Analysis of Msx1; Msx2 double mutants reveals multiple roles for Msx genes in limb development

The homeobox-containing genes Msx1 and Msx2 are highly expressed in the limb field from the earliest stages of limb formation and, subsequently, in both the apical ectodermal ridge and underlying mesenchyme. However, mice homozygous for a null mutation in either Msx1 or Msx2 do not display abnormalities in limb development. By contrast, Msx1; Msx2 double mutants exhibit a severe limb phenotype. Our analysis indicates that these genes play a role in crucial processes during limb morphogenesis along all three axes. Double mutant limbs are shorter and lack anterior skeletal elements (radius/tibia, thumb/hallux). Gene expression analysis confirms that there is no formation of regions with anterior identity. This correlates with the absence of dorsoventral boundary specification in the anterior ectoderm, which precludes apical ectodermal ridge formation anteriorly. As a result, anterior mesenchyme is not maintained, leading to oligodactyly. Paradoxically, polydactyly is also frequent and appears to be associated with extended Fgf activity in the apical ectodermal ridge, which is maintained up to 14.5 dpc. This results in a major outgrowth of the mesenchyme anteriorly, which nevertheless maintains a posterior identity, and leads to formation of extra digits. These defects are interpreted in the context of an impairment of Bmp signalling.     

FLRT3 as a key player on chick limb development

Limb outgrowth is maintained by a specialized group of cells, the apical ectodermal ridge (AER), a thickening of the limb epithelium at its distal tip. It has been shown that fibroblast growth factor (FGF) activity and activation of the Erk pathway are crucial for AER function. Recently, FLRT3, a transmembrane protein able to interact with FGF receptors, has been implicated in the activation of ERK by FGFs. In this study, we show that flrt3 expression is restricted to the AER, co-localizing its expression with fgf8 and pERK activity. Loss-of-function studies have shown that silencing of flrt3 affects the integrity of the AER and, subsequently, its proper function during limb bud outgrowth. Our data also indicate that flrt3 expression is not regulated by FGF activity in the AER, whereas ectopic WNT3A is able to induce flrt3 expression. Overall, our findings show that flrt3 is a key player during chicken limb development, being necessary but not sufficient for proper AER formation and maintenance under the control of BMP and WNT signalling.     

Function of FGF-4 in limb development

The apical ectodermal ridge plays a central role in limb development through its interactions with the underlying mesenchyme. Removal of the AER results in cessation of limb outgrowth and leads to truncation of the limb along the proximo-distal axis. The many functions attributed to the ridge include maintenance of the progress zone mesenchyme. Here, cells are stimulated to proliferate, are maintained in an undifferentiated state, and are assigned progressively more distal positional values as the limb grows. The AER also functions to maintain the activity of the polarizing region, a region of mesenchyme which is thought to provide the primary signal for patterning along the antero-posterior axis. We have begun to explore the function of fibroblast growth factor-4 (FGF-4) during limb development. FGF-4, which encodes an efficiently secreted protein, is expressed in the AER. We have previously demonstrated that FGF-4 protein can stimulate limb mesenchyme proliferation and can induce the expression of a downstream homeobox gene, Evx-1 (homologue of the Drosophila even-skipped gene), that is normally regulated by a signal from the AER. To determine to what extent FGF-4 protein can substitute for the AER to allow normal limb outgrowth, we performed experiments on the developing chick limb in ovo. Remarkably, we find that after AER removal, the FGF-4 protein can provide all the signals required for virtually normal outgrowth and patterning of the limb. Further studies indicate that proliferation of progress zone cells is not sufficient, and that an additional signal is produced by the posterior mesenchyme in response to FGF-4 which enables progress zone cells to acquire progressively more distal fates. Thus FGF-4 maintains progress zone activity through a combination of at least two signals—one that acts directly on progress zone cells to stimulate their proliferation, and one that acts indirectly by maintaining the production of patterning signal(s) by the posterior mesenchyme. We further show that failure of the posterior mesenchyme to produce this signal correlates with failure to maintain polarizing activity. This raises the possibility that the signal produced by the posterior mesenchyme and required for progressive proximo-distal limb patterning is identical to the polarizing activity. Further experiments demonstrate that retinoic acid, which mimics the activity of the polarizing region, can supply this signal. In conclusion, the finding that a single growth factor can serve as both the direct and indirect signals required to maintain progress zone activity provides a simple mechanism for ensuring that growth and pattern formation are linked in the developing limb. © 1994 Wiley-Liss, Inc.     

Flrt2 and Flrt3 have overlapping and non-overlapping expression during craniofacial development

Craniofacial morphogenesis is a complex multi-step process that involves numerous biological processes to coordinate the growth, proliferation, migration, and subsequent differentiation of the cranial neural crest cells. Members of the Fibronectin Leucine-Rich Transmembrane (Flrt) gene family have been previously reported to be widely expressed in the developing embryo. We mapped the expression of Flrt2 and Flrt3 at critical stages of craniofacial development and found that, during early craniofacial development, Flrt2 was highly expressed initially in the cranial neural crest cells and Flrt3 in the midbrain. Later both genes were expressed in the developing pharyngeal region. Flrt2 expression predominated in the neural crest-derived mesenchyme in the medial aspect of the developing frontonasal region in close relationships with the expression of Fgfr2, Shh, and Msx1, three genes shown previously to play critical roles in craniofacial development. Flrt2 was also present in the vomero-nasal organ, mandibular primodia, and the posterior aspects of the unfused and fused secondary palatal shelves. Flrt3, however, had a more restrictive expression, being present in the mesenchyme underlying the ectoderm of the medial nasal process and in the mandibular primordium and in regions undergoing outgrowth, in a pattern that overlapped with Bmp4 expression. Both Flrt2 and Flrt3 were later found to be present at sites of epithelial–mesenchymal interactions such as the developing tooth buds, hair follicles, and eye. Together the data suggested important roles for Flrt2 and Flrt3 in mediating events such as NCC migration, chondrogenesis and epithelial–mesenchymal interactions during craniofacial development.     

Growth arrest specific gene 1 acts as a region-specific mediator of the Fgf10/Fgf8 regulatory loop in the limb

Proximal-to-distal growth of the embryonic limbs requires Fgf10 in the mesenchyme to activate Fgf8 in the apical ectodermal ridge (AER), which in turn promotes mesenchymal outgrowth. We show here that the growth arrest specific gene 1 (Gas1) is required in the mesenchyme for the normal regulation of Fgf10/Fgf8. Gas1 mutant limbs have defects in the proliferation of the AER and the mesenchyme and develop with small autopods, missing phalanges and anterior digit syndactyly. At the molecular level, Fgf10 expression at the distal tip mesenchyme immediately underneath the AER is preferentially affected in the mutant limb, coinciding with the loss of Fgf8 expression in the AER. To test whether FGF10 deficiency is an underlying cause of the Gas1 mutant phenotype, we employed a limb culture system in conjunction with microinjection of recombinant proteins. In this system, FGF10 but not FGF8 protein injected into the mutant distal tip mesenchyme restores Fgf8 expression in the AER. Our data provide evidence that Gas1 acts to maintain high levels of FGF10 at the tip mesenchyme and support the proposal that Fgf10 expression in this region is crucial for maintaining Fgf8 expression in the AER.     

Differential expression of chicken CYP26 in anterior versus posterior limb bud in response to retinoic acid

Multiple studies indicate that quantitative control of the levels of all-trans-retinoic acid (RA) in the vertebrate embryo is necessary for correct development. The function of RA in cells is regulated by a number of coordinated mechanisms. One of those mechanisms involves controls on the rate of RA catabolism. Recently, enzymes capable of catabolizing RA were found to constitute a new family, called CYP26, within the cytochrome P450 superfamily. CYP26 homologues have been isolated from human, mouse, zebra fish, and recently from the chick. In this study, we examined the regulation of chicken CYP26 (cCYP26) expression by RA during the early phase of chick limb outgrowth. In the anterior limb mesenchyme and apical ectodermal ridge (AER), cCYP26 expression was induced in a concentration dependent manner by implanting beads soaked in 0.1, 1, and 5 mg/ml RA. The RA-induced expression of cCYP26 in anterior limb mesenchyme and the AER was detected as early as 1 hr after treatment and was not affected by the presence of cycloheximide. In contrast to the anterior limb, the induction of cCYP26 was dramatically reduced (or absent) when RA beads were implanted in the posterior limb mesenchyme. Furthermore, induction of cCYP26 expression in the anterior mesenchyme was inhibited by transplantations of the zone of polarizing activity (ZPA) and by Shh-soaked beads. Our data suggest that different mechanisms regulate retinoid homeostasis in the AER and mesenchyme during limb bud outgrowth. J. Exp. Zool. 290:136-147, 2001.     

Regulated expression of FLRT genes implies a functional role in the regulation of FGF signalling during mouse development

Within the mammalian genome, there are many multimember gene families that encode membrane proteins with extracellular leucine rich repeats which are thought to act as cell adhesion or signalling molecules. We previously showed that the members of the NLRR gene family are expressed in a developmentally restricted manner in the mouse with NLRR-1 being expressed in the developing myotome. The FLRT gene family shows a similar genomic layout and predicted protein secondary structure to the NLRRs so we analysed expression of the three FLRT genes during mouse development. FLRTs are glycosylated membrane proteins expressed at the cell surface which localise in a homophilic manner to cell-cell contacts expressing the focal adhesion marker vinculin. Each member of the FLRT family has a distinct, highly regulated expression pattern, as was seen for the NLRR family. FLRT3 has a provocative expression pattern during somite development being expressed in regions of the somite where muscle precursor cells migrate from the dermomyotome and move into the myotome, and later in myotomal precursors destined to migrate towards their final destination, for example, those that form the ventral body wall. FLRT3 is also expressed at the midbrain/hindbrain boundary and in the apical ectodermal ridge, regions where FGF signalling is known to be important, suggesting that the role for FLRT3 in FGF signalling identified in Xenopus is conserved in mammals. FLRT1 is expressed at brain compartmental boundaries and FLRT2 is expressed in a subset of the sclerotome, adjacent to the region that forms the syndetome, suggesting that interaction with FGF signalling may be a general property of FLRT proteins. We confirmed this by showing that all FLRTs can interact with FGFR1 and FLRTs can be induced by the activation of FGF signalling by FGF-2. We conclude that FLRT proteins act as regulators of FGF signalling, being induced by the signal and then able to interact with the signalling receptor, in many tissues during mouse embryogenesis. This process may, in part, be dependent on homophilic intercellular interactions between FLRT molecules.     

Pattern formation in epithelial development: the vertebrate limb and feather bud spacing.

The ectoderm of the vertebrate limb and feather bud are epithelia that provide good models for epithelial patterning in vertebrate development. At the tip of chick and mouse limb buds is a thickening, the apical ectodermal ridge, which is essential for limb bud outgrowth. The signal from the ridge to the underlying mesoderm involves fibroblast growth factors. The non-ridge ectoderm specifies the dorsoventral pattern of the bud and Wnt7a is a dorsalizing signal. The development of the ridge involves an interaction between dorsal cells that express radical fringe and those that do not. There are striking similarities between the signals and genes involved in patterning the limb ectoderm and the epithelia of the Drosophila imaginal disc that gives rise to the wing. The spacing of feather buds involves signals from the epidermis to the underlying mesenchyme, which again include Wnt7a and fibroblast growth factors.     

Epithelial-mesenchymal interactions in the outgrowth of limb buds and facial primordia in chick embryos.

The facial primordia in the chick embryo begin as rounded swellings that surround the primitive mouth and these grow out to form the beak. The control of proximodistal outgrowth is not well understood but may involve similar mechanisms to the limb bud. In order to test this hypothesis, combinations were made between epithelium and mesenchyme from facial primordia and limb buds. Signals from all three types of facial mesenchyme (frontonasal mass, mandibular, and maxillary) maintained the thickened apical ectodermal ridge of limb epithelium for up to 48 h. Combinations of tissues from the frontonasal mass mesenchyme and limb epithelium underwent substantial and correct morphogenesis. In contrast, poor development was observed in combinations with mandibular mesenchyme. Signals from frontonasal mass epithelium promoted outgrowth and morphogenesis of limb mesenchyme whereas mandibular and maxillary epithelium did not support joint morphogenesis. The results suggest that signals employed in the epithelial-mesenchymal interactions in facial primordia are similar but not identical to those signals used in the limb bud.     

Expression of the 9G1 antigen in the apical cap of axolotl regenerates requires nerves and mesenchyme

Monoclonal antibody 9G1 (mAb 9G1) is reactive to the wound epithelium of axolotl larvae and therefore provided the opportunity to examine the interaction between the wound epithelium, nerves, and blastemal mesenchyme during axolotl limb regeneration. In unamputated limbs, mAb 9G1 is reactive to most or all cells of the dermis, skeletal elements, blood vessels, and nerves, to a few unidentified cells in muscle, and to none in epidermis. During regeneration of axolotl limbs, mAb 9G1 reacts strongly to an intracellular antigen of the blastemal mesenchyme and of the distal-most portion of the wound epithelium, the so-called apical epithelial cap (AEC). Because this thickened wound epithelium of regenerating amphibian limbs has been suggested as functioning in a manner similar to the apical ectodermal ridge (AER) of embryonic limb buds, it was of interest to further examine the reactivity of mAb 9G1 during various stages of regeneration. Whether mAb 9G1 reactivity in the AEC depended on mesenchyme and/or nerves was also tested. Monoclonal antibody 9G1 reactivity appears in the AEC of regenerating limbs prior to outgrowth of the blastema and persists throughout blastemal stages. Apical epithelial cap reactivity to mAb 9G1 is nerve dependent during early stages of blastema development and becomes nerve-independent at later stages. When epithelium-free blastemal mesenchyme is grafted onto injured flank musculature, ectopic limb regeneration occurs and the AEC derived from flank epidermis exhibits mAb 9G1 reactivity. These results show that a mAb 9G1 reactive AEC is characteristic of regenerating limbs and that expression of the 9G1 antigen by the AEC is dependent upon underlying blastemal mesenchyme and nerves.     

The influence of epithelia on cartilage and loose connective tissue formation by limb mesenchyme cultures

In order to explain the observation that normally nonchondrogenic limb mesenchyme forms extensive cartilage in culture, the possibility that limb ectoderm inhibits chondrogenesis is examined. Small pieces of quail or chick ectoderm are grafted onto micromass cultures of wing mesenchyme from stage 23–34 chick embryos. The presence of nonridge wing or several other epithelia results in increased collagen accumulation in the underlying mesenchyme, a delay in cell differentiation, and the eventual formation of loose connective tissue, as determined by transmission electron microscopy. The influence can occur across Nuclepore filters having 0.1-μm-diameter pores and ultrathin Millipore filters, but not across Millipore filters of standard thickness. The influence is not contact dependent since cell processes do not cross these filters. The apical ectodermal ridge has the additional effect of stimulating mesenchymal outgrowth. These results support the idea that the ectoderm plays a direct but negative role in the formation of a chondrogenic core within the developing limb.     

Transient expression of a cell surface heparan sulfate proteoglycan (syndecan) during limb development

Syndecan is an integral membrane proteoglycan that contains both heparan sulfate and chondroitin sulfate chains and that links the cytoskeleton to interstitial extracellular matrix components, including collagen and fibronectin. Immunohistochemistry with a monoclonal antibody directed to the core protein of the syndecan ectodomain has been used to analyze the distribution of this proteoglycan in the developing mouse limb bud and in high-density cultures of limb mesenchyme cells. By Day 9 of gestation when the limb buds are just apparent, syndecan is detected on cells throughout the limb region, including both ectodermal and mesenchymal components. This distribution does not change as the limb bud elongates along its proximodistal axis, except for its reduction in the apical ectodermal ridge. By Day 11, the intensity of immunofluorescence in the central core decreases relative to other regions. By Day 13 immunostaining is lost in the regions destined for chondrogenesis and myogenesis but persists in the limb ectoderm and peripheral and distal mesenchyme. In the limb mesenchyme cell cultures, syndecan is initially undetected, but is found throughout the culture by 24 hr. With further culture the antigen becomes reduced in chondrogenic foci and in association with myogenic cells. When chick limb ectoderm is placed on the high-density cultures, immunoreactivity in the mouse mesenchyme is enhanced suggesting that epithelial-mesenchymal interactions modulate syndecan expression in the limb bud. Based on analysis of 35S-labeled syndecan from the cultures, syndecan from limb mesenchyme cells contains more glycosaminoglycan chains and is larger in size than the previously described polymorphic forms of syndecan from various epithelia. The high affinity of syndecan for components of the extracellular matrix and its distribution in the early limb bud are consistent with a role in maintaining the morphologic integrity of the limb bud during the period of initiation and rapid outgrowth, and in preventing the onset of chondrogenesis.     

Altered expression of the chicken homeobox-containing genes GHox-7 and GHox-8 in the limb buds of limbless mutant chick embryos.

It has been suggested that the reciprocal expression of the chicken homeobox-containing genes GHox-8 and GHox-7 by the apical ectodermal ridge and subjacent limb mesoderm might be involved in regulating the proximodistal outgrowth of the developing chick limb bud. In the present study the expression of GHox-7 and GHox-8 has been examined by in situ and dot blot hybridization in the developing limb buds of limbless mutant chick embryos. The limb buds of homozygous mutant limbless embryos form at the proper time in development (stage 17/18), but never develop an apical ectodermal ridge, fail to undergo normal elongation, and eventually degenerate. At stage 18, which is shortly following the formation of the limb bud, the expression of GHox-7 is considerably reduced (about 3-fold lower) in the mesoderm of limbless mutant limb buds compared to normal limb bud mesoderm. By stages 20 and 21, as the limb buds of limbless embryos cease outgrowth, GHox-7 expression in limbless mesoderm declines to very low levels, whereas GHox-7 expression increases in the mesoderm of normal limb buds which are undergoing outgrowth. In contrast to GHox-7, expression of GHox-8 in limbless mesoderm at stage 18 is quantitatively similar to its expression in normal limb bud mesoderm, and in limbless and normal mesoderm GHox-8 expression is highly localized in the anterior mesoderm of the limb bud. In normal limb buds, GHox-8 is also expressed in high amounts by the apical ectodermal ridge.(ABSTRACT TRUNCATED AT 250 WORDS)     

Mitogenic property of the apical ectodermal ridge

While the apical ectodermal ridge (AER) is well known for its required role in the development of distal parts of the limb and for its ability to stimulate limb duplications, the mechanism of its action is unknown. In this study we use a culture system previously developed by M. Globus and S. Vethamany-Globus (1976, Differentiation6, 91–96) in which an AER grafted onto a high-density cell culture of limb mesenchyme stimulates the formation of an outgrowth. Time-lapse movies taken during the outgrowth period demonstrated no cellular activities other than cell division. Both the mitotic index and labeling index in the mesenchyme were significantly elevated under the AER as compared to that without AER, indicating that the AER provides a growth-promoting stimulus which increases the proportion of dividing cells. On the other hand, nonridge ectoderm had no detectable effect on the mitotic index. Treatment of cultures with cytosine arabinoside both inhibited DNA synthesis and prevented AER-induced outgrowth. These results demonstrate a mitogenic capacity of AER tissue and suggest that mesenchymal outgrowth requires this activity. The mitogenic property of the AER is considered in relation to limb outgrowth in situ.     

FGF10 can induce Fgf8 expression concomitantly with En1 and R-fng expression in chick limb ectoderm, independent of its dorsoventral specification

The limb bud has a thickened epithelium at the dorsal-ventral boundary, the apical ectodermal ridge (AER), which sustains limb outgrowth and patterning. A secreted molecule fibroblast growth factor (FGF)10 is involved in inducing Fgf8 expression in the prospective AER and mutual interaction between mesenchymal FGF10 and FGF8 in the AER is essential for limb outgrowth. A secreted factor Wnt7a and a homeobox protein Lmx1 are involved in the dorsal patterning of the limb, whereas a homeobox protein Engrailed 1 (En1) is involved in the dorsal-ventral patterning as well as AER formation. Radical fringe (R-fng), a vertebrate homolog of Drosophila fringe was also found to elaborate AER formation in chicks. However, little is known about the molecular interactions between these factors during AER formation. The present study clarified the relationship between FGF10, Wnt7a, Lmx1, R-fng and En1 during limb development using a foil-barrier insertion experiment. It was found that a foil-barrier inserted into the chick prospective wing mesenchyme lateral to the mesonephric duct blocks AER induction. This experiment was expanded by implanting Fgf10-expressing cells lateral to the barrier and examined whether FGF10 could rescue the expression of the limb-patterning genes reported in AER formation. It was found that FGF10 is sufficient to induce Fgf8 expression in the ectoderm of the foil-inserted limb bud, concomitantly with R-fng and En1 expression. However, FGF10 could not rescue the expression of the dorsal marker genes, Wnt7a or Lmx1. Thus, it is suggested that epithelial factors of En1 and R-fng can induce Fgf8 expression in the limb ectoderm in cooperation with a mesenchymal factor FGF10. Some factor(s) other than FGF10, possibly from the paraxial structures medial to the limb mesoderm, is responsible for the initial dorsal-ventral specification of the limb bud.