Mesoderm induction and development of the embryonic axis in amniotes.

The mechanisms controlling the formation of the embryonic axis, and specifically those that give rise to the mesoderm, have received renewed attention recently. In the frog, some of these mechanisms have begun to be elucidated, and several factors have been found to cause uncommitted ectoderm cells to differentiate into mesoderm. All of the factors identified to date are related either to fibroblast growth factor (FGF) or to transforming growth factor beta (TGF-beta). Do the mechanisms that generate the embryonic axis of amphibians also operate in chick and mouse embryos? Here I address how amphibian and amniote embryos might provide complementary pieces of a puzzle.     

The role of growth factors in the embryogenesis and differentiation of the eye.

The vertebrate eye is composed of a variety of tissues that, embryonically, have their derivation from surface ectoderm, neural ectoderm, neural crest, and mesodermal mesenchyme. During development, these different types of cells are subjected to complex processes of induction and suppressive interactions that bring about their final differentiation and arrangement in the fully formed eye. With the changing concept of ocular development, we present a new perspective on the control of morphogenesis at the cellular and molecular levels by growth factors that include fibroblast growth factors, epidermal growth factor, nerve growth factor, platelet-derived growth factor, transforming growth factors, mesodermal growth factors, transferrin, tumor necrosis factor, neuronotrophic factors, angiogenic factors, and antiangiogenic factors. Growth factors, especially transforming growth factor-beta, have a crucial role in directing the migration and developmental patterns of the cranial neural-crest cells that contribute extensively to the structures of the eye. Some growth factors also exert an effect on the developing ocular tissues by influencing the synthesis and degradation of the extracellular matrix. The mRNAs for the growth factors that are involved in the earliest aspects of the growth and differentiation of the fertilized egg are supplied from maternal sources until embryonic tissues are able to synthesize them. Subsequently, the developing eye tissues are exposed to both endogenous and exogenous growth factors that are derived from nonocular tissues as well as from embryonic fluids and the systemic circulation. The early interaction between the surface head ectoderm and the underlying chordamesoderm confers a lens-forming bias on the ectoderm; later, the optic vesicle elicits the final phase of determination and enhances differentiation by the lens. After the blood-ocular barrier is established, the internal milieu of the eye is controlled by the interactions among the intraocular tissues; only those growth factors that selectively cross the barrier or that are synthesized by the ocular tissues can influence further development and differentiation of the cells. An understanding of the tissue interactions that are regulated by growth factors could clarify the precise mechanism of normal and abnormal ocular development.     

Induction of mesodermal tissues by acidic and basic heparin binding growth factors

The inducing activity of two heparin binding growth factors HBGF-1 (prostate epithelial cell growth factor; acidic pI) and HBGF-2 (fibroblast growth factor; basic pI) from bovine brain has been tested on totipotent ectoderm from early amphibian (Xenopus laevis, Ambystoma mexicanum) embryos. Both factors induced, at high concentrations, mostly compact spheres surrounded by a non-epidermal epithelium. When the concentration or time of incubation was reduced, large muscle inductions frequently organized as somites were formed besides endothelial vesicles, mesenchyme and smaller areas of intestine-like epithelium. Further reduction of the concentrations or the time of incubation led to an increase in size and number of endothelium-lined vesicles and of mesenchyme, whereas the induction of muscle decreased. At still lower concentrations the overall rate of inductions decreased. The relationship of the growth factors to the vegetalizing factor from chicken embryos, dilution of which shows a similar shift in induced organs, is discussed. The present and previous experiments suggest that different mesodermal and endodermal tissues are induced by secondary interactions in which additional factors are involved. The induced organs derive from dorsal as well as from ventral mesoderm.     

The specification and growth factor inducibility of the pronephric glomus in Xenopus laevis

We report a study on the specification of the glomus, the filtration device of the amphibian pronephric kidney, using an explant culturing strategy in Xenopus laevis. Explants of presumptive pronephric mesoderm were dissected from embryos of mid-gastrula to swimming tadpole stages. These explants were cultured within ectodermal wraps and analysed by RT-PCR for the presence of the Wilm's Tumour-1 gene, xWT1, a marker specific for the glomus at the stages analysed, together with other mesodermal markers. We show that the glomus is specified at stage 12.5, the same stage at which pronephric tubules are specified. We have previously shown that pronephric duct is specified somewhat later, at stage 14. Furthermore, we have analysed the growth factor inducibility of the glomus in the presence or absence of retinoic acid (RA) by RT-PCR. We define for the first time the conditions under which these growth factors induce glomus tissue in animal cap tissue. Activin together with high concentrations of RA can induce glomus tissue from animal cap ectoderm. Unlike the pronephric tubules, the glomus can also be induced by FGF and RA.     

Growth-factor-related proteins that are inducers in early amphibian development may mediate similar steps in amniote (bird) embryogenesis

Xenopus and murine activin A homologues (XTC-MIF and WEHI-MIF) and Xenopus and bovine basic fibroblast growth factor (bFGFs) are potent inducers of mesodermal and endodermal pathways of development in amphibian blastular animal cap cells. Porcine transforming growth factor beta 2 (TGF beta 2) is a weaker inducer in the same assay but human platelet-derived growth factor (PDGF) is inactive. We have assayed these factors for evidence of homologous effects in bird development. Unlike amphibians, bird embryos never exhibit a clean segregation of a cell layer that has a uniform specification when uninduced, and can be cultured in isolation as an assay after exposure to soluble factors. We have therefore performed less direct experiments, of three types. We have briefly cultured early chick epiblast cells with and without factors and then assayed their capacity to attach and spread upon fibronectin, in comparison with young streak and substreak hypoblast cells. We have asked whether similar microculture with factors alters the ability of quail epiblast cells to disrupt morphogenesis, and to integrate into the structure, of host chick blastoderms into which they are seeded. Finally, whole early chick blastoderms have been preincubated with or without factors for a brief period before setting them up to develop in vitro under circumstances usually permitting successful formation of axial pattern. Strong effects of the activin-like factors, of bFGF and of TGF beta 2 were seen in all three procedures, while PDGF was essentially inactive. In epiblast cells, effective factors at picomolar concentrations induced stable spreading upon fibronectin, and a capacity to adhere and spread upon basal epiblast surface and prevent morphogenesis in host blastoderms. Preincubation of whole early blastoderms with these factors led to characteristic deviation from normal development over the subsequent 24 h. We therefore suggest that peptides from the particular families that are active as inducers in amphibian blastula ectoderm may mediate homologous or closely related steps in respecification throughout vertebrates.     

EFFECTS OF TWO PROTEASE INHIBITORS ON SEA URCHIN AND AMPHIBIAN EGG DEVELOPMENT*

After a review of our present knowledge about the effects of proteases and protease inhibitors on cell growth and egg fertilization, the results of experiments where sea urchin and amphibian eggs were treated with two protease inhibitors (TPCK, TLCK) are described. Cleavage was hardly affected, but gastrulation quickly stopped or was incomplete. The morphogenetic abnormalities which follow can be explained by abnormal gastrulation and other factors: persistence of remnants of the fertilization membrane in sea urchin larvae and dissociation of the ectoderm cells in Xenopus embryos.     

Extraembryonic proteases regulate Nodal signalling during gastrulation

During gastrulation, a cascade of inductive tissue interactions converts pre-existing polarity in the mammalian embryo into antero-posterior pattern. This process is triggered by Nodal, a protein related to transforming growth factor-beta (TFG-beta) that is expressed in the epiblast and visceral endoderm, and its co-receptor Cripto, which is induced downstream of Nodal. Here we show that the proprotein convertases Spc1 and Spc4 (also known as Furin and Pace4, respectively) are expressed in adjacent extraembryonic ectoderm. They stimulate Nodal maturation after its secretion and are required in vivo for Nodal signalling. Embryo explants deprived of extraembryonic ectoderm phenocopy Spc1(-/-); Spc4(-/-) double mutants in that endogenous Nodal fails to induce Cripto. But recombinant mature Nodal, unlike uncleaved precursor, can efficiently rescue Cripto expression. Cripto is also expressed in explants treated with bone morphogenetic protein 4 (BMP4). This indicates that Nodal may induce Cripto through both a signalling pathway in the embryo and induction of Bmp4 in the extraembryonic ectoderm. A lack of Spc1 and Spc4 affects both pathways because these proteases also stimulate induction of Bmp4.     

Primary structure requirements for Xenopus nodal-related 3 and a comparison with regions required by Xenopus nodal-related 2

Transforming growth factor-beta superfamily members play important roles in the early development of animals. Activin and the Xenopus nodal related proteins 1, 2, and 4 induce muscle actin from Xenopus ectodermal explants, whereas the bone morphogenetic proteins 4 and 7 induce ectoderm to differentiate as epidermis. Bone morphogenetic proteins are antagonized by soluble binding proteins such as noggin and chordin, which leads to expression of neural cell adhesion molecule in animal caps. The transforming growth factor-beta superfamily member Xenopus nodal-related 3 also induces the neural cell adhesion molecule through inhibition of bone morphogenetic proteins. Therefore, whereas Xenopus nodal-related 2 and 3 share a high amount of sequence homology, they lead to very different cell fates. This study investigates the functional domains that distinguish the activities of these two factors. It was found that mutually exclusive regions of nodal-related 2 and 3 were required for activity. The central region of the mature domain is required for nodal-related 2 to induce muscle actin, whereas the N- and C-terminal ends of the mature domain are required for nodal-related 3 to induce neural cell adhesion molecule. These results help to define the minimal domains required for the unique activities of these factors.     

Peptide growth factors and their interactions during chondrogenesis

Peptide growth factors have been implicated in three aspects of cartilage growth and metabolism; the induction of mesoderm and differentiation of a cartilaginous skeleton in the early embryo, the growth and differentiation of chondrocytes within the epiphyseal growth plates leading to endochondral calcification, and the processes of articular cartilage damage and repair. Three peptide growth factor classes have been strongly implicated in these processes, the fibroblast growth factor family (FGF), the insulin-like growth factors (IGFs) including insulin, and transforming growth factor-β (TGF-β) and related molecules. Each of these peptide groups are expressed in the early embryo. Basic FGF, TGF-β and the related activin have been shown to induce the appearance of mesoderm from primitive neuroectoderm. TGF-β and related bone morphometric proteins can induce the differentiation of cartilage from primitive mesenchyme, and together with basic FGF and IGFs promote cartilage growth. Each class of growth factor is expressed within the epiphyseal growth plate where their autocrine/paracrine interactions regulate the rate of chondrocyte proliferation, matrix protein synthesis and terminal differentiation and mineralization. Basic FGF may prove useful in articular cartilage repair, while basic FGF, IGFs and TGF-β are among a number of growth factors and cytokines that have been implicated in cartilage disease.     

Polypeptide growth factors in embryogenesis and tumor growth

Polypeptide growth factors, which belong to different families (epidermal growth factors, insulin-like growth factors, fibroblast growth factors, transforming growth factors-beta, and some others), were characterized regarding their specific role in embryogenesis and tumor growth. Differences and parallels of the functioning of growth factors in these processes have been noted. Potential significance of the described characteristics of growth factors for directed modulation of embryogenesis and tumor growth is discussed.     

Neural determination in Xenopus laevis embryos: control of early neural gene expression by calcium

In amphibian embryos the central nervous system derives from the dorsal region of the ectoderm. Molecular studies led to the formulation of the "neural default model" in which neural development is under the inhibitory control of members of the BMP family. These growth factors also act as epidermis inducers. The neural fate is revealed by factors secreted by the Spemann Organizer such as noggin, chordin, follistatin, Xnr3 and cerberus which act by blocking BMP signalling. We propose a new model for neural cell determination in which a signalling pathway controlled by an increase in intracellular calcium suppresses the epidermis fate and activates the neural fate instead. This increase in calcium is due to an influx through calcium channels of the L-type, expressed in ectodermal cells during gastrulation. The possible involvement of a calcium-dependent phosphatase (calcineurin) to inhibit the epidermis fate and a calcium-calmodulin kinase (CaMkinase II) which activates the neural fate is discussed.     

Mesoderm-inducing factors: a small class of molecules

Mesoderm-inducing factors (MIF's) from chick embryos, XTC cells and WEHI-3 cells were studied using various procedures. The object was to find whether they are similar to heparin-binding growth factors (HBGFs-the only known pure mesoderm-inducing substances) and, if not, whether they are similar to each other. The major active components from all three MIF sources behave as somewhat hydrophobic, acid-stable molecules and do not bind to heparin. They all have relative molecular masses of about 13,000 measured by HPLC size exclusion chromatography. The isoelectric points measured by chromatofocusing were 6.7 (WEHI) and 7.7 (XTC). The chick MIF seemed somewhat heterogeneous by chromatofocusing and a portion of its activity bound to heparin sepharose. All three MIFs have similar effects on explants of Xenopus blastula ectoderm to the heparin-binding growth factors, causing an elongation at the time of gastrulation followed by the development of mesenchyme, mesothelium and muscle cells, the proportion of muscle increasing with dose. Unlike the HBGFs they all also induce notochord if sufficiently high concentrations are used. Our study shows that the MIFs examined here form a small group of potent agents distinct from the HBGFs and from other known growth and differentiations factors. Their occurrence in various tissues and cell lines suggests that they have functions in the adult organism as well as during early development.     

A re-examination of lens induction in chicken embryos: in vitro studies of early tissue interactions

Early studies on lens induction suggested that the optic vesicle, the precursor of the retina, was the primary inducer of the lens; however, more recent experiments with amphibians establish an important role for earlier inductive interactions between anterior neural plate and adjacent presumptive lens ectoderm in lens formation. We report here experiments assessing key inductive interactions in chicken embryos to see if features of amphibian systems are conserved in birds. We first examined the issue of specification of head ectoderm for a lens fate. A large region of head ectoderm, in addition to the presumptive lens ectoderm, is specified for a lens fate before the time of neural tube closure, well before the optic vesicle first contacts the presumptive lens ectoderm. This positive lens response was observed in cultures grown in a wide range of culture media. We also tested whether the optic vesicle can induce lenses in recombinant cultures with ectoderm and find that, at least with the ectodermal tissues we examined, it generally cannot induce a lens response. Finally, we addressed how lens potential is suppressed in non-lens head ectoderm and show an inhibitory role for head mesenchyme. This mesenchyme is infiltrated by neural crest cells in most regions of the head. Taken together, these results suggest that, as in amphibians, the optic vesicle cannot be solely responsible for lens induction in chicken embryos; other tissue interactions must send early signals required for lens specification, while inhibitory interactions from mesenchyme suppress lens-forming ability outside of the lens area.     

Differential responsiveness of myc- and ras-transfected cells to growth factors: selective stimulation of myc-transfected cells by epidermal growth factor.

To identify functional relationships between oncogenes and growth factors, we compared the effects of transfected myc and ras oncogenes on the responsiveness of Fischer rat 3T3 cells to three growth factors: epidermal growth factor (EGF), platelet-derived growth factor (PDGF), and transforming growth factor-beta (TGF-beta). Control cells did not grow in soft agar under any conditions. ras-Transfected cells grew in soft agar under all conditions tested and were insensitive to the stimulatory effects of exogenous growth factors. These cells secreted elevated levels of both EGF-like factors and TGF-beta, suggesting that the lack of responsiveness of these cells to exogenous growth factors arose from autocrine stimulation. myc-Transfected cells displayed conditional anchorage-independent growth: they formed numerous colonies in soft agar in the presence of EGF but relatively few colonies in the presence of PDGF or TGF-beta. Secretion of EGF-like factors and TGF-beta by these cells was not elevated above that of control cells. These results suggest a model for the mechanism of cooperation between myc and ras oncogenes in which ras-like genes induce growth factor production, while myc-like genes increase the responsiveness of cells to these factors.     

Behavior of transforming growth factors in serum-free media: An improved assay for transforming growth factors

Summary Transforming growth factors (TGFs) are a relatively new category of factors that induce the anchorage-independent growth of non-transformed cells. These factors are usually detected by their ability to induce normal rat kidney (NRK) fibroblasts to grow in soft agar. Until now, this assay has been performed in serum-containing medium (SCM). Unfortunately, the background activity of this assay is variable and dependent on several factors, including passage number of the cells and the serum lot used. Furthermore, the addition of either EGF or TGF-β alone results in the appearance of additional colonies, which decreases the sensitivity of the assay. To circumvent these problems, serum-free media have been developed that support the growth of the NRK cells at low density in both monolayer culture and soft agar. Long-term growth in monolayer cultures occurs in serum-free medium supplemented with laminin, insulin, transferrin, epidermal growth factor (EGF), fibroblast growth factor (FGF) and high density lipoprotein (HDL). Growth in soft agar occurs when TGFs are added to a serum-free medium, AIG medium, that contains insulin, transferrin, FGF and HDL. In contrast to the background activity observed when the assay is performed in SCM, no colonies form in the AIG medium unless TGFs are added and few, if any, colonies form if EGF or TGF-β are added alone. Thus, the AIG medium provides an improved assay for TGFs. In addition, the AIG medium should prove useful for examining other factors, including serum factors, for TGF activity. Editor's Statement This communication describes a modification of the standard assay for transforming growth factors. The techniques employed make use of advantages provided by recent advances in serum-free cell culture to provide a well-defined detection system that is more sensitive than conventional procedures. Experimental approaches described in this article also should be helpful in unraveling differences in cellular behavior encountered under anchorage-dependent vs. anchorage-independent conditions. D. W. Barnes