Downloaded from genesdev.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press PERSPECTIVE BMPs on the road to hepatogenesis Stephen A. Duncan1 and Alistair J. Watt Department of Cell Biology, Neurobiology and Anatomy, Medical College of Wisconsin, Milwaukee, Wisconsin 53226, USA onset of hepatic development. However, the time-hon- Classical transplantation studies using chick and quail ored belief that hepatic induction is solely a matter of embryos demonstrated that the vertebrate liver paren- the heart has now been overturned by further work from chyma is derived from an outgrowth of the ventral fore- the Zaret lab (Rossi et al. 2001, this issue), showing that gut endoderm, referred to as the liver bud (LeDouarin bone morphogenic proteins (BMPs) secreted from sep- 1964, 1968, 1975). This process, illustrated in Figure 1, tum transversum mesenchyme are needed in concert initiates at around gestation day 8.25 (7–8 somites) in the with cardiac-derived FGFs to induce the ventral endo- mouse (Gualdi et al. 1996). Invagination of the foregut derm to adopt a hepatic fate. positions a ventral portion of foregut endoderm next to The septum transversum mesenchyme remains the developing cardiac mesoderm. Early on, it was pro- closely associated with the liver throughout develop- posed that specification of the hepatic diverticulum re- ment and ultimately contributes toward the diaphragm lied upon the close proximity of cardiac mesoderm. Sev- and epicardium. Following specification of the hepatic eral studies, the most recent by Gualdi et al. (1996), pro- lineage, hepatoblasts rapidly proliferate and migrate vided strong evidence to support this proposal. In away from the foregut into the adjacent septum trans- general, it was found that ventral endoderm from either versum mesenchyme (Fig. 1). Using chick/quail grafts mouse or chick was unable to adopt a hepatic fate, or and coculture experiments with mouse tissues, it was express characteristic hepatic mRNAs, if it were grown found that hepatoblasts require an undefined signal from in the absence of developing cardiac tissue (LeDouarin the septum transversum mesenchyme to complete their 1975; Houssaint 1980; Fukuda-Taira 1981; Gualdi et al. differentiation into mature hepatocytes (LeDouarin 1996). Moreover, this inductive capacity seemed to be 1968; Houssaint 1980). However, this signal to differen- restricted to cardiac mesoderm, because other noncar- tiate was not restricted to the septum transversum, be- diac mesoderms failed to induce hepatic development cause hepatic differentiation could also be induced by (Fukuda 1979; Fukuda-Taira 1981). culturing hepatic endoderm with heterologous mesen- chymes (Houssaint 1980). The possibility that septum Not so lonely heart transversum mesenchyme could also contribute toward the distinct act of specifying the hepatic lineage had not Although it was firmly established that the developing previously been considered, because it was believed that heart played an important inductive role during hepato- only cardiac tissue was required. However, Rossi et al. genesis, the molecular mechanisms underlying this ac- (2001) have now addressed this possibility by taking ad- tivity were defined only recently, largely through work vantage of a transgenic mouse in which expression of a carried out by the Zaret laboratory (Gualdi et al. 1996; LacZ reporter gene is controlled by the endogenous Jung et al. 1999). Jung et al. (1999) used a tissue explant Bmp4 transcriptional regulatory elements (Lawson et al. assay to demonstrate that fibroblast growth factors 1999). The LacZ expression pattern in this mouse mim- (FGFs) 1 or 2 could substitute for cardiac mesoderm in icked that of endogenous BMP4, including expression in inducing ventral endoderm to elicit a hepatogenic re- the septum transversum but not in the ventral foregut albumin sponse. In these experiments expression of endoderm. Importantly, the authors identified LacZ ex- mRNA, a marker of nascent hepatic cells, could be de- pressing septum transversum mesenchyme that closely tected within ventral endoderm that had been isolated apposed the ventral endoderm around the time of hepatic from embryos prior to specification of the liver (2–6 specification (8 somites). This implied, through guilt by somites) and cultured for 2 d either in the presence of association, that the septum transversum might act in a cardiac tissue or in the presence of FGF1 or FGF2. In juxtacrine fashion to influence hepatic specification. contrast, when prehepatic ventral endoderm was cul- Moreover, the fact that BMP4 was expressed in the sep- albumin tured alone it failed to express mRNA. These tum transversum raised the possibility that BMPs could data strongly supported a model wherein FGFs secreted mediate the proposed inductive action. from the developing heart were sufficient to induce the 1Corresponding author. BMPs and FGFs act in concert E-MAIL [email protected]; FAX (414) 456-6517. Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/ An essential role for BMP signaling was demonstrated gad.920601. directly by culturing the prehepatic ventral endoderm GENES & DEVELOPMENT 15:1879–1884 © 2001 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/01 $5.00; www.genesdev.org 1879 Downloaded from genesdev.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press Duncan and Watt Figure 1. FGFs and BMPs in liver devel- opment. Sagittal sections through em- bryos at 8.25 d postcoitum (dpc) (upper left) and 9.5 dpc (lower left), with each area undergoing hepatogenesis boxed and shown in detail on the right. By this stage (2–6somite stage), invagination of the foregut endoderm positions a portion of ventral endoderm (green) adjacent to the cardiac mesoderm (red) and the septum transversum mesenchyme (blue). At this stage, the entire endoderm expresses the transcription factor Hnf3, making the en- doderm competent to adopt a hepatic fate. BMP signaling from the septum transver- sum and FGF signaling from the cardiac mesoderm induce a portion of the ventral endoderm to initiate hepatogenesis (green), leaving the distal lip of endoderm to form the ventral pancreas (yellow). This results in the expression of transcription factors such as Hex and Gata4 necessary for subsequent developmental stages. By 8.5 dpc the hepatic endoderm thickens to form the liver bud. The continued action of BMPs and FGFs is required for the pro- liferation and outgrowth of the liver bud. By 9.5 dpc Hnf4␣, which is necessary to complete hepatocyte differentiation, is ex- pressed in the migrating hepatic cells. with cardiac and septum transversum tissue in the pres- This model is appealing on a number of levels. The ence of the Xenopus laevis noggin protein (Xnoggin). fact that two tissues must contribute signals that induce Xnoggin acts as a potent antagonist of BMP action, and hepatogenesis reinforces the dynamic, multidimensional its presence in the media of cultured explants was found nature of embryogenesis. It means, simply, that a tissue to inhibit the onset of albumin mRNA expression that is competent to follow a given fate must be posi- within the ventral endoderm relative to control cultures tioned in the right place at the right time to fulfill that lacking Xnoggin. The demonstration that BMPs were fate. In this regard, the identification of FGF and BMPs as necessary for hepatic specification raised the question of inducers of hepatic development define a mechanism why, in previous experiments, FGF alone was capable of whereby outgrowth of the liver is positioned along the inducing albumin mRNA expression in cultured ventral anterior–posterior of the ventral foregut. In addition to endoderm. The authors suggest that it is likely, in retro- deciding the position of liver development, it is likely spect, that the ventral endoderm isolates contained that the combined action of FGF and BMP signaling de- BMP-expressing septum transversum mesenchyme. In- fines the domain of endoderm that gives rise to the ven- deed, examination of cultured endoderm explants re- tral pancreas. Like the liver, ventral pancreas also origi- vealed mesenchymal cells that expressed the septum nates as a budding outgrowth of the ventral endoderm. transversum marker Mrg1 as well as BMP4. The authors Deutsch et al. (2001) recently demonstrated, again using also found that Xnoggin inhibited FGF-induced expres- explant cultures, that the liver and ventral pancreatic sion of albumin mRNA in cultured ventral endoderm. domains are nonoverlapping, that is, the ventral pan- This result was illuminating because it also demon- creas derives from a lip of endoderm that extends beyond strated that FGF alone was insufficient to induce the the immediate vicinity of the developing heart (Fig. 1). onset of a hepatic gene expression program within the Further experiments showed that the entire ventral en- ventral endoderm. Similarly, explants cultured in the doderm, including the portion fated to become liver, is presence of exogenous BMPs but in the absence of car- innately capable of adopting a pancreatic fate by default. diac tissue or FGF also failed to express albumin mRNA. Indeed, when ventral endoderm that would normally fol- These results prompt a model in which at least two sig- low a hepatic fate is cultured in the absence of cardiac naling pathways originating from different tissue mesoderm or FGF, it initiates a gene expression program sources, FGF (cardiac mesoderm) and BMP (septum that is characteristic of the
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages7 Page
-
File Size-