In 18-day-old and adult mouse spinal cords, Hoxa-2 protein was expressed predominantly by cells of the dorsal horn and only by a few cells of the ventral horn. Double labeling studies with an antibody against glial fibrillary acidic protein (GFAP, an astrocyte-specific intermediate filament protein) showed that within the adult spinal cord, astrocytes rarely expressed the Hoxa-2 protein. However, Hoxa-2 and GFAP double-labeled astrocytes were found in the neopallial cultures, although not all astrocytes expressed Hoxa-2. Hoxa-2 expressing oligodendrocyte progenitor cells were also identified after double-labeling with O4 and Hoxa-2 antibodies; although cells in this lineage that have begun to develop a more extensive array of cytoplasmic processes were less likely to be Hoxa-2 positive. The early pattern of Hoxa-2 protein expression across transverse sections of the neural tube is temporally and spatially modified as each major class of neuron is generated. This congruence in the expression of the Hoxa-2 protein and the generation of neurons in the cord suggests that the Hoxa-2 protein may contribute to dorsal-ventral patterning and/or to the specification of neuronal phenotype. Dev Dyn 1999;216:201-217.

Dlx5 regulates regional development of the branchial arches and sensory capsules

We report the generation and analysis of mice homozygous for a targeted deletion of the Dlx5 homeobox gene. Dlx5 mutant mice have multiple defects in craniofacial structures, including their ears, noses, mandibles and calvaria, and die shortly after birth. A subset (28%) exhibit exencephaly. Ectodermal expression of Dlx5 is required for the development of olfactory and otic placode-derived epithelia and surrounding capsules. The nasal capsules are hypoplastic (e.g. lacking turbinates) and, in most cases, the right side is more severely affected than the left. Dorsal otic vesicle derivatives (e. g. semicircular canals and endolymphatic duct) and the surrounding capsule, are more severely affected than ventral (cochlear) structures. Dlx5 is also required in mandibular arch ectomesenchyme, as the proximal mandibular arch skeleton is dysmorphic. Dlx5 may control craniofacial development in part through the regulation of the goosecoid homeobox gene. goosecoid expression is greatly reduced in Dlx5 mutants, and both goosecoid and Dlx5 mutants share a number of similar craniofacial malformations. Dlx5 may perform a general role in skeletal differentiation, as exemplified by hypomineralization within the calvaria. The distinct focal defects within the branchial arches of the Dlx1, Dlx2 and Dlx5 mutants, along with the nested expression of their RNAs, support a model in which these genes have both redundant and unique functions in the regulation of regional patterning of the craniofacial ectomesenchyme.

Pathogenesis of cleft palate in TGF-beta3 knockout mice

We previously reported that mutation of the transforming growth factor-beta3 (TGF-beta3) gene caused cleft palate in homozygous null (−/−) mice. TGF-beta3 is normally expressed in the medial edge epithelial (MEE) cells of the palatal shelf. In the present study, we investigated the mechanisms by which TGF-beta3 deletions caused cleft palate in 129 × CF-1 mice. For organ culture, palatal shelves were dissected from embryonic day 13.5 (E13.5) mouse embryos. Palatal shelves were placed singly or in pairs on Millipore filters and cultured in DMEM/F12 medium. Shelves were placed in homologous (+/+ vs +/+, −/− vs −/−, +/− vs +/−) or heterologous (+/+ vs −/−, +/− vs −/−, +/+ vs +/−) paired combinations and examined by macroscopy and histology. Pairs of −/− and −/− shelves failed to fuse over 72 hours of culture whereas pairs of +/+ (wild-type) and +/+ or +/− (heterozygote) and +/−, as well as +/+ and −/− shelves, fused within the first 48 hour period. Histological examination of the fused +/+ and +/+ shelves showed complete disappearance of the midline epithelial seam whereas −/− and +/+ shelves still had some seam remnants. In order to investigate the ability of TGF-beta family members to rescue the fusion between −/− and −/− palatal shelves in vitro, either recombinant human (rh) TGF-beta1, porcine (p) TGF-beta2, rh TGF-beta3, rh activin, or p inhibin was added to the medium in different concentrations at specific times and for various periods during the culture. In untreated organ culture −/− palate pairs completely failed to fuse, treatment with TGF-beta3 induced complete palatal fusion, TGF-beta1 or TGF-beta2 near normal fusion, but activin and inhibin had no effect. We investigated ultrastructural features of the surface of the MEE cells using SEM to compare TGF-beta3-null embryos (E 12. 5-E 16.5) with +/+ and +/− embryos in vivo and in vitro. Up to E13.5 and after E15.5, structures resembling short rods were observed in both +/+ and −/− embryos. Just before fusion, at E14.5, a lot of filopodia-like structures appeared on the surface of the MEE cells in +/+ embryos, however, none were observed in −/− embryos, either in vivo or in vitro. With TEM these filopodia are coated with material resembling proteoglycan. Interestingly, addition of TGF-beta3 to the culture medium which caused fusion between the −/− palatal shelves also induced the appearance of these filopodia on their MEE surfaces. TGF-beta1 and TGF-beta2 also induced filopodia on the −/− MEE but to a lesser extent than TGF-beta3 and additionally induced lamellipodia on their cell surfaces. These results suggest that TGF-beta3 may regulate palatal fusion by inducing filopodia on the outer cell membrane of the palatal medial edge epithelia prior to shelf contact. Exogenous recombinant TGF-beta3 can rescue fusion in −/− palatal shelves by inducing such filopodia, illustrating that the effects of TGF-beta3 are transduced by cell surface receptors which raises interesting potential therapeutic strategies to prevent and treat embryonic cleft palate.

Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development

Mice homozygous for a targeted deletion of the homeobox gene Goosecoid (Gsc) have multiple craniofacial defects. To understand the mechanisms responsible for these defects, the behavior of Gsc-null cells was examined in morula aggregation chimeras. In these chimeras, Gsc-null cells were marked with beta-galactosidase (beta-gal) activity using the ROSA26 lacZ allele. In addition, mice with a lacZ gene that had been introduced into the Gsc locus were used as a guide to visualize the location of Gsc-expressing cells. In Gsc-null<->wild-type chimeras, tissues that would normally not express Gsc were composed of both Gsc-null and wild-type cells that were well mixed, reflecting the overall genotypic composition of the chimeras. However, craniofacial tissues that would normally express Gsc were essentially devoid of Gsc-null cells. Furthermore, the nasal capsules and mandibles of the chimeras had defects similar to Gsc-null mice that varied in severity depending upon the proportion of Gsc-null cells. These results combined with the analysis of Gsc-null mice suggest that Gsc functions cell autonomously in mesenchyme-derived tissues of the head. A developmental analysis of the tympanic ring bone, a bone that is always absent in Gsc-null mice because of defects at the cell condensation stage, showed that Gsc-null cells had the capacity to form the tympanic ring condensation in the presence of wild-type cells. However, analysis of the tympanic ring bones of 18.5 d.p.c. chimeras suggests that Gsc-null cells were not maintained. The participation of Gsc-null cells in the tympanic ring condensation of chimeras may be an epigenetic phenomenon that results in a local environment in which more precursor cells are present. Thus, the skeletal defects observed in Gsc-null mice may reflect a regional reduction of precursor cells during embryonic development.

Pathogenesis of retinoic acid-induced ear malformations in primate model

13-cis retinoic acid (RA) is a causative agent for human/monkey retinoic acid embryopathy (RAE), in which the most common type of malformation is microtia or anotia. In the present study, malformed ears of monkey fetuses exposed to RA during early embryogenesis were analyzed and revealed a subtype of defects., i.e., apparent duplication of the external/middle ear. A part of the posterior auricle appeared to be ectopically formed in the anterior auricular region or in the region posterior to the auricle. Additionally, there was duplication of the zygomatic arch, malleus, and incus. In order to characterize possible pathogenetic events underlying these malformations, embryos at selected stages were collected after dosing dams with RA at 5 mg/kg/day during gestational days 12-27. Cellular retinoic acid binding protein I whole-mount immunostaining showed that RA induced specific alterations in the migration of cranial neural crest cells (NCC). NCC en route to the second pharyngeal arch were bifurcated, and some of these NCC migrated abnormally into the first and/or third arches, which may underlie external ear duplication. Scanning electron microscopy and neurofilament immunostaining provided evidence that there was partial duplication of trigeminal nerve/ganglion following RA insult. The duplication of NCC neuronal derivatives in the first pharyngeal arch is consistent with duplication of NCC mesenchymal components (zygomatic arch, malleus, and incus). Therefore, RA-induced alterations in cranial NCC migration patterns are likely to be a pathogenetic event underlying ear malformations (including duplication) of RAE in monkeys.

Development of dermal denticles in skates (Chondrichthyes, Batoidea): patterning and cellular differentiation

Patterning, cellular differentiation, and developmental sequences of dermal denticles (denticles) are described for the skate Leucoraja erinacea. Development of denticles proceeds caudo-rostrally in the tail and trunk. Once three rows of denticles form in the tail and trunk, denticles begin to appear in the region of the pelvic girdle, medio-caudal to the eyes and on the pectoral fins. Although timing of cellular differentiation of denticles differs among different locations of the body, cellular development of a denticle is identical in all locations. Thickening of the epidermis as a denticle lamina marks initiation of development. A single lamina for each denticle forms, and a small group of mesenchymal cells aggregates underneath it. The lamina then invaginates caudo-rostrally to form the inner- and outer-denticle epithelia (IDE and ODE, respectively). Before nuclei of IDE cells are polarized, enameloid matrix appears between the basement membrane of the IDE and the apical surface of the pre-odontoblasts. Pre-dentin is then laid down along with collagenous materials. Von Kossa stain visualizes initial mineralization of dentin, but not enameloid. During the growth of a denticle, dense fibrous connective tissue of the dermis forms the deep dermal tissue over the dorsal musculature. Attachment fibers and tendons anchor denticles and dorsal musculature, respectively, on deep dermal tissue. Basal tissue of the denticles develops as the denticle crown grows. If the basal tissue is bone of attachment, then the cells along the basal tissue would be osteoblasts. However, these cells could not be distinguished from odontoblasts using immunolocalization of type I pro-collagen (Col I), alkaline phosphatase (APase), and neural cell adhesion molecule (N-CAM). Well-developed dentin, (not pre-dentin), the enameloid matrix (probably when it begins to mineralize), and deep dermal tissue are Verhoeff stain-positive, suggesting that these tissues contain elastin and/or elastin-like molecules. Our study demonstrates that the cellular development of denticles resembles tooth development in elasmobranchs, but that dermal denticles differ from teeth in forming from a single denticle lamina. Whether the basal tissue of denticles is bone of attachment remains undetermined. Confirmation and function of Verhoeff-positive proteins in enameloid, dentin, and deep dermal tissue remain to be determined. We discuss these issues along with an analysis of recent findings of enamel and enameloid matrices.

Development of the vertebrate ear: insights from knockouts and mutants

The three divisions of the ear (outer, middle and inner) each have an important role in hearing, while the inner ear is also crucial for the sense of balance. How these three major components arise and coalesce to form the peripheral elements of the senses of hearing and balance is now being studied using molecular-genetic approaches. This article summarizes data from studies of knockout and mutant animals in which one or more divisions of the ear are abnormal. The data confirm that development of all three divisions of the ear depends on the genes involved in hindbrain segmentation and segment identity. Genes that are regionally expressed in the inner ear can, when absent or mutated, yield selective ablation of specific inner-ear structures or cell types.

Retinoids and mammalian development

All vertebrate embryos require retinoic acid (RA) for fulfilment of the developmental program encoded in the genome. In mammals, maternal homeostatic mechanisms minimize variation of retinoid levels reaching the embryo. Retinol is transported as a complex with retinol-binding protein (RBP): transplacental transfer of retinol and its uptake by the embryonic tissues involves binding to an RBP receptor at the cell surface. Embryonic tissues in which this receptor is present also contain the retinol-binding protein CRBP I and the enzymes involved in RA synthesis; the same tissues are particularly vulnerable to vitamin A deficiency. In the nucleus, the RA signal is transduced by binding to a heterodimeric pair of retinoid receptors (RAR/RXR). In general, the receptors show functional plasticity, disruption of one RAR or RXR gene having minor or no effects on embryogenesis. However, genetic studies indicate that RXR alpha is essential for normal development of the heart and eye. Excess RA causes abnormalities of many systems; altered susceptibility to RA excess in mice lacking RAR gamma or RXR alpha suggests that the teratogenic signal is transduced through different receptors compared with physiological RA function in the same tissue.

Neural crest can form cartilages normally derived from mesoderm during development of the avian head skeleton

The lateral wall of the avian braincase, which is indicative of the primitive amniote condition, is formed from mesoderm. In contrast, mammals have replaced this portion of their head skeleton with a nonhomologous bone of neural crest origin. Features that characterize the local developmental environment may have enabled a neural crest-derived skeletal element to be integrated into a mesodermal region of the braincase during the course of evolution. The lateral wall of the braincase lies along a boundary in the head that separates neural crest from mesoderm, and also, neural crest cells migrate through this region on their way to the first visceral arch. Differences in the availability of one skeletogenic population versus the other may determine the final composition of the lateral wall of the braincase. Using the quail-chick chimeric system, this investigation tests if populations of neural crest, when augmented and expanded within populations of mesoderm, will give rise to the lateral wall of the braincase. Results demonstrate that neural crest can produce cartilages that are morphologically indistinguishable from elements normally generated by mesoderm. These findings (1) indicate that neural crest can respond to the same cues that both promote skeletogenesis and enable proper patterning in mesoderm, (2) challenge hypotheses on the nature of the boundary between neural crest and mesoderm in the head, and (3) suggest that changes in the allocation of migrating cells could have enabled a neural crest-derived skeletal element to replace a mesodermal portion of the braincase during evolution.

Hoxa2 and Hoxb2 control dorsoventral patterns of neuronal development in the rostral hindbrain

Little is known about how the generation of specific neuronal types at stereotypic positions within the hindbrain is linked to Hox gene-mediated patterning. Here, we show that during neurogenesis, Hox paralog group 2 genes control both anteroposterior (A-P) and dorsoventral (D-V) patterning. Hoxa2 and Hoxb2 differentially regulate, in a rhombomere-specific manner, the expression of several genes in broad D-V-restricted domains or narrower longitudinal columns of neuronal progenitors, immature neurons, and differentiating neuronal subtypes. Moreover, Hoxa2 and Hoxb2 can functionally synergize in controlling the development of ventral neuronal subtypes in rhombomere 3 (r3). Thus, in addition to their roles in A-P patterning, Hoxa2 and Hoxb2 have distinct and restricted functions along the D-V axis during neurogenesis, providing insights into how neuronal fates are assigned at stereotypic positions within the hindbrain.

Regulation of Hoxa2 in cranial neural crest cells involves members of the AP-2 family

Hoxa2 is expressed in cranial neural crest cells that migrate into the second branchial arch and is essential for proper patterning of neural-crest-derived structures in this region. We have used transgenic analysis to begin to address the regulatory mechanisms which underlie neural-crest-specific expression of Hoxa2. By performing a deletion analysis on an enhancer from the Hoxa2 gene that is capable of mediating expression in neural crest cells in a manner similar to the endogenous gene, we demonstrated that multiple cis-acting elements are required for neural-crest-specific activity. One of these elements consists of a sequence that binds to the three transcription factor AP-2 family members. Mutation or deletion of this site in the Hoxa2 enhancer abrogates reporter expression in cranial neural crest cells but not in the hindbrain. In both cell culture co-transfection assays and transgenic embryos AP-2 family members are able to trans-activate reporter expression, showing that this enhancer functions as an AP-2-responsive element in vivo. Reporter expression is not abolished in an AP-2(alpha) null mutant embryos, suggesting redundancy with other AP-2 family members for activation of the Hoxa2 enhancer. Other cis-elements identified in this study critical for neural-crest-specific expression include an element that influences levels of expression and a conserved sequence, which when multimerized directs expression in a broad subset of neural crest cells. These elements work together to co-ordinate and restrict neural crest expression to the second branchial arch and more posterior regions. Our findings have identified the cis-components that allow Hoxa2 to be regulated independently in rhombomeres and cranial neural crest cells.

Development of neural crest in Xenopus

The neural crest is a unique cell population among embryonic cell types, displaying properties of both ectodermal and mesodermal lineages. Most of the recent studies examining the neural crest have been performed in avian embryos. Only in the first half of this century were amphibians extensively used. We first summarize this important older source of information, reviewing studies made since the turn of the century. Due to the increasingly detailed in cellular and molecular knowledge of the early development of Xenopus laevis, the remainder of the review focuses on this species. We describe the route of migration and fate of the neural crest and propose a new model of neural crest induction in which prospective cells are induced independently of the neural plate by a double gradient of a morphogen that patterns the entire ectoderm. This model is also discussed in a more general context in connection with the dorsoventral patterning of the neural tube. Finally, we discuss some ideas concerning neural crest evolution and propose a novel hypothesis about its phylogenetic origin.

Fgf-8 determines rostral-caudal polarity in the first branchial arch

In mammals, rostral ectomesenchyme cells of the mandibular arch give rise to odontogenic cells, while more caudal cells form the distal skeletal elements of the lower jaw. Signals from the epithelium are required for the development of odontogenic and skeletogenic mesenchyme cells. We show that rostral-caudal polarity is first established in mandibular branchial arch ectomesenchymal cells by a signal, Fgf-8, from the rostral epithelium. All neural crest-derived ectomesenchymal cells are equicompetent to respond to Fgf-8. The restriction into rostral (Lhx-7-expressing) and caudal (Gsc-expressing) domains is achieved by cells responding differently according to their proximity to the source of the signal. Once established, spatial expression domains and cell fates are fixed and maintained by Fgf-8 in conjunction with another epithelial signal, endothelin-1, and by positional changes in ectomesenchymal cell competence to respond to the signal.

Ectopic expression of hoxb2 after retinoic acid treatment or mRNA injection: disruption of hindbrain and craniofacial morphogenesis in zebrafish embryos

To investigate pattern formation in the vertebrate hindbrain, we isolated a full length hoxb2 cDNA clone from zebrafish. In a gene phylogeny, zebrafish hoxb2 clusters with human HOXB2, and it maps on linkage group 3 along with several other loci whose orthologues are syntenic with human HOXB2. In the hindbrain, hoxb2 is expressed at high levels in rhombomere 3 (r3), lower levels in r4, still lower in r5, and at undetectable levels in r6. In r7, r8, and the rostral spinal cord, hoxb2 is expressed at a lower level than in r5. Lateral cells appearing to emanate from r4 express both hoxb2 and dlx2, suggesting that they are neural crest. Overexpression of hoxb2 by mRNA injections into early cleavage stage embryos resulted in abnormal morphogenesis of the midbrain and rostral hindbrain, abnormal patterning in r4, fusion of cartilage elements arising from pharyngeal arches 1 and 2, and ectopic expression of krx20 and valentino (but not pax2, rtk1, or hoxb1) in the rostral hindbrain, midbrain, and, surprisingly, the eye. Treatments with retinoic acid produced a phenotype similar to that of ectopic hoxb2 expression, including ectopic krx20 (but not valentino) expression in the eye, and fusion of cartilages from pharyngeal arches 1 and 2. The results suggest that hoxb2 plays an important role in the patterning of hindbrain and pharyngeal arches in the zebrafish.

The shaping of pharyngeal cartilages during early development of the zebrafish

In zebrafish the cartilages of the pharynx develop during late embryogenesis and grow extensively in the larva before eventually being replaced by bone. Here we examine chondrocyte arrangements, shapes, numbers, and divisions in the young hyoid cartilages. We observe two distinct developmental phases, morphogenesis and growth. The first phase generates stereotypically oriented chondrocyte stacks that might form by intercalations among cells within the precartilage condensations. In mutants that have deformed cartilages the orientation of the stacks is changed, and we propose that their correct formation underlies the correct initial shaping of the organ. The following period of rapid, nearly isometric cartilage growth occurs by divisions of chondrocytes that are largely located near the joints, and appears to be under quite separate regulation.

Programmed cell death in the development of the mouse external auditory canal

Programmed cell death (PCD) is an essential event for development. The purpose of this work was to ascertain how PCD, in vivo designated apoptosis, is involved in the development of the external auditory canal. We performed a time sequence study of the distribution of apoptosis during the development of external auditory canal (EAC) of the mouse. ICR mice ranging in age from embryonic day 11.5 (E11.5) to 12 days after birth (DAB) were used in the present study. A part of each head including both ears was removed and was processed according to its purpose. Light and electron microscopy for morphological studies and TUNEL method (Gavrieli et al. [1992] J Cell Biol., 119:493-501) for histochemical studies were used. On E11.5, distinct TUNEL-positive staining occurred in the branchial arch. Between E15.5 and 1DAB, TUNEL-positive cells were observed throughout the EAC and the number of these cells decreased with age. On E15.5 and E16.5, numerous TUNEL-positive cells were observed in a cavity remained in the epithelial plate. Transmission electron microscopy revealed that these cells had the features of apoptosis. From 3-12 DAB, no apoptosis was observed in the EAC except for the terminal differentiation of the skin of the EAC. Apoptosis was not observed during recanalization of the EAC, but occurred during the formation of the epithelial plate. The investigation established that PCD is involved in the formation of the epithelial plate, whereas only cornification of the epithelium of the EAC is associated with recanalization.

Expression of rat homeobox gene, rHOX, in developing and adult tissues in mice and regulation of its mRNA expression in osteoblasts by bone morphogenetic protein 2 and parathyroid hormone-related protein

The rat homeobox gene, rHox, was cloned from a rat osteosarcoma cDNA library. Southwestern and gel mobility shift analyses showed that rHox binds to the promoter regions of collagen (alpha1)I and osteocalcin genes while transient transfection with rHox resulted in repression of their respective promoter activities. In situ hybridization studies showed that rHox mRNA was widely expressed in osteoblasts, chondrocytes, skeletal muscle, skin epidermis, and bronchial and intestinal epithelial cells, as well as cardiac muscle in embryonic and newborn mice. However in 3-month-old mice, rHox mRNA expression was restricted to osteoblasts, megakaryocytes, and myocardium. Bone morphogenetic protein 2, a growth factor that commits mesenchymal progenitor cells to differentiate into osteoblasts, down-regulated rHox mRNA expression by 40-50% in UMR 201, a rat preosteoblast cell line, in a time- and dose-dependent manner. In contrast, PTH-related protein (PTHrP), recently shown to be a negative regulator of chondrocyte differentiation, significantly enhanced rHox mRNA expression in UMR 106-06 osteoblastic cells by 3-fold at 24 h while at the same time down-regulating expression of pro-alpha1(I) collagen mRNA by 60%. Expression of rHox mRNA in calvarial osteoblasts derived from PTHrP -/- mice was approximately 15% of that observed in similar cells obtained from normal mice. In conclusion, current evidence suggests that rHox acts as a negative regulator of osteoblast differentiation. Furthermore, down-regulation of rHox mRNA by bone morphogenetic protein 2 and its up-regulation by PTHrP support a role of the homeodomain protein, rHox, in osteoblast differentiation.

How to build a vertebrate hindbrain. Lessons from genetics

During vertebrate embryogenesis, the hindbrain is the site of a segmentation process which leads to the formation, along the anterior-posterior axis, of 7-8 metameres called rhombomeres. This phenomenon plays an essential role in early hindbrain regionalisation and in the specification of the pattern of developing structures in this region of the brain. Data accumulated during the last 10 years have also shown that rhombomeres are units of gene expression and of cell lineage. Hence, a number of regulatory genes are expressed according to segment-specific patterns in the hindbrain and have been implicated in the pattern formation process. In this review, we focus on the analysis of the function and regulation of these genes along the different steps of hindbrain segmentation, from segment delimitation to acquisition of positional identity. On this basis, we propose a model for the control of early hindbrain development.

Analysis of Hoxa7/Hoxb7 mutants suggests periodicity in the generation of the different sets of vertebrae

Mice were generated with targeted disruptions in Hoxa7 and Hoxb7, respectively. Mice carrying the Hoxa7 mutation are healthy. No abnormalities in the formation of the skeleton or other tissues were found in these mutants. Twelve percent of Hoxb7-/- mutants show first and second rib defects similar to those observed in mice homozygous for a Hoxb9 mutation (Chen, F., Capecchi, M.R., 1997. Dev. Biol. 181, 186-196). Hoxb7-/- mice are also fertile and were used to generate double mutants with Hoxa7 to reveal potential interactions between these two paralogous genes. Mice homozygous for both mutations have first and second rib defects with higher penetrance and increased expressivity, indicating a functional role for Hoxa7 in the patterning of the upper thoracic region. Although Hoxb6, Hoxa7, Hoxb7, and Hoxb9 have distinctive anterior expression limits in axial mesoderm, the disruptions of these genes all yield first and second rib defects. A hypothesis is suggested to explain the observation that axial defects in these and other mouse Hox mutants appear to concentrate along the axial column at zones of transition between vertebral types.

MuSC, a novel member of the immunoglobulin superfamily, is expressed in neurons of a subset of cranial sensory ganglia in the mouse embryo

In contrast to the spinal sensory ganglia which reiterate a basic organizational and functional unit, each cranial ganglion mediates a distinct sensory modality and exhibits a characteristic pattern of peripheral and central neuronal connectivity. Molecules responsible for establishment and maintenance of the cranial ganglion-specific networks are not known. Our hamster monoclonal antibody 802C11 strongly stained neurons and their processes of the VIIIth cranial ganglion (hearing and equilibrium), but not of the Vth cranial (somatosensory) or spinal ganglia in the mouse embryo. The cellular staining pattern of positive neurons suggested that the antigen was associated with the cell membrane, and biochemical analyses of the antigen from adult mouse brain showed the antigen to be a glycosylated intrinsic membrane protein of approximately 100 kDa. The antigen was purified, and based on the partial amino acid sequences, its entire cDNA was cloned. A bacterially expressed polypeptide encoded by the cDNA was recognized by the antibody. The deduced amino acid sequence revealed that the antigen belongs to the immunoglobulin superfamily with a significant homology (73.5% identity) to chicken SC1 protein. Chicken SC1 has been shown to be a cell-cell adhesion molecule in vitro with a proposed role in neurite extension of spinal motor neurons. These results suggest that our murine SC1-related protein (MuSC) is involved in the pathfinding and/or fasciculation of specific cranial sensory nerve fibres.

Determination of the identity of the derivatives of the cephalic neural crest: incompatibility between Hox gene expression and lower jaw development

In addition to pigment cells, and neural and endocrine derivatives, the neural crest is characterized by its ability to yield mesenchymal cells. In amniotes, this property is restricted to the cephalic region from the mid-diencephalon to the end of rhombomere 8 (level of somites 4/5). The cephalic neural crest is divided into two domains: an anterior region corresponding to the diencephalon, mesencephalon and metencephalon (r1, r2) in which expression of Hox genes is never observed, and a posterior domain in which neural crest cells exhibit (with a few exceptions) the same Hox code as the rhombomeres from which they originate. By altering the normal distribution of neural crest cells in the branchial arches through appropriate embryonic manipulations, we have investigated the relationships between Hox gene expression and the level of plasticity that neural crest cells display when they are led to migrate to an ectopic environment. We made the following observations. (i) Hox gene expression is not altered in neural crest cells by their transposition to ectopic sites. (ii) Expression of Hox genes by the BA ectoderm does not depend upon an induction by the neural crest. This second finding further supports the concept of segmentation of the cephalic ectoderm into ectomeres (Couly and Le Douarin, 1990). According to this concept, metameres can be defined in large bands of ectoderm including not only the CNS and the neural crest but also the corresponding superficial ectoderm fated to cover craniofacial primordia. (iii) The construction of a lower jaw requires the environment provided by the ectomesodermal components of BA1 or BA2 associated with the Hox gene non-expressing neural crest cells. Hox gene-expressing neural crest cells are unable to yield the lower jaw apparatus including the entoglossum and basihyal even in the BA1 environment. In contrast, the posterior part of the hyoid bone can be constructed by any region of the neural crest cells whether or not they are under the regulatory control of Hox genes. Such is also the case for the neural and connective tissues (including those comprising the cardiovascular system) of neural crest origin, upon which no segmental restriction is imposed. The latter finding confirms the plasticity observed 24 years ago (Le Douarin and Teillet, 1974) for the precursors of the PNS.

Induction and patterning of the neural crest, a stem cell-like precursor population

The neural crest is a multipotent precursor population which ultimately generates much of the peripheral nervous system, epidermal pigment cells, and a variety of mesectodermal derivatives. Individual multipotent neural crest cells are capable of some self-renewing divisions, and based upon this criteria can be considered stem cells. Considerable progress has been made in recent years toward understanding how this important population of progenitor cells is initially established in the early embryo, and how cell-intrinsic and non-cell-intrinsic factors mediate their subsequent lineage segregation and differentiation.

Homeobox genes in cardiovascular development

As summarized earlier, a surprisingly large number of different homeobox genes are expressed in the developing heart. Some are clearly important, as demonstrated by mouse gene ablation studies. For example, knockout of Nkx2-5 or Hoxa-3 function is embryonic lethal due to defects in cardiovascular development. However, gene ablation studies indicate that other homeobox genes that show cardiovascular expression are either not required for heart development or their function is effectively complemented by a redundant gene activity. Given the number of closely related homeobox genes that are expressed in the heart (and the rate at which new genes are being discovered), this is very likely to be the case for at least some homeobox gene activities. At present little is known of the precise mechanism of action of homeobox genes in embryonic development. This statement applies to homeobox genes in general, not just to genes involved in cardiovascular development. There is a popular view that homeobox genes are master regulators that control expression of a large number of downstream genes. In at least some cases, e.g., the eyeless gene of Drosophila (Holder et al., 1995), homeobox genes appear to be capable of activating and maintaining a very complex developmental program. Significantly, the eyeless gene is able to initiate eye development at numerous ectopic locations. Increasing evidence, however, suggests that genes of this type may be rather rare. Certainly there is no evidence to date that any of the homeobox genes expressed in the heart are able to initiate the complete heart development pathway. This is probably best understood in the case of the tinman gene in Drosophila, which, although absolutely required for heart development, is not capable of initiating the cardiac development pathway in ectopic locations (Bodmer, 1993). This conclusion is supported by studies of the vertebrate tinman-related gene Nkx2-5. Gene ablation studies show that Nkx2-5 is essential for correct cardiac development (Lyons et al., 1995) but is not able to initiate the regulatory pathway leading to cardiac development when expressed ectopically (Cleaver et al., 1996; Chen and Fishman, 1996). If most homeodomain proteins are not direct regulators of a differentiation pathway, what is their role during organogenesis? The cardiovascular homeobox gene about which most is known at the mechanistic level is gax (Smith et al., 1997). A number of experiments indicate that the Gax protein is involved in the regulation of cell proliferation and that it interacts with components of the cell cycle regulation machinery. Indeed, over recent years, the idea that at least some homeobox genes play their role in organogenesis through regulation of proliferation has been developed in some detail by Duboule (1995). Further evidence that this mechanism of homeobox activity is important, especially during organogenesis, comes from studies of the Hox11 homeobox gene, which is absolutely required for development of the spleen in mouse (Roberts et al., 1994). Studies indicate that Hox11 is able to interact with at least two different protein phosphatases, PP2A and PP1, which in turn, are involved in cell cycle regulation (Kawabe et al., 1997). It is quite clear that research in future years will need to focus on the precise mode of action of the different homeodomain proteins if we are to understand their role in the development of the cardiovascular system.

Hoxa-2 restricts the chondrogenic domain and inhibits bone formation during development of the branchial area

In Hoxa-2(−/−)embryos, the normal skeletal elements of the second branchial arch are replaced by a duplicated set of first arch elements. We show here that Hoxa-2 directs proper skeletal formation in the second arch by preventing chondrogenesis and intramembranous ossification. In normal embryos, Hoxa-2 is expressed throughout the second arch mesenchyme, but is excluded from the chondrogenic condensations. In the absence of Hoxa-2, chondrogenesis is activated ectopically within the rostral Hoxa-2 expression domain to form the mutant set of cartilages. In Hoxa-2(−/−)embryos the Sox9 expression domain is shifted into the normal Hoxa-2 domain. Misexpression of Sox9 in this area produces a phenotype resembling that of the Hoxa-2 mutants. These results indicate that Hoxa-2 acts at early stages of the chondrogenic pathway, upstream of Sox9 induction. We also show that Hoxa-2 inhibits dermal bone formation when misexpressed in its precursors. Furthermore, molecular analyses indicate that Cbfa1 is upregulated in the second branchial arches of Hoxa-2 mutant embryos suggesting that prevention of Cbfa1 induction might mediate Hoxa-2 inhibition of dermal bone formation during normal second arch development. The implications of these results on the patterning of the branchial area are discussed.

Signalling interactions during facial development

The development of the vertebrate face is a dynamic multi-step process which starts with the formation of neural crest cells in the developing brain and their subsequent migration to form, together with mesodermal cells, the facial primordia. Signalling interactions co-ordinate the outgrowth of the facial primordia from buds of undifferentiated mesenchyme into the intricate series of bones and cartilage structures that, together with muscle and other tissues, form the adult face. Some of the molecules that are thought to be involved have been identified through the use of mouse mutants, data from human craniofacial syndromes and by expression studies of signalling molecules during facial development. However, the way that these molecules control the epithelial-mesenchymal interactions which mediate facial outgrowth and morphogenesis is unclear. The role of neural crest cells in these processes has also not yet been well defined. In this review we discuss the complex interaction of all these processes during face development and describe the candidate signalling molecules and their possible target genes.

Role of Endothelin-1/Endothelin-A receptor-mediated signaling pathway in the aortic arch patterning in mice

The intercellular signaling mediated by endothelins and their G protein-coupled receptors has recently been shown to be essential for the normal embryonic development of subsets of neural crest cell derivatives. Endothelin-1 (ET-1) is proteolytically generated from its inactive precursor by endothelin-converting enzyme-1 (ECE-1) and acts on the endothelin-A (ETA) receptor. Genetic disruption of this ET-1/ECE-1/ETA pathway results in defects in branchial arch- derived craniofacial tissues, as well as defects in cardiac outflow and great vessel structures, which are derived from cephalic (cardiac) neural crest. In this study, in situ hybridization of ETA-/- and ECE-1(-)/- embryos with a cardiac neural crest marker, cellular retinoic acid-binding protein-1, shows that the migration of neural crest cells from the neural tube to cardiac outflow tract is not affected in these embryos. Immunostaining of an endothelial marker, platelet endothelial cell adhesion molecule CD-31, shows that the initial formation of the branchial arch arteries is not disturbed in ETA-/- or ECE-1(-)/- embryos. To visualize the subsequent patterning of arch vessels in detail, we generated ETA-/- or ECE-1(-)/- embryos that expressed an SM22alpha-lacZ marker transgene in arterial smooth muscle cells. Wholemount X-gal staining of these mutant embryos reveals that the abnormal regression and persistence of specific arch arteries results in disturbance of asymmetrical remodeling of the arch arteries. These defects include abnormal regression of arch arteries 4 and 6, enlargement of arch artery 3, and abnormal persistence of the bilateral ductus caroticus and right dorsal aorta. These abnormalities eventually lead to various types of great vessel malformations highly similar to those seen in neural crest-ablated chick embryos and human congenital cardiac defects. This study demonstrates that ET-1/ETA-mediated signaling plays an essential role in a complex process of aortic arch patterning by affecting the postmigratory cardiac neural crest cell development.

Stability and plasticity of neural crest patterning and branchial arch Hox code after extensive cephalic crest rotation

The extent to which the spatial organisation of craniofacial development is due to intrinsic properties of the neural crest is at present unclear. There is some experimental evidence supporting the concept of a prepattern established within crest while contiguous with the neural plate. In experiments in which the neural tube and premigratory crest are relocated within the branchial region, crest cells retain patterns of gene expression appropriate for their position of origin after migration into the branchial arches, resulting in skeletal abnormalities. But in apparent conflict with these findings, when crest is rerouted by late deletion of adjacent crest, infilling crest alters its pattern of gene expression to match its new location, and a normal facial skeleton results. In order to reconcile these findings thus identify processes of relevance to the course of normal development, we have performed a series of neural tube and crest rotations producing a more extensive reorganisation of cephalic crest than has been previously described. Lineage analysis using DiI labelling of crest derived from the rotated hindbrain reveals that crest does not migrate into the branchial arch it would have colonised in normal development, rather it simply populates the nearest available branchial arches. We also find that crest adjacent to the grafted region contributes to a greater number of branchial arches than it would in normal development, resulting in branchial arches containing mixed cell populations not occurring in normal development. We find that after exchange of first and third arch crest by rotation of r1-7, crest alters its expression of hoxa-2 and hoxa-3 to match its new location within the embryo resulting in the reestablishment of the normal branchial arch Hox code. A facial skeleton in which all the normal components are present, with some additional ectopic first arch structures, is formed in this situation. In contrast, when second and third arch crest are exchanged by rotation of r3 to 7, ectopic Hox gene expression is stable, resulting in the persistence of an abnormal branchial arch Hox code and extensive defects in the hyoid skeleton. We suggest that the intrinsic properties of crest have an effect on the spatial organisation of structures derived from the branchial arches, but that exposure to increasingly novel environments within the branchial region or "community effects" within mixed populations of cells can result in alterations to crest Hox code and morphogenetic fate. In both classes of operation we find that there is a tight link between the resulting branchial arch Hox code and a particular skeletal morphology.

Cloning, characterization, and mapping of the mouse homeobox gene Hmx1

Homeobox-containing genes play an important role in development, including positional specification of the body plan and organogenesis. We previously isolated the human HMX1 (H6) gene, a novel homeobox-containing gene of the HMX family, from a human embryonic craniofacial cDNA library. The closely related mouse genes Hmx3 (Nkx5.1) and Hmx2 (Nkx5.2) are in the same class as the HMX1 gene and are expressed in the craniofacial region of the developing embryo. To provide a resource for further characterization of the human HMX1 gene, we isolated the mouse Hmx1 genomic clone. We show here the mouse Hmx1 genomic sequence, its gene mapping, and its expression pattern in the developing mouse embryo. Evidence is presented showing that the three known Hmx genes in the mouse likely play complementary roles in the development of the second arch, retina, sympathetic nerve ganglia, and cranial neural ganglia. Hmx1 may play an important role in the development of craniofacial structures and may interact with Hoxa-2 and Dlx-2 in the second branchial arch.

Hoxa1 and Hoxb1 synergize in patterning the hindbrain, cranial nerves and second pharyngeal arch

The analysis of Hoxa1 and Hoxb1 null mutants suggested that these genes are involved in distinct aspects of hindbrain segmentation and specification. Here we investigate the possible functional synergy of the two genes. The generation of Hoxa1(3′RARE)/Hoxb1(3′RARE) compound mutants resulted in mild facial motor nerve defects reminiscent of those present in the Hoxb1 null mutants. Strong genetic interactions between Hoxa1 and Hoxb1 were uncovered by introducing the Hoxb1(3′RARE) and Hoxb1 null mutations into the Hoxa1 null genetic background. Hoxa1(null)/Hoxb1(3′RARE) and Hoxa1(null)/Hoxb1(null)double homozygous embryos showed additional patterning defects in the r4-r6 region but maintained a molecularly distinct r4-like territory. Neurofilament staining and retrograde labelling of motor neurons indicated that Hoxa1 and Hoxb1 synergise in patterning the VIIth through XIth cranial nerves. The second arch expression of neural crest cell markers was abolished or dramatically reduced, suggesting a defect in this cell population. Strikingly, the second arch of the double mutant embryos involuted by 10.5 dpc and this resulted in loss of all second arch-derived elements and complete disruption of external and middle ear development. Additional defects, most notably the lack of tympanic ring, were found in first arch-derived elements, suggesting that interactions between first and second arch take place during development. Taken together, our results unveil an extensive functional synergy between Hoxa1 and Hoxb1 that was not anticipated from the phenotypes of the simple null mutants.

Hox group 3 paralogs regulate the development and migration of the thymus, thyroid, and parathyroid glands

The thymus, thyroid, and parathyroid glands in vertebrates develop from the pharyngeal region, with contributions both from pharyngeal endoderm and from neural crest cells in the pharyngeal arches. Hoxa3 mutant homozygotes have defects in the development of all three organs. Roles for the Hoxa3 paralogs, Hoxb3 and Hoxd3, were investigated by examining various mutant combinations. The thyroid defects seen in Hoxa3 single mutants are exacerbated in double mutants with either of its paralogs, although none of the double-mutant combinations resulted in thyroid agenesis. The results indicate that the primary role of these genes in thyroid development is their effect on the development and migration of the ultimobranchial bodies, which contribute the parafollicular or C-cells to the thyroid. Hoxb3, Hoxd3 double mutants show no obvious defects in the thymus or parathyroids. However, the removal of one functional copy of Hoxa3 from the Hoxb3, Hoxd3 double mutants (Hoxa3 +/-, Hoxb3-/-, Hoxd3-/-) results in the failure of the thymus and parathyroid glands to migrate to their normal positions in the throat. Very little is known about the molecular mechanisms used to mediate the movement of tissues during development. These results indicate that Hoxa3, Hoxb3, and Hoxd3 have highly overlapping functions in mediating the migration of pharyngeal organ primordia. In addition, Hoxa3 has a unique function with respect to its paralogs in thymus, parathyroid, and thyroid development. This unique function may be conferred by the expression of Hoxa3, but not Hoxb3 nor Hoxd3, in the pharyngeal pouch endoderm.

Have gene knockouts caused evolutionary reversals in the mammalian first arch?

Many recent gene knockout experiments cause anatomical changes to the jaw region of mice that several investigators claim are evolutionary reversals. Here we evaluate these mutant phenotypes and the assertions of atavism. We argue that following the knockout of Hoxa-2, Dlx-2, MHox, Otx2, and RAR genes, ectopic cartilages arise as secondary consequences of disruptions in normal processes of cell specification, migration, or differentiation. These disruptions cause an excess of mesenchyme to accumulate in a region through which skeletal progenitor cells usually migrate, and at a site of condensation that is normally present in mammals but that is too small to chondrify. We find little evidence that these genes, when disrupted, cause a reversion to any primitive condition and although changes in their expression may have played a role in the evolution of the mammalian jaw, their function during morphogenesis is not sufficiently understood to confirm such hypotheses.

Equivalence in the genetic control of hindbrain segmentation in fish and mouse

The vertebrate hindbrain is subdivided into a series of rhombomeres whose segmental organization serves to pattern the architecture and innervation of the developing head. The zebrafish gene valentino is required cell-autonomously in the development of rhombomeres 5 and 6, and valentino mutants lack visible hindbrain segmentation caudal to the r3/4 boundary (Moens, C. B., Yan, Y.-L., Appel, B., Force, A. G., and Kimmel, C. B. (1996) Development 122, 3981–3990). Here we show that valentino is the zebrafish homologue of the mouse segmentation gene kreisler, which encodes a bZip transcription factor. The valentino gene is expressed in a manner consistent with its proposed role in subdividing rhombomeres 5 and 6 from their common precursor ‘proto-segment’ in the presumptive hindbrain, a process that we also demonstrate is reflected in the normal order of appearance of rhombomere boundaries. As well as having similar phenotypes with respect to visible hindbrain segmentation and patterns of marker gene expression, valentino and kreisler mutants have similar pharyngeal arch and inner ear defects, consistent with a conserved role for this gene in hindbrain segmentation and in patterning of the head periphery.

Early expression of Osteopontin in the chick is restricted to rhombomeres 5 and 6 and to a subpopulation of neural crest cells that arise from these segments

BACKGROUND

Osteopontin previously has been shown to be expressed in the early notochord and rostral hindbrain during formation of the rat neuraxis. Both of these sites are known to have important organizing capacities associated with the expression of a number of transcription factors and secreted signaling molecules. Thus multiple types of molecules are involved in neuraxis patterning. Osteopontin is an extracellular matrix molecule, whose expression within the early hindbrain and notochord suggests that it too could play a role in patterning of the neuraxis.

METHODS

By wholemount in situ hybridization, we analyzed osteopontin expression in the chick embryo at Hamburger and Hamilton (HH) stages 5-15, and patterns of expression were compared to those of other hindbrain and neural crest cell markers. Early developmental expression was further examined in the rat by wholemount in situ hybridization and immunohistochemistry.

RESULTS

Our results have demonstrated that whereas notochord expression of osteopontin is not conserved between rat and chick, expression was clearly present in the chick caudal hindbrain between HH stages 9 and 11. In the chick, osteopontin expression was restricted to rhombomeres 5 and 6 and to a subpopulation of neural crest cells arising from these rhombomeres, up to HH stage 13. Osteopontin expression was present in a cranial-to-caudal gradient, and in the dorsoventral plane, expression was absent from both the floor plate and roof plate, with the exception of occasional cells within the roof plate.

CONCLUSIONS

Localized expression in the 5th and 6th rhombomeres suggests that osteopontin may have a role in patterning of postotic neural crest cells that arise from these segments of the hindbrain. In addition, comparison of our current data with those obtained earlier in the rat suggests that there are significant differences either between the developmental role(s) of osteopontin in the chick and in the rat, or between osteopontin gene products in the two species.

Zebrafish hox genes: expression in the hindbrain region of wild-type and mutants of the segmentation gene, valentino

The developing hindbrain is organized into a series of segments termed rhombomeres which represent lineage restricted compartments correlating with domains of gene expression and neuronal differentiation. In this study, we investigate the processes of hindbrain segmentation and the acquisition of segmental identity by analyzing the expression of zebrafish hox genes in the hindbrains of normal fish and fish with a loss-of-function mutation in the segmentation gene valentino (val, the homologue of mouse kreisler; Moens, C. B., Cordes, S. P. Giorgianni, M. W., Barsh, G. S. and Kimmel, C. B. (1998). Development 125, 381–391). We find that zebrafish hox genes generally have similar expression profiles to their murine and avian counterparts, although there are several differences in timing and spatial extent of expression which may underlie some of the functional changes that have occurred along the separate evolutionary lineages of teleosts and tetrapods. Our analysis of hox gene expression in val- embryos confirms that the val gene product is important for subdivision of the presumptive rhombomere 5 and 6 territory into definitive rhombomeres, suggests that the val gene product plays a critical role in regulating hox gene transcription, and indicates that some neural crest cells are inappropriately specified in val- embryos. Our analysis of gene expression at several developmental stages has allowed us to infer differences between primary and secondary defects in the val mutant: we find that extended domains of expression for some hox genes are secondary, late phenomena potentially resulting from inappropriate cell mixing or lack of normal inter-rhombomeric interactions in the caudal hindbrain.

Segmental expression of the EphA4 (Sek-1) receptor tyrosine kinase in the hindbrain is under direct transcriptional control of Krox-20

Segmentation of the vertebrate hindbrain leads to the formation of a series of rhombomeres (r) with distinct identities. Recent studies have uncovered regulatory links between transcription factors governing this process, but little is known of how these relate to molecules mediating cell-cell signalling. The Eph receptor tyrosine kinase gene EphA4 (Sek-1) is expressed in r3 and r5, and function-blocking experiments suggest that it is involved in restricting intermingling of cells between odd- and even-numbered rhombomeres. We have analysed the cis-acting regulatory sequences of the EphA4 gene in transgenic mice and identified a 470 bp enhancer element that drives specific expression in r3 and r5. Within this element, we have identified eight binding sites for the Krox-20 transcription factor that is also expressed in r3 and r5. Mutation of these binding sites abolishes r3/r5 enhancer activity and ectopic expression of Krox-20 leads to ectopic activation of the enhancer. These data indicate that Krox-20 is a direct transcriptional activator of EphA4. Together with evidence that Krox-20 regulates Hox gene expression, our findings reveal a mechanism by which the identity and movement of cells are coupled such that sharply restricted segmental domains are generated.

DeltaEF1, a zinc finger and homeodomain transcription factor, is required for skeleton patterning in multiple lineages

DeltaEF1 is a DNA binding protein containing a homeodomain and two zinc finger clusters, and is regarded as a vertebrate homologue of zfh-1 (zinc finger homeodomain-containing factor-1) in Drosophila. In the developing embryo, deltaEF1 is expressed in the notochord, somites, limb, neural crest derivatives and a few restricted sites of the brain and spinal cord. To elucidate the regulatory function of deltaEF1 in mouse embryogenesis, we generated deltaEF1 null mutant (deltaEF1null(lacZ)) mice. The deltaEF1null(lacZ) homozygotes developed to term, but never survived postnatally. In addition to severe T cell deficiency of the thymus, the deltaEF1null(lacZ) homozygotes exhibited skeletal defects of various lineages. (1) Craniofacial abnormalities of neural crest origin: cleft palate, hyperplasia of Meckel's cartilage, dysplasia of nasal septum and shortened mandible. (2) Limb defects: shortening and broadening of long bones, fusion of carpal/tarsal bone and fusion of joints. (3) Fusion of ribs. (4) Sternum defects: split and asymmetric ossification pattern of the sternebrae associated with irregular sternocostal junctions. (5) Hypoplasia of intervertebral discs. These results indicate that deltaEF1 has an essential role in regulating development of these skeletal structures. Since the skeletal defects were not observed in deltaEF1deltaC727 mice, deltaEF1 bears distinct regulatory activities which are dependent on different domains of the molecule.

Early postnatal lethality in Hoxa-5 mutant mice is attributable to respiratory tract defects

To uncover roles for the Hoxa-5 gene during embryogenesis, we have focused on identifying structural and functional defects in organ systems underlying the perinatal lethality in Hoxa-5 homozygous mutants. Analysis of the mutant phenotype shows that Hoxa-5 is essential for normal organogenesis and function of the respiratory tract. In homozygous newborn mutants, improper tracheal and lung morphogenesis can lead to tracheal occlusion, and to respiratory distress associated with a marked decrease in the production of surfactant proteins. Collectively, these defects likely underlie the pronounced mortality of homozygous mutant pups. Furthermore, the loss of Hoxa-5 function results in altered TTF-1, HNF-3 beta, and N-myc gene expression in the pulmonary epithelium. Since expression of Hoxa-5 is confined to the mesenchymal component of the developing trachea and lung, the effects observed in epithelial cells may result from a disruption of normal epithelial-mesenchymal interactions.

Hox group 3 paralogous genes act synergistically in the formation of somitic and neural crest-derived structures

Hox genes encode transcription factors that are used to regionalize the mammalian embryo. Analysis of mice carrying targeted mutations in individual and multiple Hox genes is beginning to reveal a complex network of interactions among these closely related genes which is responsible for directing the formation of spatially restricted tissues and structures. In this report we present an analysis of the genetic interactions between all members of the third paralogous group, Hoxa3, Hoxb3, and Hoxd3. Previous analysis has shown that although mice homozygous for loss-of-function mutations in either Hoxa3 or Hoxd3 have no defects in common, mice mutant for both genes demonstrate that these two genes strongly interact in a dosage-dependent manner. To complete the analysis of this paralogous gene family, mice with a targeted disruption of the Hoxb3 gene were generated. Homozygous mutants have minor defects at low penetrance in the formation of both the cervical vertebrae and the IXth cranial nerve. Analysis and comparison of all double-mutant combinations demonstrate that all three members of this paralogous group interact synergistically to affect the development of both neuronal and mesenchymal neural crest-derived structures, as well as somitic mesoderm-derived structures. Surprisingly, with respect to the formation of the cervical vertebrae, mice doubly mutant for Hoxa3 and Hoxd3 or Hoxb3 and Hoxd3 show an indistinguishable defect, loss of the entire atlas. This suggests that the identity of the specific Hox genes that are functional in a given region may not be as critical as the total number of Hox genes operating in that region.

Dlx genes encode DNA-binding proteins that are expressed in an overlapping and sequential pattern during basal ganglia differentiation

The Dlx gene family encodes homeodomain proteins that are required for forebrain and craniofacial development. Towards elucidating the roles for each of these genes, we have isolated cDNA clones encoding the full-coding sequence for murine Dlx-5 and partial coding sequence for murine Dlx-6. Three different classes of sense Dlx-5 cDNA clones were characterized, two of which lack the homeobox. We also identified an antisense Dlx-6 transcript. Genomic analysis shows that the Dlx-5 and -6 genes are linked. Biochemical analysis using gel shift assays demonstrate that DLX-1, -2 and -5 have very similar DNA-binding properties. The expression of Dlx-1, -2, -5, -6 and antisense Dlx-1 and -6 was studied in the midgestation mouse brain. We found that the Dlx genes are expressed in overlapping patterns at different stages of differentiation within the primordia of the basal ganglia. Dlx-1 and -2 are expressed in the least mature cells (in the ventricular and subventricular zones). Dlx-5 appears to be co-expressed with Dlx-1 and -2 in the SVZ, but is also expressed in the postmitotic cells of the mantle. Dlx-6 expression is strongest in the mantle. Antisense Dlx-1 and -6 have their highest expression in the SVZ. These results suggest that each of these Dlx genes may have a distinct role in different steps of differentiation in the basal ganglia.

Role of Dlx-1 and Dlx-2 genes in patterning of the murine dentition

The molecular events of odontogenic induction are beginning to be elucidated, but until now nothing was known about the molecular basis of the patterning of the dentition. A role for Dlx-1 and Dlx-2 genes in patterning of the dentition has been proposed with the genes envisaged as participating in an ‘odontogenic homeobox gene code’ by specifying molar development. This proposal was based on the restricted expression of the genes in molar ectomesenchyme derived from cranial neural crest cells prior to tooth initiation. Mice with targeted null mutations of both Dlx-1 and Dlx-2 homeobox genes do not develop maxillary molar teeth but incisors and mandibular molars are normal. We have carried out heterologous recombinations between mutant and wild-type maxillary epithelium and mesenchyme and show that the ectomesenchyme underlying the maxillary molar epithelium has lost its odontogenic potential. Using molecular markers of branchial arch neural crest (Barx1) and commitment to chondrogenic differentiation (Sox9), we show that this population alters its fate from odontogenic to become chondrogenic. These results provide evidence that a subpopulation of cranial neural crest is specified as odontogenic by Dlx-1 and Dlx-2 genes. Loss of function of these genes results in reprogramming of this population of ectomesenchyme cells into chondrocytes. This is the first indication that the development of different shaped teeth at different positions in the jaws is determined by independent genetic pathways.

Gene dosage-dependent effects of the Hoxa-13 and Hoxd-13 mutations on morphogenesis of the terminal parts of the digestive and urogenital tracts

Gene targeting experiments have shown that the murine Hoxa-13 and Hoxd-13 paralogous genes control skeletal patterning in the distal region of the developing limbs. However, both genes are also expressed in the terminal part of the digestive and urogenital tracts during embryogenesis and postnatal development. Here, we report the abnormalities occuring in these systems in Hoxa-13(−/−) and Hoxa-13/Hoxd-13 compound mutant mice. Hoxa-13(−/−) mutant fetuses show agenesis of the caudal portion of the Mullerian ducts, lack of development of the presumptive urinary bladder and premature stenosis of the umbilical arteries, which could account for the lethality of this mutation at mid-gestational stages. Due to such lethality, only Hoxa-13(+/−)/Hoxd-13(−/−) compound mutants can reach adulthood. These compound mutants display: (i) agenesis or hypoplasia of some of the male accessory sex glands, (ii) malpositioning of the vaginal, urethral and anal openings, and improper separation of the vagina from the urogenital sinus, (iii) hydronephrosis and (iv) anomalies of the muscular and epithelial layers of the rectum. Thus, Hoxa-13 and Hoxd-13 play important roles in the morphogenesis of the terminal part of the gut and urogenital tract. While Hoxa-13(−/−)/Hoxd-13(+/−) fetuses show severely impaired development of the urogenital sinus, double null (Hoxa-13[−/−]/Hoxd-13[−/−]) fetuses display no separation of the terminal (cloacal) hindgut cavity into a urogenital sinus and presumptive rectum, and no development of the genital bud, thereby demonstrating that both genes act, in a partly redundant manner, during early morphogenesis of posterior trunk structures.

Molecular characterization of four induced alleles at the Ednrb locus

The piebald locus on mouse chromosome 14 encodes the endothelin-B receptor (EDNRB), a G protein-coupled, seven-transmembrane domain protein, which is required for neural crest-derived melanocyte and enteric neuron development. A spontaneous null allele of Ednrb results in homozygous mice that are predominantly white and die as juveniles from megacolon. To identify the important domains for EDNRB function, four recessive juvenile lethal alleles created by either radiation or chemical mutagens (Ednrb27Pub, Ednrb17FrS, Ednrb1Chlc, and Ednrb3Chlo) were examined at the molecular level. Ednrb27Pub mice harbor a mutation at a critical proline residue in the fifth transmembrane domain of the EDNRB protein. A gross genomic alteration within the Ednrb gene in Ednrb3Chlo results in the production of aberrantly sized transcripts and no authentic Ednrb mRNA. Ednrb17FrS mice exhibited a decreased level of Ednrb mRNA, supporting previous observations that the degree of spotting in piebald mice is dependent on the amount of EDNRB expressed. Finally, no molecular defect was detected in Ednrb1Chlc mice, which produce normal levels of Ednrb mRNA in adult brain, suggesting that the mutation affects important regulatory elements that mediate the expression of the gene during development.

Cis-acting elements conserved between mouse and pufferfish Otx2 genes govern the expression in mesencephalic neural crest cells

Previous studies suggested that the Otx2 gene plays an essential role in the development of cranial skeletons and nerves of mesencephalic neural crest origin. To clarify this role, we have identified the cis-acting elements in mouse and pufferfish Otx2 genes responsible for the expression in the crest cells using a transgenic approach with the lacZ reporter gene. In mouse, 49 bp sequences in the proximal 5′ region upstream were essential and sufficient to direct the transgene expression in the cephalic mesenchyme. In pufferfish, the 1.1 kb distal region, located far downstream (from +14.4 to +15.5 kb), had almost identical activity. Between them, several DNA sequences were conserved, and mutational analyses indicated that motif A was critical for the transgene expression in the premandibular region while motif B was critical in both premandibular and mandibular regions. Motif B, CTAATTA, contains the core motif for binding of homeodomain proteins while motif A, TAAATCTG, does not match any known consensus binding sequences for transcriptional factors. The cephalic mesenchyme that expressed beta-galactosidase under these cis-elements is most likely to correspond to mesencephalic crest cells. Thus the molecular machinery regulating Otx2 expression in these cells appears to be conserved between mouse and fish, implying a crucial role of the Otx2 gene in development of the neural-crest-derived structures of the gnathostome rostral head.

Segmental identity can change independently in the hindbrain and rhombencephalic neural crest

In this study we tested whether the segmental identities of the hindbrain and its derived neural crest are necessarily linked or, instead, if they can be altered independently. Using morphological criteria, we show that the hindbrains of Hoxa-2 mutant mice, in which the second arch skeletal derivatives assume first arch characteristics (Gendron-Maguire et al. [1993] Cell 75:1317-1331; Rijli et al. [1993] Cell 75:1333-1349), retain normal segmental identities. Also, by phenotypic analysis, we show that, with retinoic acid, changes can be induced in the identity of the preotic hindbrain without effects in its derived neural crest. Our data thus indicate that identity changes in the hindbrain and branchial arch neural crest can occur independently. Moreover, if Hoxa-2 is concomitantly induced by retinoic acid in the first branchial arch, the proximal derivatives of this arch are also affected. We propose a model for the patterning of the branchial region, according to which the segmental identity in this area is provided mainly by the branchial arches.

Homeobox genes in embryogenesis and pathogenesis

The homeobox, a 60-amino acid-encoding DNA sequence, originally discovered in the genome of the fruit fly Drosophila, was subsequently identified throughout the three kingdoms of multicellular organisms. Homeobox-containing genes encode DNA-binding proteins that regulate gene expression and control various aspects of morphogenesis and cell differentiation. In particular, the Hox family of clustered homeobox genes plays a fundamental role in the morphogenesis of the vertebrate embryo, providing cells with regional information along the main body axis. The nonclustered or divergent homeobox genes include a large number of genes scattered throughout the genome that, nevertheless, can be organized in distinct families based on their homologies and functional similarities. This review will provide the reader with a brief overview on some recent studies aimed at understanding the functional role of homeobox genes in normal mammalian development as well as their involvement in congenital malformations and oncogenesis.

Role of Hoxa-2 in axon pathfinding and rostral hindbrain patterning

Segmentation plays an important role in neuronal diversification and organisation in the developing hindbrain. For instance, cranial nerve branchiomotor nuclei are organised segmentally within the basal plates of successive pairs of rhombomeres. To reach their targets, motor axons follow highly stereotyped pathways exiting the hindbrain only via specific exit points in the even-numbered rhombomeres. Hox genes are good candidates for controlling this pathfinding, since they are segmentally expressed and involved in rhombomeric patterning. Here we report that in Hoxa-2(−/−) embryos, the segmental identities of rhombomere (r) 2 and r3 are molecularly as well as anatomically altered. Cellular analysis by retrograde dye labelling reveals that r2 and r3 trigeminal motor axons turn caudally and exit the hindbrain from the r4 facial nerve exit point and not from their normal exit point in r2. Furthermore, dorsal r2-r3 patterning is affected, with loss of cochlear nuclei and enlargement of the lateral part of the cerebellum. These results point to a novel role for Hoxa-2 in the control of r2-r3 motor axon guidance, and also suggest that its absence may lead to homeotic changes in the alar plates of these rhombomeres.

Targeted disruption of the mouse homologue of the Drosophila polyhomeotic gene leads to altered anteroposterior patterning and neural crest defects

The rae28 gene is a mouse homologue of the Drosophila polyhomeotic gene (Nomura, M., Takihara, Y. and Shimada, K. (1994) Differentiation 57, 39–50), which is a member of the Polycomb group (Pc-G) of genes (DeCamillis, M., Cheng, N., Pierre, D. and Brock, H.W. (1992) Genes Dev. 6, 223–232). The Pc-G genes are required for the correct expression of the Homeotic complex genes and segment specification during Drosophila embryogenesis and larval development. To study the role of the rae28 gene in mouse development, we generated rae28-deficient mice by gene targeting in embryonic stem cells. The rae28−/− homozygous mice exhibited perinatal lethality, posterior skeletal transformations and defects in neural crest-related tissues, including ocular abnormalities, cleft palate, parathyroid and thymic hypoplasia and cardiac anomalies. The anterior boundaries of Hoxa-3, a-4, a-5, b-3, b-4 and d-4 expression were shifted rostrally in the paraxial mesoderm of the rae28−/− homozygous embryos, and those of Hoxb-3 and b-4 expression were also similarly altered in the rhombomeres and/or pharyngeal arches. These altered Hox codes were presumed to be correlated with the posterior skeletal transformations and neural crest defects observed in the rae28−/− homozygous mice. These results indicate that the rae28 gene is involved in the regulation of Hox gene expression and segment specification during paraxial mesoderm and neural crest development.