Interplay between Wnt2 and Wnt2bb controls multiple steps of early foregut-derived organ development

The vertebrate liver, pancreas and lung arise in close proximity from the multipotent foregut endoderm. Tissue-explant experiments uncovered instructive signals emanating from the neighbouring lateral plate mesoderm, directing the endoderm towards specific organ fates. This suggested that an intricate network of signals is required to control the specification and differentiation of each organ. Here, we show that sequential functions of Wnt2bb and Wnt2 control liver specification and proliferation in zebrafish. Their combined specific activities are essential for liver specification, as their loss of function causes liver agenesis. Conversely, excess wnt2bb or wnt2 induces ectopic liver tissue at the expense of pancreatic and anterior intestinal tissues, revealing the competence of intestinal endoderm to respond to hepatogenic signals. Epistasis experiments revealed that the receptor frizzled homolog 5 (fzd5) mediates part of the broader hepatic competence of the alimentary canal. fzd5 is required for early liver formation and interacts genetically with wnt2 as well as wnt2bb. In addition, lack of both ligands causes agenesis of the swim bladder, the structural homolog of the mammalian lung. Thus, tightly regulated spatiotemporal expression of wnt2bb, wnt2 and fzd5 is central to coordinating early liver, pancreas and swim bladder development from a multipotent foregut endoderm.     

Regulation of hhex expression in the yolk syncytial layer, the potential Nieuwkoop center homolog in zebrafish

The Nieuwkoop center is the earliest signaling center during dorsal-ventral pattern formation in amphibian embryos and has been implied to function in induction of the Spemann-Mangold organizer. In zebrafish, Nieuwkoop-center-like activity resides in the dorsal yolk syncytial layer (YSL) at the interface of the vegetal yolk cell and the blastoderm. hex homologs are expressed in the anterior endomesoderm in frogs (Xhex), the anterior visceral endoderm in mice, and the dorsal YSL in zebrafish (hhex). Here, we investigate the control of hhex expression in the YSL. We demonstrate that bozozok (boz) is absolutely required for early hhex expression, while overexpression of boz causes ectopic hhex expression. Activation of Wnt/beta-catenin signaling by LiCl induces hhex expression in wild-type YSL but not in boz mutant embryos, revealing that boz activity is required downstream of Wnt/beta-catenin signaling for hhex expression. Further, we show that the boz-mediated induction of hhex is independent of the Boz-mediated repression of bmp2b. Our data reveal that repressive effects of both Vega1 and Vega2 may be responsible for the exclusion of hhex expression from the ventral and lateral parts of the YSL. In summary, zebrafish hhex appears to be activated by Wnt/beta-catenin in the dorsal YSL, where Boz acts in a permissive way to limit repression of hhex by Vega1 and Vega2.     

Zebrafish hhex, nk2.1a, and pax2.1 regulate thyroid growth and differentiation downstream of Nodal-dependent transcription factors

During zebrafish development, the thyroid primordium initiates expression of molecular markers such as hhex and nk2.1a in the endoderm prior to pharynx formation. As expected for an endodermally derived organ, initiation of thyroid development depends on Nodal signalling. We find that it also depends on three downstream effectors of Nodal activity, casanova (cas), bonnie and clyde (bon), and faust (fau)/gata5. Despite their early Nodal-dependent expression in the endoderm, both hhex and nk2.1a are only required relatively late during thyroid development. In hhex and nk2.1a loss-of-function phenotypes, thyroid development is initiated and arrests only after the primordium has evaginated from the pharyngeal epithelium. Thus, like pax2.1, both hhex and nk2.1a have similarly late roles in differentiation or growth of thyroid follicular cells, and here, we show that all three genes act in parallel rather than in a single pathway. Our functional analysis suggests that these genes have similar roles as in mammalian thyroid development, albeit in a different temporal mode of organogenesis.     

Anterior endomesoderm specification in Xenopus by Wnt/beta-catenin and TGF-beta signalling pathways.

In Xenopus, XHex and cerberus are early marker genes of the anterior endomesoderm (AE), a subset of endoderm cells fated to form the liver and foregut and implicated in head induction. Using XHex and cerberus as markers we have examined the signals underlying AE induction. We show that the AE is specified by the early blastula in the absence of mesodermal signals but that cell-cell contact between presumptive AE cells is required. In overexpression experiments maternal Wnt/beta-catenin and TGF-beta signals (Vg1, Xnr1-2) can induce ectopic XHex and cerberus. Inhibiting these pathways with dominant interfering signalling components blocks endogenous XHex and cerberus expression. We assess the role of signals from the organiser and show that the BMP antagonists noggin and chordin are important for maintaining XHex and cerberus expression. Finally, ventral injection of XHex mRNA can induce ectopic cerberus. Our results indicate that endodermal and mesodermal patterning are closely coordinated and that the AE is likely to be specified by the combined action of dorsal Wnt/beta-catenin signals and endoderm-specific factors mediated by TGF-beta signalling. These results provide a starting point for understanding the molecular events underlying the progressive determination of endodermally derived organs, such as the liver and foregut.     

Disruption of early proximodistal patterning and AVE formation in Apc mutants

In the postimplantation mouse embryo, axial patterning begins with the restriction of expression of a set of genes to the distal visceral endoderm (DVE). This proximodistal (PD) axis is subsequently transformed into an anteroposterior axis as the VE migrates anteriorly to form the anterior visceral endoderm (AVE). Both Nodal and Wnt signaling pathways are involved in these events. We show here that loss of function in the adenomatous polyposis coli gene (Apc) leads to constitutive beta-catenin activity that induces a proximalization of the epiblast with the activation of a subset of posterior mesendodermal genes, and loss of ability to induce the DVE. The loss of some DVE genes such as Hex and goosecoid is rescued in chimeras where only the epiblast was wild type; however, these DVE markers were no longer restricted distally but covered the entire epiblast. Thus, the Apc gene is needed in both embryonic and extraembryonic lineages for normal PD patterning around implantation, suggesting that early restricted activation of the Wnt pathway may be important for initiating axial asymmetries. In addition, we found that nuclear beta-catenin and other molecular markers are asymmetrically expressed by 4.5 days.     

Barx1-mediated inhibition of Wnt signaling in the mouse thoracic foregut controls tracheo-esophageal septation and epithelial differentiation

Mesenchymal cells underlying the definitive endoderm in vertebrate animals play a vital role in digestive and respiratory organogenesis. Although several signaling pathways are implicated in foregut patterning and morphogenesis, and despite the clinical importance of congenital tracheal and esophageal malformations in humans, understanding of molecular mechanisms that allow a single tube to separate correctly into the trachea and esophagus is incomplete. The homoebox gene Barx1 is highly expressed in prospective stomach mesenchyme and required to specify this organ. We observed lower Barx1 expression extending contiguously from the proximal stomach domain, along the dorsal anterior foregut mesenchyme and in mesenchymal cells between the nascent esophagus and trachea. This expression pattern exactly mirrors the decline in Wnt signaling activity in late development of the adjacent dorsal foregut endoderm and medial mainstem bronchi. The hypopharynx in Barx1^−/− mouse embryos is abnormally elongated and the point of esophago-tracheal separation shows marked caudal displacement, resulting in a common foregut tube that is similar to human congenital tracheo-esophageal fistula and explains neonatal lethality. Moreover, the Barx1^−/− esophagus displays molecular and cytologic features of respiratory endoderm, phenocopying abnormalities observed in mouse embryos with activated ß-catenin. The zone of canonical Wnt signaling is abnormally prolonged and expanded in the proximal Barx1^−/− foregut. Thus, as in the developing stomach, but distinct from the spleen, Barx1 control of thoracic foregut specification and tracheo-esophageal septation is tightly associated with down-regulation of adjacent Wnt pathway activity.     

Sizzled-Tolloid Interactions Maintain Foregut Progenitors by Regulating Fibronectin-Dependent BMP Signaling

The liver, pancreas, and lungs are induced from endoderm progenitors by a series of dynamic growth factor signals from the mesoderm, but how the temporal-spatial activity of these signals is controlled is poorly understood. We have identified an extracellular regulatory loop required for robust bone morphogenetic protein (BMP) signaling in the Xenopus foregut. We show that BMP signaling is required to maintain foregut progenitors and induce expression of the secreted frizzled related protein Sizzled (Szl) and the extracellular metalloprotease Tolloid-like 1 (Tll1). Szl negatively regulates Tll activity to control deposition of a fibronectin (FN) matrix between the mesoderm and endoderm, which is required to maintain BMP signaling. Foregut-specific Szl depletion results in a loss of the FN matrix and failure to maintain robust pSmad1 levels, causing a loss of foregut gene expression and organ agenesis. These results have implications for BMP signaling in diverse contexts and the differentiation of foregut tissue from stem cells.     

Unique and conserved aspects of gut development in zebrafish

Although the development of the digestive system of humans and vertebrate model organisms has been well characterized, relatively little is known about how the zebrafish digestive system forms. We define developmental milestones during organogenesis of the zebrafish digestive tract, liver, and pancreas and identify important differences in the way the digestive endoderm of zebrafish and amniotes is organized. Such differences account for the finding that the zebrafish digestive system is assembled from individual organ anlagen, whereas the digestive anlagen of amniotes arise from a primitive gut tube. Despite differences of organ morphogenesis, conserved molecular programs regulate pharynx, esophagus, liver, and pancreas development in teleosts and mammals. Specifically, we show that zebrafish faust/gata-5 is a functional ortholog of gata-4, a gene that is essential for the formation of the mammalian and avian foregut. Further, extraembryonic gata activity is required for this function in zebrafish as has been shown in other vertebrates. We also show that a loss-of-function mutation that perturbs sonic hedgehog causes defects in the development of the esophagus that parallel those associated with targeted disruption of this gene in mammals. Perturbation of sonic hedgehog also affects zebrafish liver and pancreas development, and these effects occur in a reciprocal fashion, as has been described during mammalian liver and ventral pancreas development. Together, these data define aspects of digestive system development necessary for the characterization of zebrafish mutants. Given the similarities of teleost and mammalian digestive physiology and anatomy, these findings have implications for developmental and evolutionary studies as well as research of human diseases, such as diabetes, liver cirrhosis, and cancer.     

Pancreas-specific deletion of beta-catenin reveals Wnt-dependent and Wnt-independent functions during development

Mutations and deregulation of adenomatous polyposis coli (APC) and beta-catenin are implicated in specific cancers of the pancreas, but the role of Wnt pathway in normal pancreas development and homeostasis is unknown. This article reports a comprehensive investigation of the activity and the role of the Wnt pathway in pancreas organogenesis. We have used two reporter lines to monitor canonical Wnt pathway activity during development and after birth and demonstrate activity in endocrine cells and in the mesenchyme. We have specifically deleted the beta-catenin gene in the epithelium of the pancreas and duodenum by using Pdx1-Cre mice. In agreement with Wnt pathway activity in pancreatic endocrine cells, we find a reduction in endocrine islet numbers. Our study reveals that beta-catenin deletion also affects cells in which Wnt pathway activity is not detected. Indeed, beta-catenin mutant cells have a competitive disadvantage during development that also affects the exocrine compartment. Moreover, the conditional knockout (KO) mice develop acute edematous pancreatitis perinatally due to the disruption of the epithelial structure of acini. These effects are likely to be due to the function of beta-catenin at the membrane. Mice later recover from pancreatitis and regenerate normal pancreas and duodenal villi from the wild-type (wt) cells that escape beta-catenin deletion.     

Wnt6 induces the specification and epithelialization of F9 embryonal carcinoma cells to primitive endoderm

Epithelial-to-mesenchymal transitions (EMTs) play key roles in the normal development of an organism as well as its demise following the metastasis of a malignant tumour. An EMT during early mouse development results in the differentiation of primitive endoderm into the parietal endoderm that forms part of the parietal yolk sac. In the embryo, primitive endoderm develops from cells in the inner cell mass, but the signals that instruct these cells to become specified and adopt an epithelial fate are poorly understood. The mouse F9 teratocarcinoma cell line, a model that can recapitulate the in vivo primitive to parietal endoderm EMT, has been used extensively to elucidate the signalling cascades involved in extraembryonic endoderm differentiation. Here, we identified Wnt6 as a gene up-regulated in F9 cells in response to RA and show that Wnt6 expressing cells or cells exposed to Wnt6 conditioned media form primitive endoderm. Wnt6 induction of primitive endoderm is accompanied by beta-catenin and Snail1 translocation to the nucleus and the appearance of cytokeratin intermediate filaments. Attenuating glycogen synthase kinase 3 activity using LiCl gave similar results, but the fact that cells de-differentiate when LiCl is removed reveals that other signalling pathways are required to maintain cells as primitive endoderm. Finally, Wnt6-induced primitive endodermal cells were tested to determine their competency to complete the EMT and differentiate into parietal endoderm. Towards that end, results show that up-regulating protein kinase A activity is sufficient to induce markers of parietal endoderm. Together, these findings indicate that undifferentiated F9 cells are responsive to canonical Wnt signalling, which negatively regulates glycogen synthase kinase 3 activity leading to the epithelialization and specification of primitive endoderm competent to receive additional signals required for EMT. Considering the ability of F9 cells to mimic an in vivo EMT, the identification of this Wnt6-beta-catenin-Snail signalling cascade has broad implications for understanding EMT mechanisms in embryogenesis and metastasis.     

Sfrp5 is not essential for axis formation in the mouse

Secreted frizzled related protein (Sfrp) genes encode extracellular factors that can modulate Wnt signaling. During early post-implantation mouse development Sfrp5 is expressed in the anterior visceral endoderm (AVE) and the ventral foregut endoderm. The AVE is important in anterior-posterior axis formation and the ventral foregut endoderm contributes to multiple gut tissues. Here to determine the essential role of Sfrp5 in early mouse development we generated Sfrp5-deficient mice by gene targeting. We report that Sfrp5-deficient mice are viable and fertile. To determine whether the absence of an axis phenotype might be due to genetic redundancy with Dkk1 in the AVE we generated Sfrp5;Dkk1 double mutant mice. AVE development and primitive streak formation appeared normal in Sfrp5(-/-);Dkk1(-/-) embryos. These results indicate that Sfrp5 is not essential for axis formation or foregut morphogenesis in the mouse and also imply that Sfrp5 and Dkk1 together are not essential for AVE development.     

Expression and function of mouse Sox17 gene in the specification of gallbladder/bile-duct progenitors during early foregut morphogenesis

In early-organogenesis-stage mouse embryos, the posteroventral foregut endoderm adjacent to the heart tube gives rise to liver, ventral pancreas and gallbladder. Hepatic and pancreatic primordia become specified in the posterior segment of the ventral foregut endoderm at early somite stages. The mechanisms for demarcating gallbladder and bile duct primordium, however, are poorly understood. Here, we demonstrate that the gallbladder and bile duct progenitors are specified in the paired lateral endoderm domains outside the heart field at almost the same timing as hepatic and pancreatic induction. In the anterior definitive endoderm, Sox17 reactivation occurs in a certain population within the most lateral domains posterolateral to the anterior intestinal portal (AIP) lip on both the left and right sides. During foregut formation, the paired Sox17-positive domains expand ventromedially to merge in the midline of the AIP lip and become localized between the liver and pancreatic primordia. In Sox17-null embryos, these lateral domains are missing, resulting in a complete loss of the gallbladder/bile-duct structure. Chimera analyses revealed that Sox17-null endoderm cells in the posteroventral foregut do not display any gallbladder/bile-duct molecular characters. Our findings show that Sox17 functions cell-autonomously to specify gallbladder/bile-duct in the mouse embryo.     

Bmp and Fgf signaling are essential for liver specification in zebrafish

Based on data from in vitro tissue explant and ex vivo cell/bead implantation experiments, Bmp and Fgf signaling have been proposed to regulate hepatic specification. However, genetic evidence for this hypothesis has been lacking. Here, we provide in vivo genetic evidence that Bmp and Fgf signaling are essential for hepatic specification. We utilized transgenic zebrafish that overexpress dominant-negative forms of Bmp or Fgf receptors following heat-shock induction. These transgenes allow one to bypass the early embryonic requirements for Bmp and Fgf signaling, and also to completely block Bmp or Fgf signaling. We found that the expression of hhex and prox1, the earliest liver markers in zebrafish, was severely reduced in the liver region when Bmp or Fgf signaling was blocked just before hepatic specification. However, hhex and prox1 expression in adjacent endodermal and mesodermal tissues appeared unaffected by these manipulations. Additional genetic studies indicate that the endoderm maintains competence for Bmp-mediated hepatogenesis over an extended window of embryonic development. Altogether, these data provide the first genetic evidence that Bmp and Fgf signaling are essential for hepatic specification, and suggest that endodermal cells remain competent to differentiate into hepatocytes for longer than anticipated.     

A bipotential precursor population for pancreas and liver within the embryonic endoderm

The pancreas emerges independently from dorsal and ventral domains of embryonic gut endoderm. Gene inactivation experiments in mice have identified factors required for dorsal pancreas development, but factors that initiate the ventral pancreas have remained elusive. In this study, we investigated the hypothesis that the emergence of the ventral pancreas is related to the emergence of the liver. We find that the liver and ventral pancreas are specified at the same time and in the same general domain of cells. Using embryo tissue explantation experiments, we find that the default fate of the ventral foregut endoderm is to activate the pancreas gene program. FGF signalling from the cardiac mesoderm diverts this endoderm to express genes for liver instead of those for pancreas. No evidence was found to indicate that the cell type choice for pancreas or liver involves a selection for growth or viability. Cardiac mesoderm or FGF induces the local expression of sonic hedgehog, which in turn is inhibitory to pancreas but not to liver. The bipotential precursor cell population for pancreas and liver in embryonic development and its fate selection by FGF has features that appear to be recapitulated in the adult pancreas and are reflected in the evolution of these organs.