Primary Embryonic Induction Research Articles

Reprint of: Prelude to molecularization: The double gradient model of Sulo Toivonen and Lauri Saxén

The present molecular investigations of Organizer phenomena show a remarkable connection to the earlier classical embryological studies that used transplantation as a method for making mechanistic models of induction. One of the most prominent of these connections is the dual gradient model for anterior-posterior and dorsal-ventral polarity. This paper will discuss some of the history of how transplantation experiments provided data that could be interpreted in terms of two gradients of biologically active materials. It will highlight how the attempts to discover the elusive Induktionsstoffen gave rise to the double gradient model of Sulo Toivonen and Lauri Saxén in the 1950s and 1960s. This paper will also document how this research into the identity of these molecules gave rise to the developmental genetics that eventually would find the molecules responsible for primary embryonic induction.

Prelude to molecularization: The double gradient model of Sulo Toivonen and Lauri Saxén

The present molecular investigations of Organizer phenomena show a remarkable connection to the earlier classical embryological studies that used transplantation as a method for making mechanistic models of induction. One of the most prominent of these connections is the dual gradient model for anterior-posterior and dorsal-ventral polarity. This paper will discuss some of the history of how transplantation experiments provided data that could be interpreted in terms of two gradients of biologically active materials. It will highlight how the attempts to discover the elusive Induktionsstoffen gave rise to the double gradient model of Sulo Toivonen and Lauri Saxén in the 1950s and 1960s. This paper will also document how this research into the identity of these molecules gave rise to the developmental genetics that eventually would find the molecules responsible for primary embryonic induction.

Continuity and change: paradigm shifts in neural induction.

The problem of "primary embryonic induction" was one of the first areas of developmental biology to become "molecularized." What had been seen as an intractable series of problems became amenable to the techniques of Northern blotting, ectopic RNA insertion, and in situ hybridization. These molecular analyses showed that some of the fundamental concepts of primary embryonic induction concluded by experimental embryologists were false. First, primary embryonic induction was not primary. The organizer tissue, itself, was the product of a prior induction. Second, the neural fate of cells was not being induced. Rather, the epidermal fate was induced and the neural state was the default, uninduced, fate of ectodermal tissues. Third, primary embryonic induction was not something unique to vertebrates. Rather, the ventral neural cord of insects formed using the same mechanisms as the dorsal neural tube of vertebrates. Fourth, the brain formed in a matter distinctly different from that the spinal cord. Despite these differences, there has been a clear and strong continuity between the experimental embryological tradition and the molecular genetic tradition, and these new results are seen by many contemporary developmental geneticists as strengthening, rather than destroying, the older science.

In memoriam Lauri Saxen (1927-2005)

Dr. Lauri "Tupu" Saxen died last October (2005) at the age of 78. He was a physician, a scientist, a photographer, a naturalist, a great story-teller and a man who enjoyed science enormously. His name has become synonymous with the Finnish school of Developmental Biology, a school that focuses on reciprocal inductive interactions during vertebrate organogenesis. But many biologists probably don't know the full extent of his importance to the field. A few years ago, I had the occasion to outline some of his contributions which are so varied and important that it is difficult to believe that they are the work of one person, and I have included them in this brief eulogy. One could divide his scientific contributions into five categories: (1) the threshold hypothesis of amphibian metamorphosis; (2) the double-gradient hypothesis of primary embryonic induction; (3) the analysis of reciprocal induction during kidney development; (4) the integration of developmental biology with epidemiology and (5) the maintenance of a national infrastructure for science.

Wnts as kidney tubule inducing factors.

Since the discovery that inductive tissue interactions regulate nephrogenesis, one of the aims has been to identify the molecules that mediate this induction. The small size of embryonic tissue has limited the possibilities to identify the inducers biochemically, even though such efforts were directed to study, e.g. neural induction (for a comprehensive review, Saxén and Toivonen, Primary embryonic induction, Academic Press, London, 1962). The rapid progress in molecular biology made it possible to identify genes from minute amounts of tissue and provided techniques to generate recombinant proteins to assay their action in classic experimental systems. This led to the identification of some signals that are involved in primary and secondary inductive interactions during embryogenesis. Here, we will review evidence suggesting that secreted signaling molecules from the Wnt gene family mediate kidney tubule induction.

Naturally occurring diblastodermic eggs in the annual fish Cynolebias: Implications for developmental regulation and determination.

Annual fish development differs from that of other teleosts because a phase of blastomere dispersion-reaggregation spatially and temporally separates epiboly from embryogenesis. The fate of dispersed blastomeres was assessed in diblastodermic eggs of the annual fishes Cynolebias whitei and C. nigripinnis. In typical teleosts, blastomere determination and the events of primary embryonic induction occur prior to or during epiboly, so diblastodermic eggs produce partially or completely duplicated embryos. In the diblastodermic eggs of Cynolebias, the two blastoderms are completely separate from the one cell stage to the high blastula. Blastoderm fusion begins during midepiboly. By the end of epiboly, blastoderm fusion has been completed, and the deep, embryo-forming blastomeres of both blastoderms have completely dispersed and intermingled to form a single cell population. A typical annual fish dispersed blastomere phase ensues. Blastomeres reaggregate into a single mass, in which one embryo develops. When hatched, the young fish have no obvious structural or functional abnormalities. We suggest that the dispersed blastomeres of annual fish eggs are equivalent and that induction or determination takes place within the reaggregate. Alternatively, dispersed cells are partially determined but highly regulative, so that, when two populations fuse, the cells sort out according to tissue type and form a single embryo. In either instance, the formation of a single, normal embryo seems to corroborate the hypothesis that the dispersed cell phase of annual fishes is an adaptation that prevents environmentally induced developmental defects. © 1993 Wiley-Liss, Inc.

Patterning of connective tissues in the head: discussion report

ABSTRACT The three papers presented by Noden, Thorogood and Lumsden in this session encompassed the connective tissues as broadly defined, i.e. soft (fibrous) connective tissue, cartilage, bone, muscle and the dental tissues, enamel and dentine, and utilized a variety of experimental techniques on both avian and mammalian embryos to explore specificity and patterning of the vertebrate head. Whether similar developmental processes pattern homologous structures in different Vertebrate classes (Amphibia, Aves, Mammalia) was discussed with reference to patterning of the cranial musculature, chondrocranium and dental tissues. A number of challenging ideas emerged during this session. Does the premigratory neural crest consist of a homogenous population of totipotent cells or of subpopulations of bi- or tripotential cells? Is fundamental patterning of the head an early embryonic event, perhaps specified during primary embryonic induction or the consequence of neuroepithelial folding, brain growth, inductive interactions and/or spatially and temporally distributed extracellular matrix products? Can the fact that mesoderm and angioblasts do not display distinctive patterning that relates to their particular embryologic origins be extrapolated to patterning in general? How does the documentation of an ondontogenic trunk neural crest in mammals affect our theories of how patterning mechanisms arose or were modified during vertebrate evolution?

Early Reciprocal Interactions between Ectoderm and Chordamesoderm: Statistical Evaluation of Classical Embryological Experiments: (Ambystoma mexicanum/primary embryonic induction/ectoderm-chordamesoderm interactions/descriptive statistics).

Embryos of the urodele Ambystoma mexicanum were used in time set experiments for the analysis of ectoderm-chrodamesoderm interactions in "primary embryonic induction". The influence of chordamesoderm-age (CM-age), ectoderm-age (CM-age), ectoderm-mass (E-mass) on differentiation processes were investigated. The results revealed that early interactions exist, which, however, seem to be dominated by the ectoderm in various respect: a) the segregations of the CNS is due to the autonomous change in ectoderm competence phases; b) based on the temporal sequence of competence phases, influences are exerted by the ecto-neuroderm which support the chordamesoderm in its self-differentiation capability to chorda and other mesodermal tissues. Consequently, reciprocal impulses from these newly differentiated tissues are warranted (feedback to ectoderm); c) co-ordination of the cranialization effect and corresponding ectodermal competence phase is achieved; d) the inductive efficiency of the chordamesoderm is strongly supported. Three statistical tests (regression, correlation, information analyses) were conducted (dependant variables: quality and quantity of the differentiated structures; independant variables: experimental factors). All three tests equally rendered statistical significance for the following order of decreasing importance the experimental factors play in induction and differentiation processes: E-age, E-mass, CM-age.

Alteration of cell adhesion system in amphibian ectoderm cells during primary embryonic induction: changes in reaggregation pattern of induced neurectoderm cells and ultrastructural features of the reaggregate.

Cell adhesion was studied during primary embryonic induction. The disaggregation rate and reaggregation patterns were analysed in the ectoderm cells of various developing Cynopus gastrulae and neurulae. The neurectoderm cells disaggregated more slowly with gastrulation, and the neural plate cells of early neurula showed a lesser capacity for disaggregation. Although no differences in reaggregation were found between dorsal and ventral ectoderm at the early gastrula stage, there were significant differences between the induced neurectoderm and the non-induced ventral epidermal cells at the late gastrula stage. Neural plate cells of the early neurula stage were seen to form a chain-like reaggregate, but the ventral epidermal cells of the same embryo formed a cluster-like spherical reaggregate. Scanning electron microscope observations of reaggregates also showed significant differences in adhesive properties between induced neurectoderm and non-induced epidermal cells. The adhesion field of the induced neurectoderm cells was smooth, differing from the distinct ridges of the non-induced epidermal cells. These results suggest that changes in the cell adhesion system, resulting in the formation of a columnar cell shape, may occur immediately after a neural-inducing action.

Cell-to-cell contact in primary embryonic induction: effects of lectin on electrical coupling and neural induction

The effects of lectin (concanavalin A; ConA) on the electrical coupling between inducing chorda-mesoderm and reacting ectoderm cells, and the realization of neural induction were investigated. The electrical coupling between cells of the chorda-mesoderm of the late gastrula (stage 13b) and the competent ectoderm or Con-A-treated ectoderm of the early gastrula (stage 12a) was measured. Neural induction was tested with ectoderm explants which had been combined with the inducing chorda-mesoderm for 1, 3 and 6 h. Electrical coupling was observed after 3 h. By 6 h, the coupling ratio had recovered to the same level as that between the homogeneous germ-layer cells. However, the electrical coupling did not recover in the combinant with Con-A-treated ectoderm. This suggests that Con-A disturbs close cell contact between the ectoderm and chorda-mesoderm cells. Neural induction was realized in the ectoderm which was combined with chorda-mesoderm for more than 3 h; this occurred parallel to the recovery of electrical coupling. In contrast, Con-A treatment (50 μg/ml) of the competent ectoderm for 30 min prevented neural induction. After 3 h of contact, the neural induction of Con-A-treated ectoderm was only one-third of that of the control ectoderm. The present study suggests that cellular contact between the inducing mesoderm and the ectoderm target cells plays an important role in the realization of neural induction.

Germ Layer Interactions in Pattern Formation of Amphibian Mesoderm during Primary Embryonic Induction: (germ layer interaction/mesoderm formation/embryonic induction/amphibian (Cynops) axis).

Mesodermal differentiation of dorsal marginal zone (DMZ) before and after invagination was analyzed by a series of combination experiments with different kinds of ectoderm. Lower DMZ of early gastrula didn't show any axial-mesoderm (notochord and somitic mesoderm) but lateral mesoderm (mesenchyme, mesothelium, or blood cells) in combinant with non-competent ventral ectoderm, while combinant with competent ectoderm was found to have well-differentiated axial-mesoderm with deutero-spinocaudal neurals. The axial-mesoderms have origin in the ectoderm. Uninvaginated DMZ of middle gastrula also showed difference in mesodermal differentiation between competent and non-competent ectoderms; axial-mesoderm differentiation was much better in competent than in non-competent. The axial-mesoderm originated from the uninvaginated DMZ. Archenteron roof of late gastrula showed regional difference in mesodermal differentiation in both combinants with competent and non-competent. The present study further demonstrated that there was regionality in promoting effect of induced neurectoderm on axial-mesoderm differentiation of invaginated archenteron roof. The present experiments suggest that the cranio-caudal and dorso-ventral axis formations of amphibian mesoderm are finally determined by sequential and reciprocal interactions between the mesodermal anlage and the overlying ectoderm. It should be also shown that lower DMZ acts to trigger a series of the sequential interactions during primary embryonic induction.

Cellular Basis of Neuralization of Induced Neurectoderm in Amphibian Embryogenesis: Changes of Cell Shape, Cell Size, and Cytodifferentiation of the Neurectoderm after Neural Induction: (neuralization/primary embryonic induction/cellular changes/amphibia (Cynops)).

Cellular alterations of the neurectoderm after primary embryonic induction were examined by measuring several indices of shape, volume, and cytodifferentiation of the neurectodermal cells of Cynops embryos during gastrulation and early neurulation. Results showed that cellular alterations occurs just after the 18 hr embryo stage (stage 13b). The thickness of the neurectoderm layer decreases like that of the epidermal ectoderm during early and middle gastrulation. After the 18 hr embryo stage, however, the neurectoderm thickens, mainly due to formation of columnar cells. Measurement of cell volume indicates that the neurectoderm of the early and middle gastrulae consists of a cell population of heterogeneous size. The heterogeneity diminishes sharply after the 18 hr embryo stage and the neural plate of the 36 hr embryo (stage 18) consists of cells of homogeneous size. Stages before the 12 hr embryo (stage 12b) and after the 18 hr embryo (stage 13b) also showed differed in cell adhesion to the culture flask and in cytodifferentiating potency. Single cells dissociated from the neurectoderm of 18 hr embryos could adhere to the substratum and differentiate into both nerve-like cells and pigment cells. Both capacities increase during further development. These results are discussed in relation to the neuralizing determination of neurectoderm after primary embryonic induction.

Change in the differentiation pattern ofXenopus laevis ectoderm by variation of the incubation time and concentration of vegetalizing factor.

Early amphibian gastrula ectoderm (Xenopus laevis) has been treated with vegetalizing factor using the sandwich technique, varying the period of incubation and the inducer concentration.The pattern of induced tissues depends on three factors: the inducer concentration, the size of inducer pellet and the time of exposure of ectodermal target cells to inducer.Short treatment with inducer will result in the formation of blood cells and heart structures. An increase in incubation time or inducer concentration, or both, will cause the formation of increasing amounts of such dorsal mesodermal structures as pronephros, somites and notochord. Neural structures can only be observed in explants with considerable amounts of somites and notochord.Ectoderm treated with high concentrations of vegetalizing factor for the whole period of competence will differentiate into endoderm.Furthermore, the results show thatX. laevis ectoderm does not show any autoneuralizing tendency under our experimental conditions. It therefore seems to be a suitable tool for the study of primary embryonic induction.

A Rapid Experimental Method to Study Primary Embryonic Induction

The technique described here is a combination of the sandwich-method and cell culture. It allows one to know quickly whether induction occurred or not (48 h if cell spreading is considered, 4-5 days if morphological differentiation is observed). The dissociation of the explants soon after induction does not disturb their inductive pattern. Moreover, this method allows a daily observation of the morphological events and study at the cellular and/or molecular level.

Changes in lectin-mediated agglutinability during primary embryonic induction in the amphibian embryo

The present study was undertaken to investigate structural alterations at the surface of presumptive neural cells after primary embryonic induction. For this purpose, plant lectin-mediated agglutinability of dissociated cells from the epiblast of Bufo arenarum gastrulae was tested. Two fragments of epiblast were excised from the same mid-gastrula: one from the dorsal side of the egg, making contact with the invaginating chordamesoblast and assumed to be composed of determined cells and the other from the ventral region of the egg, facing the blastocoele cavity and assumed to be composed of undetermined cells. Cells of the pooled fragments were dissociated in calcium-free Holtfreter's solution with potassium oxalate and incubated in the presence of different concentrations of phytohemagglutinin and concanavalin A. Epiblast cells overlying the archenteron roof are less agglutinated with both lectins than undetermined cells. On the other hand, when egg fragments were removed from the dorsal and ventral regions of early gastrulae before the archenteron was formed, no significant difference in lectin-mediated agglutinability was observed, even after having been cultured in vitro in absence of inducing tissue. These results suggest that the target of the inducing signal generated in the mesoblast is likely to be located on the surface of epiblast cells. Additional experiments showed that cells pretreated with colchicine, cytochalasin B or colchicine and cytochalasin B simultaneously exhibit no significant variation in agglutinability, suggesting that the cytoskeleton was not be involved in the cell surface alteration here described. Treatment of whole embryos or sandwich explants with concanavalin A or phytohemagglutinin has no effect on neural tube formation, suggesting that the carbohydrate-containing binding sites for these lectins are not involved in primary embryonic induction. Changes in cell agglutinability described in this paper are to be interpreted thus as a secondary expression of structural alterations in the cell surface concomitant with neural determination.

Cell Contacts Between Chorda-Mesoderm and the Overlaying Neuroectoderm (Presumptive Central Nervous System) During the Period of Primary Embryonic Induction in Amphibians

Using transmission and scanning electron microscopy we were able to show that during primary embryonic induction in amphibians (Triturus alpestris) the interspace between the inducing chorda-mesoderm and the reacting ectoderm (presumptive medullary plate) of mid-gastrula stages is traversed by cell projections starting from cells of both tissue layers. In addition intimate membrane contacts between the main bodies of the ectodermal and chorda-mesodermal cells could be observed. It could be ruled out that cytoplasmic bridges (anastomosis) exist between cells of inducing chorda-mesoderm and reacting ectoderm, which would allow a free transfer of inducing substances without passing through membranes, as Eakin and Lehmann [1] have postulated. The possible role of cell to cell contact for neural induction is emphasized.

Effects of Non-histone Chromosomal Proteins on Primary Embryonic Induction and Cytodifferentiation in Salamanders

The effects of non-histone chromosomal proteins from adult liver on neural induction and differentiation of induced derivatives, were studied in the Salamanders Ambystoma mexicanum and Pleurodeles waltlii. The treatment of ectoderm or/and blastoporal lip prior to their association, shows that these proteins do not alter the competence of the reacting tissue (ectoderm) or the inductive activity of the inducing tissue (blastoporal lip). Very strong inhibitory effects on cytodifferentiation can be brought about by continuous treatment, but only with homospecific proteins. In contrast, non histone proteins extracted from the liver of other animal species have no effect. Species-specific inhibitory effects of non-histone chromosomal proteins extracted from an adult tissue, on the differentiation of another embryonic tissue are thus demonstrated. They are compared with those previously studied in cultures of differentiating embryonic cells from Pleurodeles waltlii and Ambystoma mexicanum or on the morphogenesis of these embryos.

Fish swimbladder: an excellent mesodermal inductor in primary embryonic induction

Swimbladder of the crucian carp, Carassius auratus, was found to be better as a vegatalizing tissue than other tissues, such as guinea-pig bone marrow, when presumptive ectoderm of Triturus gastrulae was used as reacting tissue. Swimbladder usually induced assemblies of highly organized mesodermal tissues, such as notochord, somites and pronephric tubules, some of which were covered by mesodermal epithelium without any epidermal covering. A special character of the effect of swimbladder was the rather frequent induction of solid balls of undifferentiated cells, which were identified as mesodermal or mesodermal and probably endodermal. These findings show that swimbladder has a strong and fast spreading vegetalizing effect on the responding presumptive ectoderm.

Mechanisms of Cell Interaction during Primary Embryonic Induction Studied in Transfilter Experiments

The transmission mechanisms operative at different stages of neutralisation during primary embryonic induction of the newt Triturus vulgaris were studied in experiments employing Nuclepore filters placed between interactive tissue explants. The transmission time of the neuralising effect was determined with 0.2 mum Nuclepore filter. In another series of experiments the transformation of neuralised ectoderm by archenteron roof mesoderm into other parts of the CNS was studied. Although sufficiently long induction times were used no transformation into hindbrain structures could be induced across filters with pore sizes from 0.1 mum to 1.0 mum. However, electron microscopy demonstrated cytoplasmic penetration into 0.6 mum filters at 15 h of induction. The results speak against free long-range diffusion of inductive material at the stage of transformation of the neuralised ectoderm to more caudal parts of CSN and warrant a more detailed structural study of the transmission phenomenon in question.

Transfllter Studies on Neural Induction in the Newt

In order to study the transmission mechanism of neuralising signals during primary embryonic induction, the interacting components (competent newt gastrula ectoderm and dorsal lip tissues) were separated by filter membranes of varying pore size. Nuclepore filters with nominal pore size from 0.1 to 8 mum were employed and the neuralising effect was shown to traverse all of these membranes. Electron microscopic examination did not reveal any cytoplasmic processes in the pores and the authors conclude that the morphogenetic signals are carried by transmissable compounds rather than through direct cytoplasmic contacts.