Table of contents

DPP homologs and dorsoventral patterning: The roles of Nodal, Nodal related, Lefty, Activin and Vg1, divergent members of the TGF-beta family

In Xenopus, mesoderm induction by endoderm at the blastula stage is well documented, but the molecular nature of the endogenous inductive signals remains unknown. The carboxy-terminal fragment of Cerberus, designated Cer-S, provides a specific secreted antagonist of mesoderm-inducing Xenopus Nodal-Related (Xnr) factors. Cer-S does not inhibit signaling by other mesoderm inducers such as Activin, Derriere, Vg1 and BMP4, nor by the neural inducer Xnr3. Cer-S blocks the induction of both dorsal and ventral mesoderm in animal-vegetal Nieuwkoop-type recombinants. During blastula stages Xnr1, Xnr2 and Xnr4 are expressed in a dorsal to ventral gradient in endodermal cells. Dose-response experiments using cer-S mRNA injections support the existence of an endogenous activity gradient of Xnrs. Xnr expression at blastula can be activated by the vegetal determinants VegT and Vg1 acting in synergy with dorsal beta-catenin. The data support a modified model for mesoderm induction in Xenopus, in which mesoderm induction is mediated by a gradient of multiple Nodal-related signals released by endoderm at the blastula stage (Agius, 2000).

The pioneering work of Nieuwkoop, Slack and colleagues has led to the current three-signal model of mesoderm induction and patterning. Using recombinants of blastula endodermal and ectodermal explants, it has been shown that mesoderm is generated via inductive signals from endoderm. The endoderm is thought to release two signals: (1) a uniform or ventral endodermal signal that induces ventral mesoderm such as lateral plate, mesenchyme and blood and (2) a signal emanating from dorsal endoderm that induces dorsal organizer tissue in the overlying mesoderm. A third signal in this model, also called the horizontal signal, emanates from dorsal organizer tissue during gastrulation and is able to induce the differentiation of dorsal histotypes such as notochord, somites and kidney in ventral mesodermal cells. A number of molecules secreted by Spemann's organizer are thought to participate in this third signal (Agius, 2000).

Maternal activities such as dorsal beta-catenin and vegetal VegT and Vg1 cooperate to set up a zygotic dorsal to ventral gradient in the endoderm composed of multiple Xnrs at stage 9, when mesoderm induction takes place. At high Nodal-related concentrations, which require a functional beta-catenin pathway in the dorsal side of the embryo, the Spemann organizer (expressing genes such as chordin, noggin and Frzb1) is induced in overlying cells by early gastrula. In the ventral side, VegT and Vg1 would lead to the production of lower levels of Nodal-related signals, and ventral mesoderm (expressing genes such as Xwnt8 and BMP4) would be induced. Similarly, in embryos ventralized by UV irradiation or by deltaN-XTcf-3, the uniformly distributed VegT and Vg1 products would produce low levels of Xnrs sufficient to induce ventral mesoderm at the gastrula stage (Agius, 2000 and references therein).

A particularly attractive aspect of the model is that it may help explain a long-standing puzzle in Xenopus embryology. A surprisingly large number of microinjected molecules are able to rescue, often completely, the UV ventralized phenotype that results from interfering with cortical rotation of the fertilized egg. The UV-rescuing gene products include such diverse molecules as beta-catenin (and other members of this signaling pathway), Vg1, Xnr1 and Xnr2, noggin and chordin. Although one can argue that each of these diverse genes acts via different redundant pathways, their common UV-rescue activity may be easier to unravel if considered as part of a cascade of sequential gene activations. In this view, overexpression of beta-catenin or Vg1 would lead to high levels of Xnr expression in blastula endoderm, which in turn would mediate the induction of Spemann organizer in overlying cells, activating genes such as noggin and chordin that execute dorsal patterning at the gastrula stage (Agius, 2000 and references therein).

This sequential model of gene activation must be considered an oversimplification of the in vivo situation. It is likely that multiple signaling pathways synergize to pattern the gastrula. For example, this model does not take into account the role that beta-catenin and its target homeobox genes siamois and Xtwn may have in the marginal zone itself. It is known that the promoter of the homeobox gene goosecoid, which is active in dorsal mesoderm, contains Siamois and Xtwn binding sites in addition to a TGF-beta responsive element. However, in embryos in which the function of beta-catenin and expression of siamois are blocked by the inhibitor DeltaN-XTcf-3, overexpression of Xnr1 mRNA is sufficient to activate goosecoid and other organizer markers. Furthermore, in animal cap experiments Xnr1 protein is able to induce ventral and dorsal mesodermal markers with sharp thresholds in the low nanomolar range. Since Activin, Vg1 and Derriere are not inhibited by Cer-S, the present experiments do not address whether these molecules may cooperate with Xnrs in mesoderm patterning in vivo. There is ample genetic support for a critical role of Nodal-related molecules in mesoderm formation in many vertebrates. In mouse, mutations in the gene nodal result in embryos severely deficient in mesodermal tissues. It has been argued that, because some mutant embryos contain patches of Brachyury expression, mouse Nodal is involved in the maintenance rather than in the initiation of mesoderm induction. In Xenopus, the present results suggest a requirement for Nodal-related signaling in the initial mesoderm induction by endoderm. In zebrafish, two Nodal-related genes, cyclops and squint have been identified to date. Both mutations affect axial mesoderm and in double mutants the effects are synergistic, leading to the loss of goosecoid expression in the organizer and lack of head and trunk mesoderm (Agius, 2000 and references therein).

In vertebrates, Nodal-related protein plays crucial roles in mesoderm and endoderm induction. Two novel Xenopus nodal-related genes, Xnr5 and Xnr6 are described; they are first zygotically expressed at the mid-blastula transition, in the dorsal-vegetal region including the Nieuwkoop center. Xnr5 and Xnr6 were isolated by expression screening of a library enriched with immediate-early-type transcripts, and are strong inducers of both mesoderm and endoderm. They also induce the other nodal-related genes in the animal cap. In embryos, cerberus-short (nodal-specific inhibitor) can inhibit Xnr1 and Xnr2 expression to the same extent as goosecoid, but not Xnr5 and Xnr6 transcription. Xnr5 and Xnr6 are regulated completely cell autonomously, differently from other Xnrs in the cell-dissociated embryos. The expression of Xnr5 and Xnr6 is regulated by maternal Veg T and beta-catenin, but does not require TGF-beta signaling. Therefore, expression of Xnr5 and Xnr6 is controlled by different mechanisms from other Xnr family genes (Takashashi, 2000).

A model of endoderm and mesoderm induction by known nodal-related genes in Xenopus is presented. At the MBT, when zygotic transcription starts, vegetally localized maternal Veg T induces Xnr5 and Xnr6 in cooperation with beta-catenin to make the Nieuwkoop center. However, Xnr5 and Xnr6 may also be regulated by other pathways. Quick accumulation of Xnr5 and Xnr6 extends throughout the deep endoderm. In this phase, only low levels of Xnr1, Xnr2 and Xnr4 expression are induced cell autonomously. Then endodermal expression of Xnr5 and Xnr6 leads to accumulation of Xnr1, Xnr2 and Xnr4 in the endodermal region, which in turn act in endoderm determination in cooperation with Veg T, and induce the expressions of Xnr1, Xnr2, Xnr4, gsc and Xbra, among other genes, in the equatorial region (mesoderm induction, including formation of organizer). beta-catenin directly regulates expression of siamois, Xtwin and Xnr3 in the dorsal side, and plays a role in determination of anterior endoderm and organizer (Takashashi, 2000).

The segregation of cells into germ layers is one of the earliest events in the establishment of cell fate in the embryo. In the zebrafish, endoderm and mesoderm are derived from cells that involute into an internal layer, the hypoblast, whereas ectoderm is derived from cells that remain in the outer layer, the epiblast. In this study, the origin of the zebrafish endoderm and its separation from the mesoderm are examined. By labeling individual cells located at the margin of the blastula, it has been demonstrated that all structures that are endodermal in origin are derived predominantly from the more dorsal and lateral cells of the blastoderm margin. Frequently marginal cells give rise to both endodermal and mesodermal derivatives, demonstrating that these two lineages have not yet separated. Cells located farther than 4 cell diameters from the margin give rise exclusively to mesoderm, and not to endoderm. Following involution, a variety of cellular changes are seen indicating the differentiation of the two germ layers. Endodermal cells gradually flatten and extend filopodial processes forming a noncontiguous inner layer of cells against the yolk. At this time, they also begin to express Forkhead-domain 2 protein, the zebrafish homolog of HNF3beta. Mesodermal cells form a coherent layer of round cells separating the endoderm and ectoderm. In cyclops-mutant embryos (cyclops codes for a nodal-related signaling molecule), which have reduced mesodermal anlage, it has been demonstrated that by late gastrulation not only mesodermal but also endodermal cells are fewer in number. This suggests that a common pathway initially specifies two germ layers together before a progressive sequence of determinative events segregates endoderm and mesoderm into morphologically distinct germ layers. A hierarchical sequence of determinative events is hypothesized: it is suggested that a cell is first specified to either the epiblast or hypoblast and, subsequently, within the hypoblast, a specification to either the endoderm or mesoderm. Before this latter separation, cells appear to be part of a uniform population of migrating cells, and no distinguishing characteristics between the endodermal and mesodermal precursors can be seen (Warga, 1999).

Shortly after implantation the mouse embryo comprises three tissue layers. The founder tissue of the embryo proper, the epiblast, forms a radially symmetric cup of epithelial cells that grows in close apposition to the extra-embryonic ectoderm and the visceral endoderm. This simple cylindrical structure exhibits a distinct molecular pattern along its proximal-distal axis. The anterior-posterior axis of the embryo is positioned later by coordinated cell movements that rotate the pre-existing proximal-distal axis. The transforming growth factor-ß family member Nodal is known to be required for formation of the anterior-posterior axis. This study shows that signals from the epiblast are responsible for the initiation of proximal-distal polarity. Nodal acts to promote posterior cell fates in the epiblast and to maintain molecular pattern in the adjacent extra-embryonic ectoderm. Both of these functions are independent of Smad2. Moreover, Nodal signals from the epiblast also pattern the visceral endoderm by activating the Smad2-dependent pathway required for specification of anterior identity in overlying epiblast cells. These experiments show that proximal-distal and subsequent anterior-posterior polarity of the pregastrulation embryo result from reciprocal cell-cell interactions between the epiblast and the two extra-embryonic tissues (Brennan, 2001).

In mouse, frog, and chicken embryos, Nodal-related factors have been implicated in mesodermal and neural patterning, and left-right asymmetry. Zebrafish nodal-related 2 (znr2) has been isolated and characterized. znr2 is expressed at low levels maternally, and zygotic transcripts localize to dorsal blastomeres at MBT. Slightly later, znr2 is also expressed dorsally in the extraembryonic yolk syncytial layer (YSL). During early gastrulation, znr2 expression expands to include deep and superficial cells in the entire marginal zone and YSL. During shield stages, expression is primarily localized to superficial noninvoluting cells of the organizer, called dorsal forerunners. Znr2 misexpression in whole fish embryos expands or duplicates dorsoanterior and axial cell fates. Furthermore, Znr2 overexpression exclusively in the YSL, a region implicated in endogenous mesodermal induction, causes broadened or duplicated goosecoid expression in the overlying blastoderm. Functional comparison of Znr2 and another recently identified zebrafish nodal-related factor, Znr1/Cyclops, reveals distinct inductive properties of each ligand. Znr2 efficiently induces organizer-type dorsoanterior mesodermal and endodermal markers, but neural markers only weakly, if at all. In contrast, while Znr1/Cyclops reproducibly induces mesodermal and neural markers, it is an inefficient inducer of organizer-type mesoderm. These results suggest that znr2 encodes a robust mesendodermal inducer that signals nonautonomously during the earliest stages of embryonic patterning, and that part of this activity arises from within the YSL (Erter, 1998).

The vertebrate brain develops from a bilaterally symmetric neural tube but later displays profound anatomical and functional asymmetries. Despite considerable progress in deciphering mechanisms of visceral organ laterality, the genetic pathways regulating brain asymmetries are unknown. In zebrafish, genes implicated in laterality of the viscera (cyclops/nodal, antivin/lefty and pitx2) are coexpressed on the left side of the embryonic dorsal diencephalon, within a region corresponding to the presumptive epiphysis or pineal organ. Asymmetric gene expression in the brain requires an intact midline and Nodal-related factors. RNA-mediated rescue of mutants defective in Nodal signaling corrects tissue patterning at gastrulation, but fails to restore left-sided gene expression in the diencephalon. Such embryos develop into viable adults with seemingly normal brain morphology. However, the pineal organ, which typically emanates at a left-to-medial site from the dorsal diencephalic roof, becomes displaced in position. Thus, a conserved signaling pathway regulating visceral laterality also underlies an anatomical asymmetry of the zebrafish forebrain (Liang, 2000).

The Xenopus nodal related-1 (Xnr1) gene has a complex expression pattern in embryos, with two temporal phases. In the first phase, transcripts are first detected in perinuclear sites in the vegetal region of the blastula. During gastrulation, this expression disappears and transcripts become localized to the dorsal marginal zone. Expression stops and then restarts in a second phase at neurula and tailbud stages, first in two symmetric patches near the posterior end of the notochord, and then asymmetrically in a large domain in the left lateral plate mesoderm. The regulation of the early phase of expression of Xnr1 has been investigated. The T-box transcription factor VegT can induce Xnr1. It had previously been shown that Xnr1 can induce VegT in ectoderm cells and it is shown that the early expression of Xnr1 is regulated by an autoregulatory loop. By inspection of the Xnr1 promoter sequence, two non-palindromic T-box-binding sites, which are 10 bp apart, has been investigated. Using mutational analysis, these elements are shown to be required for the VegT induction of Xnr1. The Xnr1 promoter shows striking homologies with the Xnr3 promoter. In particular, two elements that are required for Wnt signaling are conserved between these two promoters, but the two T-box sites are not conserved, and Xnr3 is not induced by VegT. A region of the promoter containing the T-box sites and the Wnt sites is sufficient to drive expression of a reporter gene in a dorsal domain in transgenic Xenopus at the gastrula stage. This pattern of expression of the transgene in gastrulae is not dependent on the T-box sites (Hyde, 2000).

Different types of endoderm, including primitive, definitive and mesendoderm, play a role in the induction and patterning of the vertebrate head. These three types of endoderm are defined in order to compare the mechanism of head induction in model vertebrate organisms. (1) The primitive endoderm is a prospective extraembryonic tissue present in the mouse and the chick, whereas amphibia generate no extraembryonic tissue at all. This endoderm is not a product of gastrulation, and its fate is to become the stalk of the yolk sac. (2) The definitive anterior endoderm develops into the foregut and the liver. In amphibia, it also comprises yolky cells outside the epithelial lining. (3) The precordal mesendoderm as an organizer-derived tissue migrates anteriorly to lie under the developing forebrain. The formation of the anterior neural plate has been studied in chick embryos using the homeobox gene GANF as a marker. GANF is a member of the 'Anf' (anterior neural folds) family, from which a single member has been found in vertebrates, such as fish (Danf), amphibia (Xanf), chick (GANF), mice (Hesx1/Rpx) and human (HANF/HESX1), but has not been found in Drosophila. GANF is first expressed after mesendoderm ingression from Hensen's node. After transplantation, neither the avian hypoblast nor the anterior definitive endoderm is capable of GANF induction, whereas the mesendoderm (young head process, prechordal plate) exhibits a strong inductive potential. GANF induction cannot be separated from the formation of a proper neural plate, which requires an intact lower layer and the presence of the prechordal mesendoderm. It is inhibited by BMP4 and promoted by the presence of the BMP antagonist Noggin. In order to investigate the inductive potential of the mammalian visceral endoderm, use was made of rabbit embryos which, in contrast to mouse embryos, allow the morphological recognition of the prospective anterior pole in the living, pre-primitive-streak embryo. The anterior visceral endoderm from such rabbit embryos induces neuralization and independent, ectopic GANF expression domains in the area pellucida or the area opaca of chick hosts. In terms of the timing and the location of the head organizer, the chick is more similar to the frog, where the signaling comes from the organizer and its derivative, the mesendoderm. In contrast, head-inducing signals in mammals originate from the anterior visceral endoderm. Hence, mouse embryos begin the patterning of the head long before the mesendoderm ingresses, whereas chick head development occurs only after endoderm formation. Only mammals have shifted the head-inducing signals into the primitive endoderm, and they begin the induction and patterning process of the head long before (about 24 hours in mice and rabbits) the mesendoderm ingresses. Several genes are expressed in the independent primitive endoderm domain (the anterior visceral endoderm) in the mammalian head organizer before or at the onset of primitive streak formation, prior to their expression in the axial mesendoderm or the node during gastrulation. These include the homeobox genes Hesx1 (Rpx), Goosecoid, Lim1, Hex and Otx2; the forkhead gene HNF3beta; the nuclear protein gene mrg1; the growth factor gene Nodal, and the antigen VE-1. Thus, the signals for head induction reside in the anterior visceral endoderm of mammals whereas, in birds and amphibia, they reside in the prechordal mesendoderm, indicating a heterochronic shift of the head inductive capacity during the evolution of mammalia (Knoetgen, 1999).

Nodal, a member of the TGFbeta family of secreted growth factors is essential for gastrulation in the mouse. A nodallacZ reporter allele was generated by homologous recombination in embryonic stem cells. beta-galactosidase staining in the perigastrulation-stage embryo has demonstrated that the site of highest nodal expression is localized to the prospective posterior region of the epiblast, marking the site of primitive streak formation. Shortly after implantation (5.5 days of development), low levels of nodal are detected throughout the embryonic ectoderm and associated overlying primitive endoderm. Primitive streak forms at the site of the highest nodal activity within the ectoderm. Mesodermal cells exiting the streak proximally and laterally briefly express nodal. Transient nodal expression has been documented in the visceral endoderm prior to and during early streak formation. A mosaic analysis using wild-type embryonic stem cells to rescue nodal-deficient embryos allows the documentation of functionally distinct nodal activities in the embryonic ectodermal and primitive endodermal cell lineages. nodal signaling in the ectoderm is necessary for primitive streak formation as the gastrulation defect of nodal-deficient embryos can be rescued by the inclusion of small numbers of wild-type cells. In addition, chimeric embryos composed of nodal-deficient primitive endoderm fail to develop rostral neural structures. The forebrain structures, and possibly those of the anterior midbrain, fail to form in the absence of nodal-expressing primitive endoderm. In the vertebrate CNS, because the induction of the forebrain (and possibly regions of the midbrain) is contingent upon signals provided by the prechordal plate tissue, it seems likely that formation of this tissue is adversely affected in chimeras comprised largely of wild-type cells but with mutant primitive endoderm. It is concluded that the action of nodal, a TGFbeta-related growth factor expressed in the primitive endoderm, is critical for patterning of the anterior aspects of the A-P axis (Varlet, 1996).

The TGFbeta family member Xnr3 is similar in amino acid sequence to the mouse factor nodal and is expressed in a restricted group of cells in the superficial layer of Spemann's organizer. Xnr3, unlike the related factors nodal, Xnr1 and Xnr2, lacks mesoderm-inducing activity. Xnr3 can directly induce neural tissue in Xenopus ectoderm explants (animal caps). Injection of animal caps with either Xnr3 RNA or plasmids induces the expression of the pan-neural genes NCAM and nrp1, as well as the anterior neural marker Cpl1. A growing body of evidence suggests that neural induction in Xenopus proceeds as the default in the absence of epidermis inducers. The best candidates for the endogenous epidermis inducers are BMP-4 and BMP-7. The neural inducing activity of Xnr3 can be inhibited by overexpression of BMP-4, as has been observed with the neural inducers noggin, chordin and follistatin. Furthermore, Xnr3 can block mesoderm induction by BMP-4 and activin, but not by Xnr2. The structural basis underlying the divergent activities of Xnr2 and Xnr3 was analyzed using site-directed mutagenesis. Mutations introduced to the conserved cysteine residues characteristic of the TGFbeta family were found to inactivate Xnr2, but not Xnr3. The most unique feature of Xnr3 is the absence of a conserved cysteine at the C terminus of the protein. This feature distinguishes Xnr3 from other TGFbeta family members, including Xnr2. However, changing the C terminus of Xnr3 to more closely resemble other TGFbeta family members does not significantly alter its activity, suggesting that other structural features of Xnr3 distinguish its biological activity from Xnr2 (Hansen, 1997).

Vg1 is a transforming growth factor-beta-like molecule involved in mesoderm formation. Localization of mRNA has emerged as a fundamental mechanism for generating polarity during development. In vertebrates, one example of this phenomenon is Vg1 RNA, which is localized to the vegetal cortex of Xenopus oocytes. Vegetal localization of Vg1 RNA is directed by a 340-nt sequence element contained within its 3' untranslated region. To investigate how such cis-acting elements function in the localization process, a detailed analysis of the precise sequence requirements for vegetal localization within the 340-nt localization element was undertaken. There is considerable redundancy within the localization element and critical sequences lie at the ends of the element. When duplicated, a subelement from the 5' end of the Vg1 localization element has been found to be sufficient to direct vegetal localization. It is suggested that the Vg1 localization element is composed of smaller, redundant sequence motifs and one such 6-nt motif is identified as essential for localization. These results offer insight into what constitutes an RNA localization signal and how RNA sequence elements may act in the localization process (Gautreau, 1997).

TGF-beta signaling plays a key role in induction of the Xenopus mesoderm and endoderm. Using a yeast-based selection scheme, derriere, a novel TGF-beta family member has been isolated that is closely related to Vg1 and that is required for normal mesodermal patterning, particularly in posterior regions of the embryo. Unlike Vg1, derriere is expressed zygotically, with RNA localized to the future endoderm and mesoderm by late blastula, and to the posterior mesoderm by mid-gastrula. The derriere expression pattern appears to be identical to the zygotic expression domain of VegT (Xombi, Brat, Antipodean), and can be activated by VegT as well as fibroblast growth factor (FGF). In turn, derriere activates expression of itself, VegT and eFGF, suggesting that a regulatory loop exists between these genes. derriere is a potent mesoderm and endoderm inducer, acting in a dose-dependent fashion. When misexpressed ventrally, derriere induces a secondary axis lacking a head, an effect that is due to dorsalization of the ventral marginal zone. When misexpressed dorsally, derriere suppresses head formation. derriere can also posteriorize neurectoderm, but appears to do so indirectly. Together, these data suggest that derriere expression is compatible only with posterior fates. In order to assess the in vivo function of derriere, a dominant interfering Derriere protein (Cm-Derriere) was constructed that preferentially blocks Derriere activity relative to that of other TGFbeta family members. Cm-derriere expression in embryos leads to posterior truncation, including defects in blastopore lip formation, gastrulation and neural tube closure. Normal expression of anterior and hindbrain markers is observed; however, paraxial mesodermal gene expression is ablated. This phenotype can be rescued by wild-type derriere and by VegT. These findings indicate that derriere plays a crucial role in mesodermal patterning and development of posterior regions in Xenopus (Sun, 1999).

The Xenopus nodal-related 3 gene (Xnr3) is expressed in the Spemann organizer of the embryo and encodes a member of the transforming growth factor beta family that mediates some activities of the organizer. Xnr3 is transcriptionally activated by wnt signaling during gastrulation in the Xenopus embryo. A small region of the Xnr3 promoter is sufficient to confer wnt-inducible transcription. By mutational analysis of the promoter, two distinct sequence elements have been identified that are required for the response to wnt signals. One regulatory sequence interacts with a factor that accumulates in Xenopus gastrulae independent of wnt signaling. The other functionally important site can bind mammalian LEF-1 protein, a member of the LEF-1/TCF family of transcription factors. Misexpression of LEF-1 in embryo explants induces transcription of the endogenous Xnr3 gene. Taken together, these data provide further evidence for a role of LEF-1/TCF proteins in wnt signaling and identify the Spemann organizer-specific gene Xnr3 as a direct target of these transcription factors in vertebrates. Two other genes are known to respond to wnt signals: goosecoid, and siamois (McDendry, 1997).

Mesoderm induction and patterning are mediated by members of the TGFbeta superfamily. A novel zebrafish member, antivin, has been isolated that structurally is highly related to mouse lefty. Expression of atv is first detected after the midblastula transition in cells at the dorsal blastoderm margin, which later gives rise to the zebrafish equivalent of the organizer. During late blastula stages, expression spreds around the margin, but then decreases progressively during gastrulation, except in the dorsalmost region. In the gastrula midline, atv is expressed in cells at the leading edge of the mesendoderm, which extends anteriorly under the ectoderm to form the prochordal plate. Overexpression of antivin completely abolishes mesoderm induction at blastula stage, yet resultant embryos develop well-patterned epidermal and neural derivatives. The mesoderm-inhibiting activity of antivin can be mimicked by lefty and is suppressed by increasing levels of the mesodermal inducer Activin or its receptors. On the basis of its expression and activity, it is proposed that Antivin normally functions as a competitive inhibitor of Activin to limit mesoderm induction in the early embryo. Activin receptor type IIa and a constitutively activated form of a type I activin receptor override the Antivin inhibitory effect. In conclusion, this study strongly suggests that Atv acts at the level of receptor binding to negatively regulate the Activin signaling pathway. In the absence of mesendoderm (accomplished by Antivin-mediated 'knock-out', epiboly and convergent movements occur normally at gastrulation. However, elongation of the embryonic axis is strongly disturbed and anterior cephalic neurectodem territories rather than differentiating close to the animal pole, follow the movement of the margin toward the vegetal region. This demonstrates that mesoderm is required for the correct elongation of the embryonic axis (Thisse, 1999).

Loss-of-function analysis has shown that the transforming growth factor-like signaling molecule nodal is essential for mouse mesoderm development. However, definitive proof of nodal function in other developmental processes in the mouse embryo has been lacking because the null mutation blocks gastrulation. The generation and analysis of a hypomorphic nodal allele is described. Mouse embryos heterozygous for the hypomorphic allele and a null allele undergo gastrulation but then display abnormalities that fall into three distinct mutant phenotypic classes, which may result from expression levels falling below critical thresholds in one or more domains of nodal expression. Analysis of each of these classes provides conclusive evidence for nodal-mediated regulation of several developmental processes in the mouse embryo, beyond its role in mesoderm formation. Nodal signaling is required for correct positioning of the anteroposterior axis, normal anterior and midline patterning, and the left-right asymmetric development of the heart, vasculature, lungs and stomach (Lowe, 2001).

DPP homologs: The roles of members of the TGF-beta family in left-right asymmetry

A microtubule-dependent motor, the murine protein KIF3B, was disrupted by gene targeting. The null mutants did not survive beyond midgestation, exhibiting growth retardation, pericardial sac ballooning, and neural tube disorganization. Prominently, the left-right asymmetry was randomized in the heart loop and the direction of embryonic turning. lefty-2 (a TGF beta-family member) expression was either bilateral or absent. Furthermore, the node lacked monocilia while the basal bodies were present. Immunocytochemistry revealed KIF3B localization in wild-type nodal cilia. Video microscopy showed that these cilia were motile and generated a leftward flow. These data suggest that KIF3B is essential for the left-right determination through intraciliary transportation of materials for ciliogenesis of motile primary cilia that could produce a gradient of putative morphogen along the left-right axis in the node (Nonaka, 1998).

In vertebrates, lateralization of the CNS is evident both in terms of asymmetric behaviors and in the localization of specific cognitive abilities predominantly to one side of the brain. Differences in neural structures on left and right sides of the vertebrate brain have been described at early developmental stages, suggesting genetic regulation of CNS asymmetry. However, epigenetic factors also influence asymmetry, and it has proved difficult to determine if there is a relationship between laterality of the CNS and handedness of the viscera and heart. Indeed, although a wide variety of studies have elucidated the pathways regulating heart and visceral asymmetry, these studies have shed little light on the development of CNS asymmetry (Concha, 2000 and references therein).

The Nodal pathway is one of several signaling pathways implicated in the establishment of organ asymmetry. nodal genes are expressed asymmetrically in the left lateral plate mesoderm, and their misexpression can lead to altered left/right polarity of organs. Further support for a role for the Nodal pathway in laterality decisions has come from analysis of proteins belonging to the EGF-CFC, Lefty, and Fast families, which are all implicated in Nodal signaling. For instance, the EGF-CFC protein One-eyed pinhead (Oep) is a membrane attached protein that acts as an essential cofactor for the Nodal-related signals Cyclops (Cyc) and Squint (Sqt) in zebrafish. Fish lacking late activity of Oep and mice lacking the function of a related protein, Cryptic, exhibit heterotaxia (randomization of asymmetry) and fail to establish left-sided expression of genes implicated in the Nodal signaling pathway. A very similar phenotype is observed in zebrafish schmalspur (sur) mutants, which carry mutations in Fast1, a transcriptional effector of Nodal signaling. Moreover, mice lacking Lefty1, an antagonist of Nodal signaling, exhibit altered patterns of asymmetric gene expression and left pulmonary isomerism. These results suggest that EGF-CFC/Fast/Lefty-regulated Nodal signaling is essential for the regulation of asymmetric gene expression in the lateral plate mesoderm and left-right development of internal organs. Since loss of function mutations in hCRYPTIC are associated with human left-right laterality defects, an essential role for Nodal signaling in left-right axis formation appears to be conserved from fish to humans (Concha, 2000 and references therein).

In addition to domains of expression in the lateral plate mesoderm, the genes oep, sur, cyc, the lefty gene antivin/lefty1 (atv), and the homeobox gene pitx2 are also expressed in the zebrafish brain. oep and sur are expressed bilaterally in the diencephalon. In contrast, cyc, atv, and pitx2 are expressed exclusively on the left side of the brain. These observations suggest that Nodal signaling may have a function in the brain comparable to its role in the lateral plate mesoderm. In support of this, several mutations that affect the laterality of gene expression in the mesoderm and situs of the viscera and heart also affect expression of left-sided genes in the brain. It has been unclear, however, if these changes in CNS gene expression have any consequences on the development of neuroanatomical asymmetries (Concha, 2000 and references therein).

In order to study the genetic basis of left-right asymmetry in the CNS, asymmetries in the diencephalic habenular nuclei and pineal complex of the larval zebrafish brain have been examined. The habenulae are dorsal diencephalic nuclei that receive input from the telencephalon via the stria medullaris and they relay information to the interpeduncular nuclei in the ventral midbrain via the fasciculi retroflexus. These pathways are conserved in all vertebrates and are likely to represent an evolutionarily conserved output pathway from the telencephalon. The habenular nuclei are often bilaterally symmetrical in teleosts but asymmetries are described in some species. The left habenular nucleus is considerably larger than the right nucleus in larval zebrafish (Concha, 2000 and references therein).

In addition to the habenulae, the dorsal diencephalon gives rise to several evaginations that include the epiphysis/pineal organ and the parapineal organ. In adult animals, the pineal organ has endocrine roles that include secretion of melatonin, a hormonal regulator of circadian rhythms. Additionally, in many species, the epiphysis and/or parapineal organs are photoreceptive and, in some amphibia and reptiles, may form true eye-like structures. During embryogenesis, the pineal complex develops earlier than the lateral eyes and may mediate early light-evoked behaviors. The epiphysis is usually a midline structure, while, in some species, the parapineal occupies a position to the left of the midline. The parapineal is located to the left in larval zebrafish (Concha, 2000).

The regulation of left-sided expression of genes that function in the Nodal pathway have been examined, as well as how components of this pathway influence asymmetry and laterality in the zebrafish forebrain. Analysis of zebrafish embryos with compromised Nodal signaling reveals an early role for this pathway in the repression of asymmetrically expressed genes in the diencephalon. Oep and Sur function is subsequently required to overcome this repression and an intact midline is required to limit expression to one side of the brain. Strikingly, it is found that in mutants in which expression of left-sided genes is either absent or is bilateral, anatomical asymmetry of the habenulae and parapineal is still established but laterality is randomized. These results indicate that components of the Nodal signaling pathway and the midline regulate the laterality of vertebrate brain asymmetry (Concha, 2000).

In the mouse and chick embryo, the node plays a central role in generating left-right (LR) positional information. Using several different strategies, evidence is provided in the mouse that bone morphogenetic protein 4 (Bmp4) is required independently in two different sites for node morphogenesis and for LR patterning. Bmp4 expression in the trophoblast-derived extra-embryonic ectoderm is essential for the normal formation of the node and primitive streak. However, tetraploid chimera analysis demonstrates that Bmp4 made in epiblast-derived tissues is required for robust LR patterning, even when normal node morphology is restored. In the absence of embryonic Bmp4, the expression of left-side determinants such as Nodal and Lefty2 is absent in the left lateral plate mesoderm (LPM). Noggin-mediated inhibition of Bmp activity in cultured wild-type embryos results in suppression of Nodal expression in the LPM. Thus, unlike previous models proposed in the chick embryo in which Bmp4 suppresses left-sided gene expression, these results suggest that Bmp acts as a positive facilitator of the left-sided molecular cascade and is required for Nodal induction and maintenance in the left LPM (Fujiwara, 2002).

Left-right (LR) asymmetry of the heart in vertebrates is regulated by early asymmetric signals in the embryo, including the secreted signal Sonic hedgehog (Shh), but less is known about LR asymmetries in visceral organs. Shh also specifies asymmetries in visceral precursors in the zebrafish (cardiac and visceral sidedness are independent of one another). The transcription factors fli-1 and Nkx-2.5 are expressed asymmetrically in the precardiac mesoderm and subsequently in the heart; an Eph receptor, rtk2, and an adhesion protein, DM-GRASP, mark early asymmetries in visceral endoderm. Misexpression on the right side of either shh mRNA, or a dominant negative form of protein kinase A, reverses the expression of these asymmetries in precursors of both the heart and the viscera. Reversals in the heart and gut are uncoordinated, suggesting that each organ interprets the signal independently. Misexpression of Bone Morphogenetic Protein (BMP4) on the right side reverses the heart, but visceral organs are unaffected, consistent with a function for BMPs locally in the heart field. Zebrafish mutants with midline defects show independent reversals of cardiac and visceral laterality. Thus, hh signals influence the development of multiple organ asymmetries in zebrafish and different organs appear to respond independently to a central cascade of midline signaling, which in the heart involves BMP4 (Schilling, 1999).

Determination of the left-right (L-R) axis implicates several genes, among which TGFbeta-related molecules such as Activin betaB, lefty1 and 2 and Nodal. Bmp4 and its signal transduction pathway partners BMPR IA and Smad1 are transiently expressed on the right side of Hensen's node, when L-R polarity is being established. Moreover, Smad1 is expressed asymmetrically in the nascent notochord. These observations suggest a role for a BMP4-dependent autocrine or paracrine mechanism during early L-R determination (Monsoro-Burq, 2000).

In chick embryos, the first signs of left-right asymmetry are detected in Hensen's node, essentially by left-sided Sonic Hedgehog (Shh) expression. After a gap of several hours, SHH induces polarized gene activities in the left paraxial mesoderm. During this time period, BMP4 signaling is necessary and sufficient to maintain Shh asymmetry within the node. SHH and BMP4 proteins negatively regulate each other's transcription, resulting in a strict complementarity between these two gene patterns on each side of the node. Noggin, present in the midline at this stage, limits BMP4 spreading. Moreover, BMP4 is downstream of Activin signals and controls Fgf8. Thus, early BMP4 signaling coordinates left and right pathways in Hensen's node (Monsoro-Burq, 2001).

The heart develops from a linear tubular precursor, which loops to the right and undergoes terminal differentiation to form the multichambered heart. Heart looping is the earliest manifestation of left-right asymmetry and determines the eventual heart situs. The signaling processes that impart laterality to the unlooped heart tube and thus allow the developing organ to interpret the left-right axis of the embryo are poorly understood. Recent experiments in zebrafish have led to the suggestion that bone morphogenetic protein 4 (BMP4) may impart laterality to the developing heart tube. In Xenopus, as in zebrafish, BMP4 is expressed predominantly on the left of the linear heart tube. Furthermore ectopic expression of Xenopus nodal-related protein 1 (Xnr1) RNA affects BMP4 expression in the heart, linking asymmetric BMP4 expression to the left-right axis. Transgenic embryos overexpressing BMP4 bilaterally in the heart tube tend towards a randomization of heart situs in an otherwise intact left-right axis. Additionally, inhibition of BMP signalling by expressing noggin or a truncated, dominant negative BMP receptor prevents heart looping but allows the initial events of chamber specification and anteroposterior morphogenesis to occur. Thus in Xenopus asymmetric BMP4 expression links heart development to the left-right axis, by being both controlled by Xnr1 expression and necessary for heart looping morphogenesis (Breckenridge, 2001).

Shh acts early in the development of the axial skeleton, to induce a prochondrogenic response to later BMP signaling. Somitic expression of the transcription factor Nkx3.2, homolog of Drosophila bagpipe, is initiated by Shh and sustained by BMP signals. Misexpression of Nkx3.2 in somitic tissue confers a prochondrogenic response to BMP signals. The transcriptional repressor activity of Nkx3.2 is essential for this factor to promote chondrogenesis. Conversely, a 'reverse function' mutant of Nkx3.2 that has been converted into a transcriptional activator inhibits axial chondrogenesis in vivo. It is concluded that Nkx3.2 is a critical mediator of the actions of Shh during axial cartilage formation, acting to inhibit expression of factors that interfere with the prochondrogenic effects of BMPs (Murtaugh, 2001).

A model of left-right axis formation in the chick involves inhibition of bone morphogenetic proteins by the antagonist Caronte (Car), a member of the Dan family of BMP antagonists, as a mechanism of upregulating Nodal in the left lateral plate mesoderm. By contrast, expression of CFC, a competence factor which is absolutely required for Nodal signaling in the lateral plate mesoderm, is dependent on a functional BMP signaling pathway. EGF-CFC factors are attached to the plasma membrane via glycosylphosphatidylinositol (GPI) anchors and contain a modified EGF-like domain and a cysteine-rich CFC domain. Both domains are required for physical interaction of EGF-CFC factors with Nodal and its signal-transducing receptor complex consisting of type I receptors, ALK4 or ALK7 and the type II receptors, ActRIIA or ActRIIB (Schlange, 2002 and references therein).

The relationship between BMP, a right-sided signal, and Nodal, a left-sided signal, was investigated in further detail. BMP2 and Noggin-expressing cells were implanted into the left lateral plate mesoderm and paraxial mesoderm and a strong upregulation of Nodal and its target genes Pitx2 and Nkx3.2 was observed. In addition Cfc, the Nodal type II receptor ActrIIa and Snr, a zinc finger repressor, which mediates suppression of Pitx2 in the right lateral plate mesoderm, were found to depend on BMP signaling for their expression. Comparison of the expression domains of Nodal, Bmp2, Car and Cfc revealed co-expression of Nodal, Cfc and Bmp2, while Car and Nodal only partially overlapped. Ectopic application of BMP2, Nodal, and Car as well as combinations of these signaling molecules to the right lateral plate mesoderm revealed that BMP2 and Car need to synergize in order to specify left identity. A novel model of left-right axis formation is proposed that involves BMP as a positive regulator of Nodal signaling in the chick embryo (Schlange, 2002).

Exogenous application of BMP to the lateral plate mesoderm (LPM) of chick embryos at the early somite stage has a positive effect on Nodal expression. BMP applications into the right LPM are followed by a rapid activation of Nodal, while applications into the left LPM result in expansion of the normal domain of Nodal expression. Conversely, blocking of BMP signaling by Noggin in the left LPM interfers with the activation of Nodal expression. These results support a positive role for endogenous BMP on Nodal expression in the LPM. BMP positively regulates the expression of Caronte, Snail and Cfc in both the left and right LPM. BMP-treated embryos have molecular impairment of the midline with downregulation of Lefty1, Brachyury and Shh, but it is also shown that the midline defect is not sufficient to induce ectopic Nodal expression (Piedra, 2002).

DPP homologs and head formation in Xenopus

Bone morphogenetic proteins (BMPs) and their antagonists are involved in the axial patterning of vertebrate embryos. Both BMP-3b and BMP-3 (also called GDF-10 and osteogenin respectively) dorsalize Xenopus embryos, but act as dissimilar antagonists within the BMP family. BMP-3b injected into Xenopus embryos triggers secondary head formation in an autonomous manner, whereas BMP-3 induces aberrant tail formation. At the molecular level, BMP-3b antagonizes nodal-like proteins and ventralizing BMPs, whereas BMP-3 antagonizes only the latter. These differences are due to divergence of their pro-domains. Less BMP-3b than BMP-3 precursor is proteolytically processed in embryos. BMP-3b protein associates with a monomeric form of Xnrl, a nodal-like protein, whereas BMP-3 does not. These molecular features are consistent with their expression profiles during Xenopus development. XBMP-3b is expressed in the prechordal plate, while xBMP-3 is expressed in the notochord. Using antisense morpholino oligonucleotides, it was found that the depletion of both xBMP-3b and cerberus, a head inducer, causes headless Xenopus embryos, whereas the depletion of both xBMP-3 and cerberus affects the size of the somite. These results reveal that xBMP-3b and cerberus are essential for head formation regulated by the Spemann organizer, and that xBMP-3b and perhaps xBMP-3 are involved in the axial patterning of Xenopus embryos (Hino, 2003).

DPP homologs and rostral/caudal patterning in Xenopus

Recent revisions to the Xenopus fate map challenge the interpretation of previous maps and current models of amphibian axial patterning. The rostral-most contributions to both dorsal and ventral mesoderm concomitantly from marginal zone progenitors were determined in stage 6 embryos. Data reveal an unequivocal rostral-to-caudal progression of both dorsal and ventral mesoderm across the pre-gastrula axis, historically called the dorsal-ventral axis, and a dorsal-to-ventral progression from animal-to-vegetal in the marginal zone. These findings support the proposed revisions to the fate and axis orientation maps. Most importantly, these results raise questions about the role of the organizer grafts and organizer-derived BMP antagonists in the 'induction' of secondary axes. Both phenomena were reexamined; organizer grafts and BMP antagonists evoke caudal-to-rostral mesodermal fate transformations, and not ventral-to-dorsal transformations as currently believed. BMP antagonism evokes a second axis because it stimulates precocious mediolateral intercalation of caudal, dorsal mesoderm. The implications of these findings for models of organizer function in vertebrate axial patterning are discussed (Constance Lane, 2004).

Currently, there are contradictory fate maps for the amphibian Xenopus laevis, and the discrepancies in tissue distributions between the maps led to conflicting assignments of the dorsal-ventral and rostral-caudal axes. The discrepancies involve primarily the mesoderm. In the conventional fate map, the dorsal-ventral axis runs from Spemann's organizer in the dorsal marginal zone (DMZ) to the meridian of sperm entry in the ventral marginal zone (VMZ). Mesodermal tissues are arranged notochord (most dorsal in the modern interpretation), muscle, pronephros, and blood (most ventral). The rostral-caudal axis is not assigned in the conventional map. In the revised map, it is the rostral-caudal axis that runs from Spemann's organizer to the meridian of sperm entry. The dorsal-ventral axis of the mesoderm is reassigned to the animal/vegetal axis. Since the region including and immediately surrounding Spemann's organizer (which is historically called the DMZ) is the source of anterior mesoderm, it was renamed the rostral marginal zone (RMZ). The marginal zone on the opposite side of the embryo (historically called the VMZ) was renamed the caudal marginal zone (CMZ), as it is the source of posterior mesoderm. Thus, the two maps and the axis designations are incompatible. The discrepancies between the maps must be resolved. To this end, this study reports the rostral-caudal and dorsal-ventral topographic projection from the stage 6 blastula marginal zone to the mesoderm of the tadpole (stage 32-34). The results reported here support the revised map and contradict the old map (Constance Lane, 2004).

The role of BMP antagonists secreted by the organizer were examined, since an extensive body of experimental evidence concludes that BMP antagonists from the organizer act to 'dorsalize' mesoderm. The BMP antagonist noggin behaves like an organizer graft, stimulating precocious differentiation of caudal tissue. BMP antagonists provoke secondary axis formation by mediating precocious mediolateral intercalation behavior. Several lines of experimentation reveal that the marginal zone is prepatterned into animal and vegetal morphogenetic domains, which correspond to the prospective dorsal and ventral mesoderm fields, respectively. When considered in combination with the revised mapping data, these results alter the modern view of amphibian and vertebrate axial patterning. They challenge investigators to rethink all experiments and interpretations based on the old fate maps, and to devise a new working model of vertebrate axial patterning (Constance Lane, 2004).

The BMP signaling gradient patterns dorsoventral tissues in a temporally progressive manner along the anteroposterior axis

Patterning of the vertebrate anteroposterior (AP) axis proceeds temporally from anterior to posterior. How dorsoventral (DV) axial patterning relates to AP temporal patterning is unknown. This study examined the temporal activity of BMP signaling in patterning ventrolateral cell fates along the AP axis, using transgenes that rapidly turn 'off' or 'on' BMP signaling. BMP signaling patterns rostral DV cell fates at the onset of gastrulation, whereas progressively more caudal DV cell fates are patterned at progressively later intervals during gastrulation. Increased BMP signal duration is not required to pattern more caudal DV cell fates; rather, distinct temporal intervals of signaling are required. This progressive action is regulated downstream of, or in parallel to, BMP signal transduction at the level of Smad1/5 phosphorylation. It is proposed that a temporal cue regulates a cell's competence to respond to BMP signaling, allowing the acquisition of a cell's DV and AP identity simultaneously (Tucker, 2008).

Taking into account the kinetics of the BMP inhibition system, the results are summarized in a model. The embryo is broadly divided into sectors reflecting DV and AP positions. The DV position of a single cell in each sector, and thus the level of BMP activity the cell experiences during gastrulation, dictates the cell's fate, as either neurectoderm (no BMP activity), cranial neural crest (intermediate BMP activity), or epidermis (high BMP activity). The levels of BMP activity that each prospective cell type will experience were plotted over time from midblastula, when low-level signaling is first detected, through midgastrulation. By the onset of gastrulation at 5.5 hpf, a BMP gradient is established such that ventrolateral cells exposed to an intermediate and high BMP level become cranial neural crest and epidermis, respectively, while dorsal cells, exposed to little or no signal, become neural tissue. Cells positioned along the AP axis at the midbrain-hindbrain boundary, R3, and R5 positions each interpret the BMP signaling gradient during different critical intervals. In wild-type, the shape of the gradient remains fairly constant from early to midgastrulation, thus generating the proper DV fate of each cell at the three AP positions over time (Tucker, 2008).

Table of contents

decapentaplegic: Biological Overview | Transcriptional regulation | Targets of activity | Protein Interactions | Post-transcriptional Regulation | Developmental Biology | Effect of mutation | References

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.