Background and Significance to Research in Liver Development
Related reviews:
McLin V and Zorn AM (2006) Molecular control of liver development. Clin Liver Dis. Feb; 10 (1):1-25. Review
Zaret, KS (2002) Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet. 2002 Jul;3(7):499-512. Review.
B1 Significance
The liver provides many essential functions. It is the largest exocrine gland in the body, producing bile, metabolizing lipids, and is the primary site for detoxification and elimination of body wastes. The liver also performs important endocrine functions by secreting blood proteins such as albumin that maintain homeostasis, as well as regulating blood glucose levels through glycogen storage. Although the liver has a capacity to regenerate and repair, numerous conditions and pathologies result in dysfunctional or damaged livers, which are life threatening and where liver transplantation is the only option. Pediatric liver transplants account for 10-15% of all liver transplants worldwide and many of these occur because of congenital defects. Despite the essential functions of the liver, very little is understood about its early embryonic development. In 2002 the American Liver Foundation published a "Pediatric Liver Research Agenda" advocating that a better understanding embryonic liver development would provide important insights into treatments and preventative strategies for pediatric liver disease. Furthermore, understanding the molecular mechanisms governing liver development will also be valuable for efforts to differentiate therapeutically useful hepatic tissue from stem cells in vitro.
B2 Overview of Embryonic Liver Development
Animal models have been indispensable for elucidating the molecular pathways governing development and how misregulation of these same pathways can lead to pathogenesis. Based on descriptive and experimental embryology in mouse and chicken, liver development can be conceptualized in a series of steps (Figure 1; Zaret, 2002) and from preliminary experiments these appear to be largely conserved in Xenopus (Section C; (Chalmers and Slack, 2000; Zorn and Mason, 2001). First the endoderm germ layer forms during gastrulation and eventually gives rise in the adult to the liver, pancreas, lung, thyroid and epithelial cells of the gastrointestinal and respiratory systems (Wells and Melton, 1999). The endoderm is then patterned into broad domains along the anterior-posterior axis and anterior endoderm (AE) becomes competent to give rise to foregut derivatives. Around the 4-8 somite stage of development a combination of positive inductive signals from the cardiogenic mesoderm and repressive signals from the trunk mesoderm specifies a group of cells in the ventral foregut endoderm to adopt a hepatic fate (Figure 2; (Fukuda-Taira, 1981; Gualdi et al., 1996; Le Douarin, 1975).
| Endoderm Patterning Hepatic Competence | Liver Induction And Specification | Liver bud Proliferation Morphogenesis | Hepatic Differentiation |
 |  |  |  |
 |  |  |
| gastrulation | 4-8 somite | gut tube | Liver |
| Figure 1 The top row depicts liver development in the mouse at four developmental stages; gastrula (fat left), 4-8 somite stage, isolated gut tubes at liver bud stage and adult liver (far right). The bottom row depicts the analogous steps in Xenopus, which based on our preliminary data are conserved. During gastrulation the endoderm (yellow) is specified and patterned so that the anterior endoderm (pink), acquires the competence to form foregut lineages. Between 4-8 somite stages a subset of the ventral foregut endoderm (arrows) is induced by the adjacent cardiogenic mesoderm (orange) to adopt a hepatic fate. The hepatic epithelium then invades the septum transversum mesenchyme and proliferates to form the liver bud (lb) and eventually differentiates and forms the mature liver. Anterior is to the left. |
The hepatic endoderm epithelium thickens, delaminates and invades the surrounding mesenchyme to form the liver bud. Continued epithelial/mesenchymal interactions stimulate cell proliferation and morphogenesis as the embryonic organ grows. The undifferentiated cells in the liver bud, known as hepatoblasts, are bi-potential, giving rise to hepatocytes and biliary epithelium. The adult liver also contains cells of mesodermal origin, but compared to many organs its cellular make up is relatively simple, with hepatocytes accounting for ~70% of the cells in the liver. The liver is therefore a relatively simple tissue in which to explore principles of organogenesis.
B3 Molecular pathways regulating early Liver Development.
 | Figure 2 A model of liver induction in the mouse. The foregut region of a 4-8-somite stage mouse embryo showing FGF and BMP signals from the CM (orange) induce the liver (green) in the adjacent foregut endoderm (yellow). Unknown signals from the trunk mesoderm (red) and ectoderm (white), that we hypothesize are Wnts, restrict hepatic development in the posterior endoderm. |
Promoter studies examining the transcriptional regulation of liver specific genes in tissue culture, such as transthyretin, alpha-fetoprotein and albumin have provided extensive molecular information on terminal hepatocyte differentiation (Costa et al., 2003a). At the other end of the developmental spectrum, work in Xenopus and zebrafish have begun to elucidate the molecular details of how the endoderm is formed in the first place (Stainier, 2002). Finally genetic analysis in the mouse has identified many of the genes, such as
Hex, Prox1, c-jun, Hlx, HGF, c-Met, Smad2/3 Kras, Sek1, CERB and
Xbp1 that are required for maintenance, proliferation, morphogenesis and survival of the liver bud once the hepatic lineage is specified (Zaret, 2002). The conspicuous gap in our knowledge is between gastrulation and liver specification when mammalian embryos are rather inaccessible, which is the focus of this proposal.
Classical transplantation experiments in avian embryos and in vitro mouse embryos explant cultures, indicate that the cardiogenic mesoderm (CM) induces liver development in the adjacent ventral foregut endoderm between the 4-8 somite stages of development (Fukuda-Taira, 1981; Gualdi et al., 1996; Le Douarin, 1975). Only the CM, which includes the precardiac and septum transversum mesenchyme, had this activity and mesoderm from other regions of the embryo could not stimulate hepatic development in foregut endoderm (Fukuda-Taira, 1981; Gualdi et al., 1996; Le Douarin, 1975). Furthermore, in vivo only the foregut endoderm was competent to initiate hepatogenesis. To date there is no known mutation in any organism that specifically blocks initial hepatic induction. However, Zaret and colleagues have shown that FGF and BMP growth factors, are good candidates for endogenous liver inducing factors (Zaret, 2002).
B4 Hepatic competence and Wnt signaling
| A. Fate Map | B. Wnt-antagonists | C. Wnts | D. β-catenin activity |
|---|
|  |  |  |
| | Cerberus
Dkk1
Frzb1
sFRPs | Wnt8
Wnt8b
Wnt3A | |
|  |  |  |
| Figure 3 Wnts and Wnt-antagonists. Xenopus (top row) and mouse (bottom row) gastrula embryos are shown. Hybridization of Xenopus and mouse gastrula sections with Cerberus probes (Zorn et al 1999; Biben et al 1998), exemplifies the 5-6 Wnt-antagonists expressed in the AE. In mouse, these antagonists are expressed in both the anterior visceral endoderm as well as the definitive anterior endoderm. C) In situ hybridization of Wnt8 to Xenopus and mouse gastrula exemplifies the Wnts expressed in the posterior mesoderm and ectoderm. D) The interaction of Wnts and Wnt-antagonists spatially regulates β-catenin activity. In the Xenopus high nuclear β-catenin levels (Schohl and Fagotto 2002), are depicted as dark blue and absent nuclear β-catenin, in the AE is white. The mouse embryo depicts the results from transgenic TOP:gal mice where the blue indicates high activity of the β-catenin/TCF transcriptional reporter in the posterior but not in the anterior. This reciprocal expression pattern of Wnts and Wnt-antagonists also persists during hepatic induction stages. |
In vivo transplantation experiments in chick found that only the foregut endoderm was "competent" to differentiate into liver and CM could not induce hepatic development in posterior endoderm (Fukuda-Taira, 1981; Le Douarin, 1975). However, mouse explant studies found that posterior endoderm could express liver markers once isolated from its adjacent dorsal/posterior mesoderm and cultured
in vitro. Furthermore, this dorsal mesoderm could repress hepatic induction in foregut co-cultures (Gualdi et al., 1996). This suggests that hepatic potential is normally repressed in the hindgut by dorsal mesodermal signals, while the anterior endoderm is somehow protected. As a result in the intact embryo only the foregut is "competent" to undergo hepatic development. We hypothesize that canonical Wnt signaling regulates hepatic competence. We propose a model where Wnt ligands secreted from the trunk mesoderm repress hepatic development in the posterior endoderm, while Wnt-antagonists expressed in the anterior endoderm make the developing foregut refractory to these repressive Wnts and competent to respond to hepatic induction.
Canonical Wnt signaling promotes the accumulation of nuclear β-catenin, interacts the TCF and SOX family of transcription factors to regulate the transcription of target genes (Wodarz and Nusse, 1998). Immunocytochemistry of β-catenin distribution in the Xenopus gastrula indicates that the AE, which expresses Wnt-antagonists and that fate maps to the foregut has very low levels of nuclear β-catenin as compared to the rest of the endoderm (Fig. 3; Schohl and Fagotto, 2002). Similarly a transcriptional reporter of β-catenin/TCF in transgenic mice (TOP: gal), indicates that the anterior endoderm (visceral and definitive) of the mouse gastrula is also devoid of Wnt signaling (Merrill et al 2004). Base on these observations, we propose a molecular mechanism where the β-catenin "free-zone" in the anterior endoderm promotes pre-hepatic gene expression (possibly FoxA2 and Gata4) and thus hepatic competence. In the rest of the endoderm, high levels of nuclear β-catenin would repress pre-hepatic gene expression, possibly by activating the transcription of genes that promote posterior fates.
Our approach: Xenopus embryos as a model for Liver Development
 |
| Figure 4 Fate map and time-line of Xenopus liver development. The endoderm (yellow) with the cells fated to give rise to the liver precursors (red) are shown in a 32-cell embryo (left), a gastrula, a 4-8 somite stage, a 1.5 day and an isolated gut tube (liver, red; pancreas, green; lungs, blue) of a three-day embryo. As development proceeds the tissue contributing to the liver is progressively restricted. For example, the red cells at the 32-cell stages, will contribute to the entire ventral foregut including lung and ventral pancreas precursors. The three blue lines below summarize the expression of molecular markers during the progressive developmental stages. |
The
Xenopus embryo is a powerful model for identifying and characterizing developmental pathways and has provided much of our understanding of vertebrate germ layer formation and axis specification.
Xenopus has been particularly useful in illuminating the function and regulation of Wnt, nodal and BMP signaling and this has proven relevant to how these same pathways are mis-regulated in human pathogenesis. The experimental advantages of
Xenopus include abundant, large and externally developing embryos in which gene products can be easily introduced by microinjection. Excellent endodermal fate maps exist from early cleavage, gastrula and neurula stages (Fig. 4) allowing us to identify the liver and foregut precursors as early as the 8-cell stage. Microinjection of these cells provides a simple and efficient means of exogenous gene expression in the liver precursors. The robust amphibian embryo has long been a favorite model for experimental embryologists allowing strait forward tissue transplantation and
in vitro embryonic tissue culture. In mice genetic studies of early organogenesis are often difficult because of early embryonic lethality leaving very little material to analyze. In contrast, maternal nutrients stored in the egg and oxygen from the environment allows
Xenopus embryos to survive gross abnormalities or the loss of organs such as the heart, providing ample experimental material to examine. The recent addition of rapid transgenics and tissue specific promoters now allows temporal and spatial control of gene expression during organogenesis.
We have described the histology and molecular details of early Xenopus liver development and our analysis indicates that liver bud formation and the final histological organization of the adult liver is very similar to mammals (Fig. 5).
 |
| Figure 5 Histology. (A) A 1.5 day embryo showing the endoderm (orange) and the liver region (red). (B) Midsagittal and (C) cross sections of an embryo hybridized with the liver enriched marker AMBP shows the thickened hepatic epithelium (Lv). (E) A 3 day embryo showing the gut tube (yellow) and liver bud (red). (F-I) H & E histological analysis of a 3 day embryo, (F-G) in midsagittal and (H-I) cross section illustrates the enlarged liver bud and hepatoblasts arranged in chords (I). (J) H & E stained section of an adult Xenopus liver shown a histological arrangement similar to mammals. |
Screens for novel genes involved in Liver Development.
To data only FGFs and BMPs have been implicated in liver specification, and the hepatogenic pathway they regulate is unknown. To identify additional hepatic regulators, functionally screen a embryonic foregut cDNA library, by microinjecting pools of synthetic mRNAs from the library into Xenopus embryos and identified pools that altered normal liver development. To date we have identified ~35 potentially new hepatic regulators.
 | Figure 6 Screening strategy to identify hepatic regulators. The ventral foregut (red square) contains the liver, ventral pancreas, lung and thyroid precursors, including endoderm, mesoderm and ectoderm was dissected from 2000 stage 20-22 Xenopus embryos. A plasmid cDNA library was constructed and arrayed in duplicate using a Q-bot colony-picking robot. Plasmid DNA and in vitro transcribed RNA was generated from each 96 well plate. RNA from each pool was injected into embryos, which were assayed by Hex in situ and morphology for changes in liver development. |
If you would like to join our research team, contact the Zorn Laboratory. The Zorn Lab is part of the Division of Developmental Biology at Cincinnati Children's Research Foundation. The lab is located in Location R (Research Foundation Building), Room 2509.
Division of Developmental Biology
Cincinnati Children's Research Foundation
Cincinnati, OH 45229-3039
E-mail Aaron.Zorn@chmcc.org
Phone 513-636-3770
Fax 513-636-4317
Interested in joining us as a student or a postdoc? Learn more about postdoctoral training and graduate student opportunities at Cincinnati Children's.
