Tissue specification in the early embryo requires the integration of spatial information at the promoters of developmentally important genes. Although several response elements for signalling pathways have been identified in Xenopus promoters, it is not yet understood what defines the sharp borders that restrict expression to a specific tissue. Here we use transgenic frog embryos to study the spatial and temporal regulation of the Xbra promoter. Deletion analysis and point mutations in putative transcription factor-binding sites identified two repressor modules, which exert their main effects at different stages during gastrulation. One module is defined by a bipartite binding site for a Smad-interacting protein (SIP1) of the deltaEF1 repressor family and acts to confine expression to the marginal zone early in gastrulation. The other module is defined by two homeodomain-binding sites and is responsible for repression in dorsal mesoderm and ectoderm at mid-gastrula stages. In addition, an upstream region of the promoter is necessary to repress expression in neural tissues later in development. Together, our results show that repression plays an important role in the restriction of Xbra expression to the mesoderm, and we suggest that similar mechanisms may be involved in the spatial regulation of other genes in early embryonic development.

One of the earliest markers of anterior asymmetry in vertebrate embryos is the transcription factor Hex. We find that Hex is a transcriptional repressor that can be converted to an activator by fusing full length Hex to two copies of the minimal transcriptional activation domain of VP16 together with the flexible hinge region of the (lambda) repressor (Hex-(lambda)VP2). Retention of the entire Hex open reading frame allows one to examine Hex function without disrupting potential protein-protein interactions. Expression of Hex-(lambda)VP2 in Xenopus inhibits expression of the anterior marker Cerberus and results in anterior truncations. Such embryos have multiple notochords and disorganised muscle tissue. These effects can occur in a cell non-autonomous manner, suggesting that one role of wild-type Hex is to specify anterior structures by suppressing signals that promote dorsal mesoderm formation. In support of this idea, over-expression of wild-type Hex causes cell non-autonomous dorso-anteriorization, as well as cell autonomous suppression of dorsal mesoderm. Suppression of dorsal mesoderm by Hex is accompanied by the down-regulation of Goosecoid and Chordin, while induction of dorsal mesoderm by Hex-(lambda)VP2 results in activation of these genes. Transient transfection experiments in ES cells suggest that Goosecoid is a direct target of Hex. Together, our results support a model in which Hex suppresses organiser activity and defines anterior identity.

Gastrulation in the amphibian embryo is driven by cells of the mesoderm. One of the genes that confers mesodermal identity in Xenopus is Brachyury (Xbra), which is required for normal gastrulation movements and ultimately for posterior mesoderm and notochord differentiation in the development of all vertebrates. Xbra is a transcription activator, and interference with transcription activation leads to an inhibition of morphogenetic movements during gastrulation. To understand this process, we have screened for downstream target genes of Brachyury (Tada, M., Casey, E., Fairclough, L. and Smith, J. C. (1998) Development 125, 3997–4006). This approach has now allowed us to isolate Xwnt11, whose expression pattern is almost identical to that of Xbra at gastrula and early neurula stages. Activation of Xwnt11 is induced in an immediate-early fashion by Xbra and its expression in vivo is abolished by a dominant-interfering form of Xbra, Xbra-En(R). Overexpression of a dominant-negative form of Xwnt11, like overexpression of Xbra-En(R), inhibits convergent extension movements. This inhibition can be rescued by Dsh, a component of the Wnt signalling pathway and also by a truncated form of Dsh which cannot signal through the canonical Wnt pathway involving GSK-3 and (beta)-catenin. Together, our results suggest that the regulation of morphogenetic movements by Xwnt11 occurs through a pathway similar to that involved in planar polarity signalling in Drosophila.

The maternal T-box gene VegT, whose transcripts are restricted to the vegetal hemisphere of the Xenopus embryo, plays an essential role in early development. Depletion of maternal VegT transcripts causes embryos to develop with no endoderm, while vegetal blastomeres lose the ability to induce mesoderm (Zhang, J., Houston, D. W., King, M. L., Payne, C., Wylie, C. and Heasman, J. (1998) Cell 94, 515–524). The targets of VegT, a transcription activator, must therefore include genes involved both in the specification of endoderm and in the production of mesoderm-inducing signals. We recently reported that the upstream regulatory region of the homeobox-containing gene Bix4 contains T-box binding sites. Here we show that expression of Bix4 requires maternal VegT and that two T-box binding sites are necessary and sufficient for mesodermal and endodermal expression of reporter genes driven by the Bix4 promoter in transgenic Xenopus embryos. Remarkably, a single T-box binding site is able to act as a mesoderm-specific enhancer when placed upstream of a minimal promoter. Finally, we show that Bix4 rescues the formation of endodermal markers in embryos in which VegT transcripts have been ablated but does not restore the ability of vegetal pole blastomeres to induce mesoderm. These results demonstrate that Bix4 acts directly downstream of VegT to specify endodermal differentiation in Xenopus embryos.

The endoderm forms the gut and associated organs, and develops from a layer of cells which emerges during gastrula stages in the vertebrate embryo. In comparison to mesoderm and ectoderm, little is known about the signals which induce the endoderm. The origin of the endoderm is intimately linked with that of mesoderm, both by their position in the embryo, and by the molecules that can induce them. We characterised a gene, zebrafish gata5, which is expressed in the endoderm from blastula stages and show that its transcription is induced by signals originating from the yolk cell. These signals also induce the mesoderm-expressed transcription factor no tail (ntl), whose initial expression coincides with gata5 in the cells closest to the blastoderm margin, then spreads to encompass the germ ring. We have characterised the induction of these genes and show that ectopic expression of activin induces gata5 and ntl in a pattern which mimics the endogenous expression, while expression of a dominant negative activin receptor abolishes ntl and gata5 expression. Injection of RNA encoding a constitutively active activin receptor leads to ectopic expression of gata5 and ntl. gata5 is activated cell-autonomously, whereas ntl is induced in cells distant from those which have received the RNA, showing that although expression of both genes is induced by a TGF-beta signal, expression of ntl then spreads by a relay mechanism. Expression of a fibroblast growth factor (eFGF) or a dominant negatively acting FGF receptor shows that ntl but not gata5 is regulated by FGF signalling, implying that this may be the relay signal leading to the spread of ntl expression. In embryos lacking both squint and cyclops, members of the nodal group of TGF-beta related molecules, gata5 expression in the blastoderm is abolished, making these factors primary candidates for the endogenous TGF-beta signal inducing gata5.

The Xenopus homologue of Brachyury, Xbra, is expressed in the presumptive mesoderm of the early gastrula. Induction of Xbra in animal pole tissue by activin occurs only in a narrow window of activin concentrations; if the level of inducer is too high, or too low, the gene is not expressed. Previously, we have suggested that the suppression of Xbra by high concentrations of activin is due to the action of genes such as goosecoid and Mix.1. Here, we examine the roles played by goosecoid and Mix.1 during normal development, first in the control of Xbra expression and then in the formation of the mesendoderm. Consistent with the model outlined above, inhibition of the function of either gene product leads to transient ectopic expression of Xbra. Such embryos later develop dorsoanterior defects and, in the case of interference with Mix.1, additional defects in heart and gut formation. Goosecoid, a transcriptional repressor, appears to act directly on transcription of Xbra. In contrast, Mix.1, which functions as a transcriptional activator, may act on Xbra indirectly, in part through activation of goosecoid.

Brachyury, a member of the T-box gene family, is required for posterior mesoderm and notochord differentiation in vertebrate development, and mis-expression of Xenopus Brachyury causes ectopic mesoderm formation. Brachyury is a transcription activator, and its ability to activate transcription is essential for its biological function, but Brachyury target genes have proved difficult to identify. Here we employ a hormone-inducible Brachyury construct and subtractive hybridization to search for such targets. Using this approach we have isolated Bix1, a homeobox gene expressed both in the marginal zone of Xenopus and in the vegetal hemisphere. Expression of Bix1 is induced in an immediate-early fashion by mesoderm-inducing factors such as activin as well as by the products of the T-box genes Xbra and VegT (also known as Antipodean, Brat and Xombi). Activation of Bix1 in response to Xbra is direct in the sense that it does not require protein synthesis, and both Xbra and VegT activate expression of a reporter gene driven by the Bix 5′ regulatory region, which contains an Xbra/VegT binding site. Mis-expression of low levels of Bix1 causes formation of ventral mesoderm, while high levels induce endodermal differentiation. These results suggest that Bix1 acts downstream of both VegT and Xbra to induce formation of mesoderm and endoderm.

Brachyury is a member of the T-box gene family and is required for formation of posterior mesoderm and notochord during vertebrate development. The ability of Brachyury to activate transcription is essential for its biological function, but nothing is known about its target genes. Here we demonstrate that Xenopus Brachyury directly regulates expression of eFGF by binding to an element positioned approximately 1 kb upstream of the eFGF transcription start site. This site comprises half of the palindromic sequence previously identified by binding site selection and is also present in the promoters of the human and mouse homologues of eFGF.

The TGFbeta family member activin induces different mesodermal cell types in a dose-dependent fashion in the Xenopus animal cap assay. High concentrations of activin induce dorsal and anterior cell types such as notochord and muscle, while low concentrations induce ventral and posterior tissues such as mesenchyme and mesothelium. In this paper we investigate whether this threshold phenomenon involves the differential effects of the two type I activin receptors ALK-2 and ALK-4. Injection of RNA encoding constitutively active forms of the receptors (here designated ALK-2* and ALK-4*) reveals that ALK-4* strongly induces the more posterior mesodermal marker Xbra and the dorsoanterior marker goosecoid in animal cap explants. Maximal levels of Xbra expression are attained using lower concentrations of RNA than are required for the strongest activation of goosecoid, and at the highest doses of ALK-4*, levels of Xbra transcription decrease, as is seen with high concentrations of activin. By contrast, the ALK-2* receptor activates Xbra but fails to induce goosecoid to significant levels. Analysis at later stages reveals that ALK-4* signalling induces the formation of a variety of mesodermal derivatives, including dorsal cell types, in a dose-dependent fashion, and that high levels also induce endoderm. By contrast, the ALK-2* receptor induces only ventral mesodermal markers. Consistent with these observations, ALK-4* is capable of inducing a secondary axis when injected into the ventral side of 32-cell stage embryos whilst ALK-2* cannot. Co-injection of RNAs encoding constitutively active forms of both receptors reveals that ventralising signals from ALK-2* antagonise the dorsal mesoderm-inducing signal derived from ALK-4*, suggesting that the two receptors use distinct and interfering signalling pathways. Together, these results show that although ALK-2* and ALK-4* transduce distinct signals, the threshold responses characteristic of activin cannot be due to interactions between these two pathways; rather, thresholds can be established by ALK-4* alone. Furthermore, the effects of ALK-2* signalling are at odds with it behaving as an activin receptor in the early Xenopus embryo.

Analysis of gene function in Xenopus development frequently involves over-expression experiments, in which RNA encoding the protein of interest is microinjected into the early embryo. By taking advantage of the fate map of Xenopus, it is possible to direct expression of the protein to particular regions of the embryo, but it has not been possible to exert control over the timing of expression; the protein is translated immediately after injection. To overcome this problem in our analysis of the role of Brachyury in Xenopus development, we have, like Kolm and Sive (1995; Dev. Biol. 171, 267–272), explored the use of hormone-inducible constructs. Animal pole regions derived from embryos expressing a fusion protein (Xbra-GR) in which the Xbra open reading frame is fused to the ligand-binding domain of the human glucocorticoid receptor develop as atypical epidermis, presumably because Xbra is sequestered by the heat-shock apparatus of the cell. Addition of dexamethasone, which binds to the glucocorticoid receptor and releases Xbra, causes formation of mesoderm. We have used this approach to investigate the competence of animal pole explants to respond to Xbra-GR, and have found that competence persists until late gastrula stages, even though by this time animal caps have lost the ability to respond to mesoderm-inducing factors such as activin and FGF. In a second series of experiments, we demonstrate that Xbra is capable of inducing its own expression, but that this auto-induction requires intercellular signals and FGF signalling. Finally, we suggest that the use of inducible constructs may assist in the search for target genes of Brachyury.

Bone morphogenetic protein-4 (BMP-4) is thought to play an important role in early Xenopus development by acting as a ‘ventralizing factor’ and as an epidermal determinant: local inhibition of BMP-4 function in whole embryos causes the formation of an additional dorsal axis, and inhibition of BMP-4 function in isolated ectodermal cells causes the formation of neural tissue. In this paper we describe a homeobox-containing gene whose expression pattern is similar to that of BMP-4, whose expression requires BMP-4 signalling and which, when over-expressed, causes a phenotype similar to that caused by over-expression of BMP-4. We suggest that this gene, which we call Xom, acts downstream of BMP-4 to mediate its effects.

The Brachyury (T) gene is required for formation of posterior mesoderm and for axial development in both mouse and zebrafish embryos. In this paper, we first show that the Xenopus homologue of Brachyury, Xbra, and the zebrafish homologue, no tail (ntl), both function as transcription activators. The activation domains of both proteins map to their carboxy terminal regions, and we note that the activation domain is absent in two zebrafish Brachyury mutations, suggesting that it is required for gene function. A dominant-interfering Xbra construct was generated by replacing the activation domain of Xbra with the repressor domain of the Drosophila engrailed protein. Microinjection of RNA encoding this fusion protein allowed us to generate Xenopus and zebrafish embryos which show striking similarities to genetic mutants in mouse and fish. These results indicate that the function of Brachyury during vertebrate gastrulation is to activate transcription of mesoderm-specific genes. Additional experiments show that Xbra transcription activation is required for regulation of Xbra itself in dorsal, but not ventral, mesoderm. The approach described in this paper, in which the DNA-binding domain of a transcription activator is fused to the engrailed repressor domain, should assist in the analysis of other Xenopus and zebrafish transcription factors.

Injection of RNA encoding BMP-4 into the early Xenopus embryo suppresses formation of dorsal and anterior cell types. To understand this phenomenon, it is necessary to know the stage at which BMP-4 acts. In this paper, we present three lines of evidence showing that BMP-4 misexpression has no effect on the initial steps of mesoderm induction, either dorsal or ventral, but instead causes ventralization during gastrulation. Firstly, activation of organizer-specific genes such as goosecoid, Xnot, pintallavis and noggin occurs normally in embryos injected with BMP-4 RNA, but transcript levels are then rapidly down-regulated as gastrulation proceeds. Similarly, BMP-4 does not affect the initial activation of goosecoid by activin in animal caps, but expression then declines precipitously. Secondly, embryos made ventral by injection with BMP-4 RNA cannot be rescued by grafts of Spemann's organizer at gastrula stages. Such embryos therefore differ from those made ventral by UV-irradiation, where the defect occurs early and rescue can be effected by the organizer. Finally, the dorsalizing effects of the organizer, and of the candidate dorsalizing signal noggin, both of which exert their effects during gastrulation, can be counteracted by BMP-4. Together, these experiments demonstrate that BMP-4 can act during gastrulation both to promote ventral mesoderm differentiation and to attenuate dorsalizing signals derived from the organizer.

Mouse embryos homozygous for a null mutation in nodal arrest development at early gastrulation and contain little or no embryonic mesoderm. Here, two Xenopus nodal-related genes (Xnr-1 and Xnr-2) are identified and shown to be expressed transiently during embryogenesis, first within the vegetal region of late blastulae and later in the marginal zone during gastrulation, with enrichment in the dorsal lip. Xnrs and mouse nodal function as dose-dependent dorsoanterior and ventral mesoderm inducers in whole embryos and explanted animal caps. Using a plasmid vector to produce Xnr proteins during gastrulation, we show that, in contrast to activin and other TGF beta-like molecules, Xnr-1 and Xnr-2 can dorsalize ventral marginal zone explants and induce muscle differentiation. Xnr signalling also rescues a complete embryonic axis in UV-ventralized embryos. The patterns of Xnr expression, the activities of the proteins and the phenotype of mouse nodal mutants, all argue strongly that a signaling pathway involving nodal, or nodal-related peptides, is an essential conserved element in mesoderm differentiation associated with vertebrate gastrulation and axial patterning.

Widespread expression of the DNA-binding protein Brachyury in Xenopus animal caps causes ectopic mesoderm formation. In this paper, we first show that two types of mesoderm are induced by different concentrations of Brachyury. Animal pole explants from embryos injected with low doses of Xbra RNA differentiate into vesicles containing mesothelial smooth muscle and mesenchyme. At higher concentrations somitic muscle is formed. The transition from smooth muscle formation to that of somitic muscle occurs over a two-fold increase in Brachyury concentration. Brachyury is required for differentiation of notochord in mouse and fish embryos, but even the highest concentrations of Brachyury do not induce this tissue in Xenopus animal caps. Co-expression of Brachyury with the secreted glycoprotein noggin does cause notochord formation, but it is difficult to understand the molecular basis of this phenomenon without knowing more about the noggin signal transduction pathway. To overcome this difficulty, we have now tested mesoderm-specific transcription factors for the ability to synergize with Brachyury. We find that co-expression of Pintallavis, but not goosecoid, with Brachyury causes formation of dorsal mesoderm, including notochord. Furthermore, the effect of Pintallavis, like that of Brachyury, is dose-dependent: a two-fold increase in Pintallavis RNA causes a transition from ventral mesoderm formation to that of muscle, and a further two-fold increase induces notochord and neural tissue. These results suggest that Pintallavis cooperates with Brachyury to pattern the mesoderm in Xenopus.

To increase our understanding of haematopoiesis during early vertebrate development, we have studied the expression pattern of the transcription factor GATA-2 in Xenopus embryos, and asked how this is regulated. We show that the blood island precursors of the ventral mesoderm express GATA-2 RNA at neural tube stages, some 5 hours before globin RNA is detected in their derivatives. Prior to this however, GATA-2 is expressed much more widely within the embryo. Maternal transcripts are uniformly distributed, and zygotic transcription is activated during gastrulation throughout ventral and lateral regions of the embryo, with expression highest in the sensorial ectoderm and only weak in the ventral mesoderm. The domain of GATA-2 expression in neurulae outlines the region of the neural plate and suggests a possible wider role in dorsoventral patterning. To identify the signals involved in regulating this pattern of expression, we performed experiments with embryo explants. GATA-2 is activated autonomously in isolated animal caps and this activation is suppressed by the mesoderm-inducing factor activin, but not by FGF. Thus, the down-regulation of GATA-2 observed in the region of the Spemann organiser may be a response to an activin-like signal emanating from the dorsal-vegetal region or Nieuwkoop centre. GATA-2 activation in animal caps and ventral marginal zones was suppressed by co-culturing with dorsal marginal zones, suggesting that a signal from the Spemann organiser is involved in suppression of GATA-2 in the dorsal region of the embryo. Expression of a candidate for this signal, noggin, had the same effect. Taken together, the observations presented here suggest that GATA-2 activation occurs by default in the absence of signals, that the restriction of its expression within the early embryo is controlled by negative signals emanating from the Nieuwkoop centre and the organiser, and that noggin and activin-like molecules play a role in these signalling pathways.

The growth factor activin elicits mesodermal fates when applied to prospective ectodermal cells of the Xenopus blastula stage embryo. Previous experiments with dissociated cells showed that there are at least five different responses separated by closely spaced, sharp dose thresholds. Here we investigate this multithreshold activin response further using probes for genes expressed at early gastrula stages, namely Xbra, goosecoid, noggin, Xwnt-8 and Mix.1. We show that initial dose-response profiles are broad and smooth in contrast to the later threshold-bound patterns. For Xbra, goosecoid and noggin, the later expression ranges are subsets of earlier ones. Unexpectedly, Xwnt-8 is initially induced at high doses only, but later appears only in cells that have received a low dose of activin. Keeping the cells dissociated after activin treatment, rather than allowing them to reaggregate, prevents sustained expression of Xbra and Xwnt-8 but allows that of goosecoid and noggin. However, cell contact is required for sharpening the dose-response threshold of goosecoid. Finally, we show that a previously reported dorsoventral prepattern in the animal cap is also cell-contact dependent and it is not required for the multi-threshold response to activin.

Members of the activin family are believed to act as mesoderm-inducing factors during early amphibian development. Little is known, however, about mesoderm formation in the mammalian embryo, and as one approach to investigating this we have studied activin and follistatin expression during early mouse development. Activins are homo- or heterodimers of the beta A or beta B subunits of inhibin, itself a heterodimer consisting of one of the beta subunits together with an alpha subunit. Follistatin is a single-chain polypeptide which inhibits activin function. Expression of the inhibin alpha chain could not be detected in embryonic or extraembryonic tissues at any of the stages studied (5.5 to 8.5 days) and expression of the beta A and beta B subunits could only be observed in the deciduum in cells surrounding the embryo. Expression of follistatin could also be detected in the deciduum, but in a pattern complementary to that of the beta subunits. Embryonic expression of follistatin first occurred in the primitive streak, and at later stages transcripts were detectable in the somites and in rhombomeres 2, 4 and 6 of the hindbrain. These results are consistent with a role for activin in mesoderm formation in the mouse embryo, and suggest functions for follistatin in addition to its role as an inhibitor of activin.

Members of the activin family have been suggested to act as mesoderm-inducing factors during early amphibian development. Little is known, however, about mesoderm formation in the mammalian embryo, and as one approach to investigating this we have studied activin expression during early mouse development. Activins are homo- or heterodimers of the beta A or beta B subunits of inhibin, itself a heterodimer consisting of one of the beta subunits together with an alpha subunit. Our results indicate that the oocyte contains mRNA encoding all three subunits, and antibody staining demonstrates the presence of both alpha and beta protein chains. From the fertilized egg stage onwards, alpha subunit protein cannot be detected, so the presence of beta subunits reflects the presence of activin rather than inhibin. Maternal levels of activin protein decline during early cleavage stages but increase, presumably due to zygotic transcription (see below), in the compacted morula. By 3.5 days, only the inner cell mass (ICM) cells of the blastocyst express activin, but at 4.5 days the situation is reversed; activin expression is confined to the trophectoderm. Using reverse transcription-PCR, neither beta A nor beta B mRNA was detectable at the two-cell stage but transcripts encoding both subunits were detectable at the morula stage, with beta B mRNA persisting into the blastocyst. We have also analyzed activin and inhibin expression in ES and EC cells. Consistent with the observation that activins are expressed in the ICM of 3.5-day blastocysts, we find high levels of beta A and beta B mRNA in all eight ES cell lines tested. F9 EC cells express only activin beta B, together with low levels of the inhibin alpha chain. When ES and EC cells are induced to differentiate, levels of activin fall dramatically. These results are consistent with a role for activins in mesoderm formation and other steps of early mouse development.

In Xenopus embryos, the extracellular matrix (ECM) protein tenascin (TN) is expressed dorsally in a very restricted pattern. We have studied the spatial and temporal expression of TN mRNA in tailbud-stage embryos by RNAase protection and in situ hybridization using a cDNA probe for Xenopus TN obtained by PCR amplification. We report that TN transcripts are principally expressed in cells dispersed around the neural tube and notochord as well as in myotome and sclerotome cells. No TN mRNA could be detected in lateral plate mesoderm, but expression was detectable beneath tail fin epidermis. In a second series of experiments, we studied the expression of TN mRNA and protein in combinations between animal and vegetal stage-6 blastomeres and in stage-8 blastula animal caps treated with activin A or basic fibroblastic growth factor (b-FGF). Isolated animal cap tissue cultured alone differentiates into epidermis, which expresses neither TN protein nor TN mRNA. TN expression is, however, elicited in response to isolated dorsal vegetal blastomeres and in response to high concentrations of activin, both of which treatments lead to formation of muscle and/or notochord. Low concentrations of activin, and ventral vegetal blastomeres, treatments that induce mesoderm of ventral character, are poor inducers of TN. However, b-FGF, which also induces ventral mesoderm, elicits strong expression. These results indicate that TN regionalization is a complex process, dependent both on the pattern of differentiation of mesodermal tissues and on the agent with which they are induced. The data further show that “ventral mesoderm” induced by low concentrations of activin is distinct from that induced by b-FGF, and imply that activin induces ventral mesoderm of the trunk while b-FGF induces posterior mesoderm of the tailbud.