Search

Embryonic stem cells are uniquely endowed with the capacity of self-renewal and the potential to give rise to all possible cell types, including germ cells. These qualities have made mouse embryonic stem cells a valuable resource for genetic manipulation of the mouse genome. In addition, they present a powerful system for the in vitro dissection of mammalian embryonic development. The recent isolation of human embryonic stem cells has raised a lot of interest for the potential of transposing our knowledge of lineage-specific differentiation of embryonic stem cells to cell-based therapy of human disease. Recent reports have provided insights into the specific differentiation of embryonic stem cells to different cell types of the embryo. However, progress in this direction seems to depend on the knowledge of the mechanisms controlling lineage decisions during embryogenesis.

Mammalian and Drosophila homologues of Baf57 have been previously isolated as being a subunit of SWI/SNF-like chromatin remodeling complexes. Here, we report the cloning and developmental expression of Xenopus Baf57. We isolated XBaf57 by using an expression cloning approach to identify novel modulators of Xenopus Smad7. XBaf57 co-operates with XSmad7 by increasing the expression of neural markers in ectodermal explants. XBaf57 is expressed in the ectoderm and pre-involuting mesoderm during gastrula stages and in the central nervous system during neurula and tailbud stages. These results raise the possibility that XBaf57 (or XBaf57-containing chromatin remodelling complexes) may be involved in the process of neural induction during Xenopus embryonic development.

In Xenopus, endodermal cell fate is determined gradually from late blastula to early gastrula stages; cell-cell interaction plays an important role in this process. Here we use a cell dissociation assay to show that extracellular signaling is required continuously before endoderm determination. Activin and Vg1, but not BMP2 or basic FGF, rescue the expression of endodermal markers in dissociated cells when provided at the mid-blastula transition (MBT, the time in which zygotic transcription begins). Removal of exogenously added activin or Vg1 before MBT results in reduction of endodermal gene expression in dissociated vegetal cells. In vivo, endogenous endodermal markers are reduced in vegetal explants when activin-like signaling is blocked with dominant negative receptors. VegT, a maternal transcription factor shown to be critical for endoderm specification, relies on an active TGFbeta pathway to induce endoderm in animal caps. These results indicate that TGFbeta signaling may be activated by the maternally expressed VegT to participate in endoderm determination. In addition, VegT function seems to be required in parallel with the TGFbeta pathway, as overexpression of activin does not relieve endoderm repression by a dominant negative VegT mutant in vegetal cells. Our data suggest that maternal VegT first activates a zygotic TGFbeta signal, then cooperates with this signal to determine the endodermal cell fate.

Screening a maternal Xenopus expression library for activities that synergize with low levels of injected β -catenin, we have isolated a clone encoding the C-terminal end of x- β TrCP-2, a highly conserved protein belonging to the F-box/WD40 family of ubiquitin-ligase specificity factors. We show that x- β TrCP-2 expression reduces dorsal axis formation in Xenopus embryos. A dominant negative mutant lacking the F-box triggers the opposite effect, inducing secondary axes and activating the expression of Wnt responsive genes in ectodermal explants. In light of the existence of β TrCP transcripts associated with the vegetal cortex, we propose that β TrCP plays a fundamental role in the establishment of the dorsal determinants during cortical rotation in Xenopus .

We have identified a short segment of the mouse Pax-6 gene 5' flanking region that is necessary and sufficient for reporter construct expression in components of the eye derived from non-neural ectoderm. This transcriptional control element has a highly conserved nucleotide sequence over 341 bp and is located approximately 3.5 kb upstream of the start-point for transcription from the most proximal promoter (PO) of the Pax-6 gene. The level of identity between the human and mouse Pax-6 genes in this region is 93%. When combined either with its natural promoter or a heterologous minimal promoter, this element directs reporter construct expression to a region of surface ectoderm overlying the optic cup beginning at E8.5-9.0 (12-14 somites). Subsequently, expression is restricted to the lens (primarily the lens epithelium) and the corneal epithelium. This element will provide an important tool in future transgenic analyses of lens formation and will allow identification of transcription factors with a central function in lens development.

The specification of neural fate in Xenopus embryos has been shown to be under regulation by negative factors. The secreted protein bone morphogenetic protein-4 (BMP4) has been identified as one of these factors: in the early gastrula ectoderm, BMP4 can both inhibit neural fate and induce epidermis. In this study, we show that two other Xenopus BMP genes, BMP2 and BMP7, are endowed with the same types of activities. First, we show that expression of a dominant negative form of the BMP2 ligand, which blocks normal processing of the wild-type ligand, causes neuralization of Xenopus ectoderm. Second, we have isolated the Xenopus BMP2/7 receptor (XALK2) and generated a constitutively active mutant that signals in a ligand-independent manner. We show that signals from the activated BMP2/7 receptor also inhibit neuralization and induce epidermis in dissociated ectoderm cells. Consistent with both findings we show that secreted BMP2 and BMP7 ligands can also mediate neural inhibition and epidermal induction. These results suggest that both BMP2 and BMP7 may be involved independently or together with BMP4 in the inhibition of the neural fate and the onset of the epidermal induction pathway in vivo. This further supports the idea that epidermal induction is due to the effects of multiple signals from heterogeneous BMP genes.

The Smad proteins have been implicated in the intracellular signaling of transforming growth factor-β (TGF-β) ligands. Here we describe the function of Smad5 in early Xenopus development. Misexpression of Smad5 in the embryo causes ventralization and induces ventral mesoderm. Moreover, Smad5 induces epidermis in dissociated ectoderm cells which would otherwise form neural tissue. Both of these activities require Smad4 (DPC4) activity. We propose that Smad5 acts downstream of the BMP4 signaling pathway in Xenopus embryos and directs the formation of ventral mesoderm and epidermis.

Although noggin, follistatin, and chordin can neuralize by antagonizing BMP4 epidermalizing activity, is it possible that, in addition, they transduce a signal via a receptor, as was originally postulated for noggin and chordin? The default model would predict not. There are indeed three lines of evidence that strongly argue against the existence of receptors for follistatin, noggin, or chordin, at least in the pathway mediating neuralization, and that inhibition of BMP signaling is sufficient to unveil the neural fate. First, there is the evidence from cell dissociation experiments discussed above: when embryonic cells are dissociated for several hours they will make neural tissue (3xNeural differentation of Xenopus laevis ectoderm takes place after disaggregation and delayed reaggregation without inducer. Grunz, H and Tacke, L. Cell Diff. Dev. 1989; 28: 211–218Crossref | PubMed | Scopus (157)See all References, 2xSingle cell analysis of mesoderm formation in the Xenopus embryo. Godsave, S.F and Slack, J.M.W. Development. 1991; 111: 523–530PubMedSee all References), and the addition of BMP4 will inhibit this effect and induce epidermis (Wilson and Hemmati-Brivanlou 1995xInduction of epidermis and inhibition of neural fate by Bmp-4. Wilson, P.A and Hemmati-Brivanlou, A. Nature. 1995; 376: 331–333Crossref | PubMedSee all ReferencesWilson and Hemmati-Brivanlou 1995). The second evidence comes from experiments recently performed in Drosophila. Holley et al. 1996xThe Xenopus dorsalizing factor noggin ventralizes Drosophila embryos by preventing DPP from activating its receptor. Holley, S, Neul, J, Attisano, L, Wrana, J, Sasai, Y, O'Connor, M, De Robertis, E, and Ferguson, E. Cell. 1996; 86: 607–617Abstract | Full Text | Full Text PDF | PubMed | Scopus (180)See all ReferencesHolley et al. 1996 demonstrate that at least for SOG/chd, and perhaps for noggin, binding DPP/BMP is their only function. They showed that while noggin inhibits DPP and phenocopies a dpp− mutation, it can only operate outside of the cell, and in the presence of an activated DPP receptor, its effect is abolished. More compelling is the fact that the double sog− dpp− mutant has the same phenotype as the dpp− mutant. If SOG had any other function than just inhibiting DPP, the double mutant phenotype should have been different than that of dpp− alone. The final line of evidence comes from the fact that, while chordin can reverse the osteogenic induction caused by BMP4 in 10T1/2 cell lines, it cannot block the one mediated by retinoic acid (Piccolo et al. 1996xDorsoventral patterning in Xenopus (inhibition of ventral signals by direct binding of chordin to BMP-4) . Piccolo, S, Sasai, Y, Lu, B, and De Robertis, E. Cell. 1996; 86: 589–598Abstract | Full Text | Full Text PDF | PubMed | Scopus (797)See all ReferencesPiccolo et al. 1996). Taken together, these observations strongly suggest that inhibition of BMP signaling is sufficient for neuralization.Contribution from many groups working with the amphibian system has provided a molecular solution to the problem of vertebrate neural induction originally defined by Spemann and Mangold. The challenge for the future will inevitably include the establishment of a link between the early neural specification process, described above, and the function of neurogenic genes operating downstream of these signaling events, ultimately leading to the generation of a mature neuron.

In Xenopus embryos growth cones of the mandibular trigeminal nerve contact cells located in the posterior end of the cement gland. We report three lines of evidence suggesting that cues from the posterior cells of the target influence the behavior of trigeminal neurites. First, target ablation in vivo results in failure of axons to stop and to arborize. Second, 180 degree rotation in the anteroposterior axis of the target results in anterior innervation of the cement gland. Third, ectopic cement gland implantation in the path of neurites migration leads to ectopic innervation. These results provide evidence for short-range cues (50 microns) in the posterior part of the cement gland controlling the stopping and the branching of the mandibular trigeminal nerve. Furthermore, we show that an ectodermic explant expressing follistatin mimics the natural target in homing trigeminal neurites. We finally propose to use this novel in vivo assay to isolate the trigeminal nerve target-recognition molecule(s).

Activins are members of the transforming growth factor-beta superfamily, a class of peptide growth factors that can regulate the growth and differentiation of a variety of cell types. In mesoderm induction assays, activins A and B were shown to be very potent inducers and it was only recently demonstrated that they are crucial for initial mesoderm induction in Xenopus embryos. To determine the source of activin protein for initial mesoderm induction and to investigate whether activins may play further roles in embryonic development we have examined the localization of the mRNAs encoding the activin beta A and beta B subunits in Xenopus embryos. Activin beta A and beta B mRNAs are found in the follicle cells surrounding oocytes but not in oocytes themselves or fertilized eggs. During embryogenesis activin mRNA is first detected after the midblastula transition and expression increases as development proceeds. Activin beta B mRNA is homogeneously distributed during blastula and early gastrula stages but restricted to the dorso-anterior region in neurula stage embryos. At the early tailbud stage activin expression becomes confined to the brain, eye analgen, visceral pouches, otic vesicles, and the anterior notochord.

We have isolated and characterized cDNAs that encode a protein expressed in the axons and growth cones of a subset of Xenopus embryonic neurons. The protein is also expressed in a subset of cells of the brain, including cells in even-numbered rhombomeres, the eye, and the heart. The sequence of the cDNA suggests that the protein belongs to a new class of neural-specific intermediate filaments. Both the RNA and the protein are expressed in the neurula and persist during embryogenesis in the brain, cranial nerves, and spinal cord. Because of the predicted structure of the protein, we have named it tanabin (from the Persian word for rope).